显示样式:     当前期刊: ISPRS Journal of Photogrammetry and Remote Sensing    加入关注       排序: 导出
  • Big earth observation time series analysis for monitoring Brazilian agriculture
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-08-17
    Michelle Cristina Araujo Picoli, Gilberto Camara, Ieda Sanches, Rolf Simões, Alexandre Carvalho, Adeline Maciel, Alexandre Coutinho, Julio Esquerdo, João Antunes, Rodrigo Anzolin Begotti, Damien Arvor, Claudio Almeida

    This paper presents innovative methods for using satellite image time series to produce land use and land cover classification over large areas in Brazil from 2001 to 2016. We used Moderate Resolution Imaging Spectroradiometer (MODIS) time series data to classify natural and human-transformed land areas in the state of Mato Grosso, Brazil’s agricultural frontier. Our hypothesis is that building high-dimensional spaces using all values of the time series, coupled with advanced statistical learning methods, is a robust and efficient approach for land cover classification of large data sets. We used the full depth of satellite image time series to create large dimensional spaces for statistical classification. The data consist of MODIS MOD13Q1 time series with 23 samples per year per pixel, and 4 bands (Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), near-infrared (nir) and mid-infrared (mir)). By taking a series of labelled time series, we fed a 92-dimensional attribute space into a support vector machine model. Using a 5-fold cross validation, we obtained an overall accuracy of 94% for discriminating among nine land cover classes: forest, cerrado, pasture, soybean-fallow, fallow-cotton, soybean-cotton, soybean-corn, soybean-millet, and soybean-sunflower. Producer and user accuracies for all classes were close to or better than 90%. The results highlight important trends in agricultural intensification in Mato Grosso. Double crop systems are now the most common production system in the state, sparing land from agricultural production. Pasture expansion and intensification has been less studied than crop expansion, although it has a stronger impact on deforestation and greenhouse gas (GHG) emissions. Our results point to a significant increase in the stocking rate in Mato Grosso and to the possible abandonment of pasture areas opened in the state’s frontier. The detailed land cover maps contribute to an assessment of the interplay between production and protection in the Brazilian Amazon and Cerrado biomes.

  • Improvements of the MODIS Gross Primary Productivity model based on a comprehensive uncertainty assessment over the Brazilian Amazonia
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-08-02
    Catherine Torres de Almeida, Rafael Coll Delgado, Lênio Sores Galvão, Luiz Eduardo de Oliveira Cruz e Aragão, María Concepción Ramos

    Tropical forests and savannas are responsible for the largest proportion of global Gross Primary Productivity (GPP), a major component of the global carbon cycle. However, there are still deficiencies in the spatial and temporal information of tropical photosynthesis and its relations with environmental controls. The MOD17 product, based on the Light Use Efficiency (LUE) concept, has been updated to provide GPP estimates around the globe. In this research, the MOD17 GPP collections 5.0, 5.5 and 6.0 and their sources of uncertainties were assessed by using measurements of meteorology and eddy covariance GPP from eight flux towers in Brazilian tropical ecosystems, from 2000 to 2006. Results showed that the MOD17 collections tend to overestimate GPP at low productivity sites (bias between 111% and 584%) and underestimate it at high productivity sites (bias between −2% and −18%). Overall, the MOD17 product was not able to capture the GPP seasonality, especially in the equatorial sites. Recalculations of MOD17 GPP using site-specific meteorological data, corrected land use/land cover (LULC) classification, and tower-based LUE parameter showed improvements for some sites. However, the improvements were not sufficient to estimate the GPP seasonality in the equatorial forest sites. The use of a new soil moisture constraint on the LUE, based on the Evaporative Fraction, just showed improvements in water-limited sites. Modifications in the algorithm to account for separate LUE for cloudy and clear sky days presented noticeably improved GPP estimates in the tropical ecosystems investigated, both in magnitude and in seasonality. The results suggest that the high cloudiness makes the diffuse radiation an important factor to be considered in the LUE control, especially over dense forests. Thus, the MOD17 GPP algorithm needs more updates to accurately estimate productivity in tropical ecosystems.

  • Spatial and temporal variation of human appropriation of net primary production in the Rio de la Plata grasslands
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-07-31
    Santiago Baeza, José M. Paruelo

    Latin America, and particularly, the Rio de la Plata Grasslands (RPG), are one of the regions with the highest rates of land use change worldwide. These changes drastically alter ecosystems energy flows, affecting biodiversity, atmospheric composition, and the ecosystem capacity to provide services. In this work we evaluated the impact of these changes on Net Primary Production (NPP), one of the most important and integrative ecosystem attributes, through the calculation of Human Appropriation of NPP (HANPP), a very complete indicator of human impact on ecosystems. Our results provide a comprehensive and fine grained description of HANPP patterns over an entire biogeographycal region for two periods that encompass a strong agricultural intensification process. We used medium resolution land use maps and NPP estimates from sub-national level agricultural statistics and remotely sensed data modeling. Results show that the human impact on the energy flow in RPG ecosystems reached very high levels compared to other regions of the world. The average appropriation of was 42% of the potential vegetation NPP in 2001/2002 and it increased 4.5% during the last years due to an intense land use changes. Most of the HANPP was explained by harvest rather than by land use changes, mainly in the last period due to crops yield increase and the expansion of double crop system as a common agronomic practice. High HANPP values found were associated to a set of environmental impacts that affect ecosystems sustainability and their ability to provide ecosystem services.

  • A simple terrain relief index for tuning slope-related parameters of LiDAR ground filtering algorithms
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-04-02
    Peng Wan, Wuming Zhang, Andrew K. Skidmore, Jianbo Qi, Xiuliang Jin, Guangjian Yan, Tiejun Wang

    Ground filtering is an essential procedure in almost all LiDAR applications. However, most existing ground filtering algorithms require different amounts of user input to manually set up initial parameters, such as terrain relief amplitude and average slope, which is subjective, time consuming, and prone to errors. Here, we propose a simple terrain relief index derived from raw airborne LiDAR data to automatically tune the slope-related parameters of ground filtering algorithms. The terrain relief index is a ratio between the height difference of the entire point cloud and the maximum above ground level of non-ground points. The latter variable can be estimated with the maximum local height difference of raw LiDAR data through gridding. We validated our method using the benchmark airborne LiDAR datasets provided by the International Society for Photogrammetry and Remote Sensing. The results showed a high correlation (r = 0.876) between the terrain relief index and the referential terrain relief amplitude. The degree of correlation was greater across larger areas (r = 0.926) than small areas (r = 0.861) regardless of the type of land cover (e.g., city or forest). The terrain relief index was introduced into two existing filtering algorithms: Cloth Simulation Filtering (CSF) and Progressive Morphological (PM) Filter, by relating the terrain relief index to the cloth rigidness of the CSF and the slope threshold of the PM filter. To compare the results, the two algorithms were implemented both with manually tuned parameters and with the parameters derived from the terrain relief index. The results showed that there was only a slight discrepancy in average Total Error (0.1%) between them in the CSF, which means that the terrain relief index can automatically determine the cloth rigidness without noticeable loss of accuracy. The average difference between the slope threshold provided by the terrain relief index and the manually tuned optimal slope threshold was 0.142 rad (8.136°) for the PM filter, which is acceptable relative to manually parameter setting without any prior knowledge. The terrain relief index can estimate the terrain relief amplitude from raw airborne LiDAR data, and the parameter settings suggested in this paper for the filtering algorithm can improve automation.

  • Monitoring Andean high altitude wetlands in central Chile with seasonal optical data: A comparison between Worldview-2 and Sentinel-2 imagery
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-04-13
    Rocío A. Araya-López, Javier Lopatin, Fabian E. Fassnacht, H. Jaime Hernández

    In the Maipo watershed, situated in central Chile, mining activities are impacting high altitude Andean wetlands through the consumption and exploitation of water and land. As wetlands are vulnerable and particularly susceptible to changes of water supply, alterations and modifications in the hydrological regime have direct effects on their ecophysiological condition and vegetation cover. The aim of this study was to evaluate the potential of Worldview-2 and Sentinel-2 sensors to identify and map Andean wetlands through the use of the one-class classifier Bias support vector machines (BSVM), and then to estimate soil moisture content of the identified wetlands during snow-free summer using partial least square regression.The results obtained in this research showed that the combination of remote sensing data and a small sample of ground reference measurements enables to map Andean high altitude wetlands with high accuracies. BSVM was capable to classify the meadow areas with an overall accuracy of over ∼78% for both sensors. Our results also indicate that it is feasible to map surface soil moisture with optical remote sensing data and simple regression approaches in the examined environment. Surface soil moisture estimates reached r2 values of up to 0.58, and normalized mean square errors of 19% using Sentinel-2 data, while Worldview-2 estimates resulted in non-satisfying results. The presented approach is particularly valuable for monitoring high-mountain wetland areas with limited accessibility such as in the Andes.

  • Semantic line framework-based indoor building modeling using backpacked laser scanning point cloud
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-04-25
    Cheng Wang, Shiwei Hou, Chenglu Wen, Zheng Gong, Qing Li, Xiaotian Sun, Jonathan Li

    Indoor building models are essential in many indoor applications. These models are composed of the primitives of the buildings, such as the ceilings, floors, walls, windows, and doors, but not the movable objects in the indoor spaces, such as furniture. This paper presents, for indoor environments, a novel semantic line framework-based modeling building method using backpacked laser scanning point cloud data. The proposed method first semantically labels the raw point clouds into the walls, ceiling, floor, and other objects. Then line structures are extracted from the labeled points to achieve an initial description of the building line framework. To optimize the detected line structures caused by furniture occlusion, a conditional Generative Adversarial Nets (cGAN) deep learning model is constructed. The line framework optimization model includes structure completion, extrusion removal, and regularization. The result of optimization is also derived from a quality evaluation of the point cloud. Thus, the data collection and building model representation become a united task-driven loop. The proposed method eventually outputs a semantic line framework model and provides a layout for the interior of the building. Experiments show that the proposed method effectively extracts the line framework from different indoor scenes.

  • An analytical approach to evaluate point cloud registration error utilizing targets
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-06-22
    Ronghua Yang, Xiaolin Meng, Yibin Yao, Bi Yu Chen, Yangsheng You, Zejun Xiang

    Point cloud registration is essential for processing terrestrial laser scanning (TLS) point cloud datasets. The registration precision directly influences and determines the practical usefulness of TLS surveys. However, in terms of target based registration, analytical point cloud registration error models employed by scanner manufactures are only suitable to evaluate target registration error, rather than point cloud registration error. This paper proposes an new analytical approach called the registration error (RE) model to directly evaluate point cloud registration error. We verify the proposed model by comparing RE and root mean square error (RMSE) for all points in three point clouds that are approximately equivalent.

  • Removing non-static objects from 3D laser scan data
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-06-19
    Johannes Schauer, Andreas Nüchter

    For the purpose of visualization and further post-processing of 3D point cloud data, it is often desirable to remove moving objects from a given data set. Common examples for these moving objects are pedestrians, bicycles and motor vehicles in outdoor scans or manufactured goods and employees in indoor scans of factories. We present a new change detection method which is able to partition the points of multiple registered 3D scans into two sets: points belonging to stationary (static) objects and points belonging to moving (dynamic) objects. Our approach does not require any object detection or tracking the movement of objects over time. Instead, we traverse a voxel grid to find differences in volumetric occupancy for “explicit” change detection. Our main contribution is the introduction of the concept of “point shadows” and how to efficiently compute them. Without them, using voxel grids for explicit change detection is known to suffer from a high number of false positives when applied to terrestrial scan data. Our solution achieves similar quantitative results in terms of F1-score as competing methods while at the same time being faster.

  • Developing a multi-filter convolutional neural network for semantic segmentation using high-resolution aerial imagery and LiDAR data
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-06-19
    Ying Sun, Xinchang Zhang, Qinchuan Xin, Jianfeng Huang

    Semantic segmentation of LiDAR and high-resolution aerial imagery is one of the most challenging topics in the remote sensing domain. Deep convolutional neural network (CNN) and its derivatives have recently shown the abilities in pixel-wise prediction of remote sensing data. Many existing deep learning methods fuse LiDAR and high-resolution aerial imagery towards an inter-modal mode and thus overlook the intra-modal statistical characteristics. Additionally, the patch-based CNNs could generate the salt-and-pepper artifacts as characterized by isolated and spurious pixels on the object boundaries and patch edges leading to unsatisfied labelling results. This paper presents a semantic segmentation scheme that combines multi-filter CNN and multi-resolution segmentation (MRS). The multi-filter CNN aggregates LiDAR data and high-resolution optical imagery by multi-modal data fusion for semantic labelling, and the MRS is further used to delineate object boundaries for reducing the salt-and-pepper artifacts. The proposed method is validated against two datasets: the ISPRS 2D semantic labelling contest of Potsdam and an area of Guangzhou in China labelled based on existing geodatabases. Various designs of data fusion strategy, CNN architecture and MRS scale are analyzed and discussed. Compared with other classification methods, our method improves the overall accuracies. Experiment results show that our combined method is an efficient solution for the semantic segmentation of LiDAR and high-resolution imagery.

  • In-situ measurements from mobile platforms: An emerging approach to address the old challenges associated with forest inventories
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-06-18
    Xinlian Liang, Antero Kukko, Juha Hyyppä, Matti Lehtomäki, Jiri Pyörälä, Xiaowei Yu, Harri Kaartinen, Anttoni Jaakkola, Yunsheng Wang

    Accurate assessments of forest resources rely on ground truth data that are collected via in-situ measurements, which are fundamental for all other statistical- and/or remote-sensing-based deductions on quantified forest attributes. The major bottleneck of the current in-situ observation system is that the data collection is time consuming, and, thus, limited in extent, which potentially biases any further inferences made. Consequently, conventional field-data-collection approaches can hardly keep pace with the coverage, scale and frequency required for contemporary and future forest inventories. In-situ measurements from mobile platforms seem to be a promising technique to solve this problem and are estimated at least 10 times faster than static techniques (e.g., terrestrial laser scanning, TLS) at the plot level. However, the mobile platforms are still at the very early stages of development, and it is unclear which three-dimensional (3D) forest measurements the mobile systems can provide and at what accuracy. This study presents a quantitative evaluation of the performance of mobile platforms in a variety of forest conditions and through a comparison with state-of-the-art static in-situ observations. Two mobile platforms were used to collect field data, where the same laser-scanning system was both mounted on top of a vehicle and wore by an operator. The static in-situ observation from TLS is used as a baseline for the evaluation. All point clouds involved were processed through the same processing chain and compared to conventional manual measurement. The evaluation results indicate that the mobile platforms can assess homogeneous forests as well as static observations, but they cannot yet assess heterogeneous forest as required by practical applications. The major challenge is twofold: mobile-data coverage and accuracy. Future research should focus on the robust registration techniques between strips, especially in complex forest conditions, since errors of data registration results in significant impacts on tree attributes estimation accuracy. In cases that the spatial inconstancy cannot be eliminated, attributes estimation in single strips, i.e., the multi-single-scan approach, is an alternative. Meanwhile, operator training deserves attention since the data quality from mobile platforms is partly determined by the operators’ selection of trajectory in the field.

  • A semi-supervised generative framework with deep learning features for high-resolution remote sensing image scene classification
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-11-14
    Wei Han, Ruyi Feng, Lizhe Wang, Yafan Cheng

    High resolution remote sensing (HRRS) image scene classification plays a crucial role in a wide range of applications and has been receiving significant attention. Recently, remarkable efforts have been made to develop a variety of approaches for HRRS scene classification, wherein deep-learning-based methods have achieved considerable performance in comparison with state-of-the-art methods. However, the deep-learning-based methods have faced a severe limitation that a great number of manually-annotated HRRS samples are needed to obtain a reliable model. However, there are still not sufficient annotation datasets in the field of remote sensing. In addition, it is a challenge to get a large scale HRRS image dataset due to the abundant diversities and variations in HRRS images. In order to address the problem, we propose a semi-supervised generative framework (SSGF), which combines the deep learning features, a self-label technique, and a discriminative evaluation method to complete the task of scene classification and annotating datasets. On this basis, we further develop an extended algorithm (SSGA-E) and evaluate it by exclusive experiments. The experimental results show that the SSGA-E outperforms most of the fully-supervised methods and semi-supervised methods. It has achieved the third best accuracy on the UCM dataset, the second best accuracy on the WHU-RS, the NWPU-RESISC45, and the AID datasets. The impressive results demonstrate that the proposed SSGF and the extended method is effective to solve the problem of lacking an annotated HRRS dataset, which can learn valuable information from unlabeled samples to improve classification ability and obtain a reliable annotation dataset for supervised learning.

  • A rigorous fastener inspection approach for high-speed railway from structured light sensors
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-11-17
    Qingzhou Mao, Hao Cui, Qingwu Hu, Xiaochun Ren

    Rail fasteners are critical components in high-speed railway. Therefore, they are inspected periodically to ensure the safety of high-speed trains. Manual inspection and two-dimensional visual inspection are the commonly used methods. However, both of them have drawbacks. In this paper, a rigorous high-speed railway fastener inspection approach from structured light sensors is proposed to detect damaged and loose fasteners. Firstly, precise and extremely dense point cloud of fasteners are obtained from commercial structured light sensors. With a decision tree classifier, the defects of the fasteners are classified in detail. Furthermore, a normal vector based center extraction method for complex cylindrical surface is proposed to extract the centerline of the metal clip of normal fasteners. Lastly, the looseness of the fastener is evaluated based on the extracted centerline of the metal clip. Experiments were conducted on high-speed railways to evaluate the accuracy, effectiveness, and the influence of the parameters of the proposed method. The overall precision of the decision tree classifier is over 99.8% and the root-mean-square error of looseness check is 0.15 mm, demonstrating a reliable and effective solution for high-speed railway fastener maintenance.

  • MugNet: Deep learning for hyperspectral image classification using limited samples
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-11-20
    Bin Pan, Zhenwei Shi, Xia Xu

    In recent years, deep learning based methods have attracted broad attention in the field of hyperspectral image classification. However, due to the massive parameters and the complex network structure, deep learning methods may not perform well when only few training samples are available. In this paper, we propose a small-scale data based method, multi-grained network (MugNet), to explore the application of deep learning approaches in hyperspectral image classification. MugNet could be considered as a simplified deep learning model which mainly targets at limited samples based hyperspectral image classification. Three novel strategies are proposed to construct MugNet. First, the spectral relationship among different bands, as well as the spatial correlation within neighboring pixels, are both utilized via a multi-grained scanning approach. The proposed multi-grained scanning strategy could not only extract the joint spectral-spatial information, but also combine different grains’ spectral and spatial relationship. Second, because there are abundant unlabeled pixels available in hyperspectral images, we take full advantage of these samples, and adopt a semi-supervised manner in the process of generating convolution kernels. At last, the MugNet is built upon the basis of a very simple network which does not include many hyperparameters for tuning. The performance of MugNet is evaluated on a popular and two challenging data sets, and comparison experiments with several state-of-the-art hyperspectral image classification methods are revealed.

  • A new deep convolutional neural network for fast hyperspectral image classification
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-12-06
    M.E. Paoletti, J.M. Haut, J. Plaza, A. Plaza

    Artificial neural networks (ANNs) have been widely used for the analysis of remotely sensed imagery. In particular, convolutional neural networks (CNNs) are gaining more and more attention in this field. CNNs have proved to be very effective in areas such as image recognition and classification, especially for the classification of large sets composed by two-dimensional images. However, their application to multispectral and hyperspectral images faces some challenges, especially related to the processing of the high-dimensional information contained in multidimensional data cubes. This results in a significant increase in computation time. In this paper, we present a new CNN architecture for the classification of hyperspectral images. The proposed CNN is a 3-D network that uses both spectral and spatial information. It also implements a border mirroring strategy to effectively process border areas in the image, and has been efficiently implemented using graphics processing units (GPUs). Our experimental results indicate that the proposed network performs accurately and efficiently, achieving a reduction of the computation time and increasing the accuracy in the classification of hyperspectral images when compared to other traditional ANN techniques.

  • Semantic labeling in very high resolution images via a self-cascaded convolutional neural network
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-12-21
    Yongcheng Liu, Bin Fan, Lingfeng Wang, Jun Bai, Shiming Xiang, Chunhong Pan

    Semantic labeling for very high resolution (VHR) images in urban areas, is of significant importance in a wide range of remote sensing applications. However, many confusing manmade objects and intricate fine-structured objects make it very difficult to obtain both coherent and accurate labeling results. For this challenging task, we propose a novel deep model with convolutional neural networks (CNNs), i.e., an end-to-end self-cascaded network (ScasNet). Specifically, for confusing manmade objects, ScasNet improves the labeling coherence with sequential global-to-local contexts aggregation. Technically, multi-scale contexts are captured on the output of a CNN encoder, and then they are successively aggregated in a self-cascaded manner. Meanwhile, for fine-structured objects, ScasNet boosts the labeling accuracy with a coarse-to-fine refinement strategy. It progressively refines the target objects using the low-level features learned by CNN’s shallow layers. In addition, to correct the latent fitting residual caused by multi-feature fusion inside ScasNet, a dedicated residual correction scheme is proposed. It greatly improves the effectiveness of ScasNet. Extensive experimental results on three public datasets, including two challenging benchmarks, show that ScasNet achieves the state-of-the-art performance.

  • A deep learning framework for remote sensing image registration
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-01-05
    Shuang Wang, Dou Quan, Xuefeng Liang, Mengdan Ning, Yanhe Guo, Licheng Jiao

    We propose an effective deep neural network aiming at remote sensing image registration problem. Unlike conventional methods doing feature extraction and feature matching separately, we pair patches from sensed and reference images, and then learn the mapping directly between these patch-pairs and their matching labels for later registration. This end-to-end architecture allows us to optimize the whole processing (learning mapping function) through information feedback when training the network, which is lacking in conventional methods. In addition, to alleviate the small data issue of remote sensing images for training, our proposal introduces a self-learning by learning the mapping function using images and their transformed copies. Moreover, we apply a transfer learning to reduce the huge computation cost in the training stage. It does not only speed up our framework, but also get extra performance gains. The comprehensive experiments conducted on seven sets of remote sensing images, acquired by Radarsat, SPOT and Landsat, show that our proposal improves the registration accuracy up to 2.4–53.7%.

  • Geoinformatics for the conservation and promotion of cultural heritage in support of the UN Sustainable Development Goals
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-01-12
    Wen Xiao, Jon Mills, Gabriele Guidi, Pablo Rodríguez-Gonzálvez, Sara Gonizzi Barsanti, Diego González-Aguilera

    Cultural Heritage (CH) is recognised as being of historical, social, and anthropological value and is considered as an enabler of sustainable development. As a result, it is included in the United Nations’ Sustainable Development Goals (SDGs) 11 and 8. SDG 11.4 emphasises the protection and safeguarding of heritage, and SDG 8.9 aims to promote sustainable tourism that creates jobs and promotes local culture and products. This paper briefly reviews the geoinformatics technologies of photogrammetry, remote sensing, and spatial information science and their application to CH. Detailed aspects of CH-related SDGs, comprising protection and safeguarding, as well as the promotion of sustainable tourism are outlined. Contributions of geoinformatics technologies to each of these aspects are then identified and analysed. Case studies in both developing and developed countries, supported by funding directed at the UN SDGs, are presented to illustrate the challenges and opportunities of geoinformatics to enhance CH protection and to promote sustainable tourism. The potential and impact of geoinformatics for the measurement of official SDG indicators, as well as UNESCO’s Culture for Development Indicators, are discussed. Based on analysis of the review and the presented case studies, it is concluded that the contribution of geoinformatics to the achievement of CH SDGs is necessary, significant and evident. Moreover, following the UNESCO initiative to introduce CH into the sustainable development agenda and related ICOMOS action plan, the concept of Sustainable Cultural Heritage is defined, reflecting the significance of CH to the United Nations’ ambition to “transform our world”.

  • 更新日期:2018-06-03
  • One-two-one networks for compression artifacts reduction in remote sensing
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-02-17
    Baochang Zhang, Jiaxin Gu, Chen Chen, Jungong Han, Xiangbo Su, Xianbin Cao, Jianzhuang Liu

    Compression artifacts reduction (CAR) is a challenging problem in the field of remote sensing. Most recent deep learning based methods have demonstrated superior performance over the previous hand-crafted methods. In this paper, we propose an end-to-end one-two-one (OTO) network, to combine different deep models, i.e., summation and difference models, to solve the CAR problem. Particularly, the difference model motivated by the Laplacian pyramid is designed to obtain the high frequency information, while the summation model aggregates the low frequency information. We provide an in-depth investigation into our OTO architecture based on the Taylor expansion, which shows that these two kinds of information can be fused in a nonlinear scheme to gain more capacity of handling complicated image compression artifacts, especially the blocking effect in compression. Extensive experiments are conducted to demonstrate the superior performance of the OTO networks, as compared to the state-of-the-arts on remote sensing datasets and other benchmark datasets. The source code will be available here: https://github.com/bczhangbczhang/.

  • Pan-sharpening via deep metric learning
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-02-17
    Yinghui Xing, Min Wang, Shuyuan Yang, Licheng Jiao

    Neighbors Embedding based pansharpening methods have received increasing interests in recent years. However, image patches do not strictly follow the similar structure in the shallow MultiSpectral (MS) and PANchromatic (PAN) image spaces, consequently leading to a bias to the pansharpening. In this paper, a new deep metric learning method is proposed to learn a refined geometric multi-manifold neighbor embedding, by exploring the hierarchical features of patches via multiple nonlinear deep neural networks. First of all, down-sampled PAN images from different satellites are divided into a large number of training image patches and are then grouped coarsely according to their shallow geometric structures. Afterwards, several Stacked Sparse AutoEncoders (SSAE) with similar structures are separately constructed and trained by these grouped patches. In the fusion, image patches of the source PAN image pass through the networks to extract features for formulating a deep distance metric and thus deriving their geometric labels. Then, patches with the same geometric labels are grouped to form geometric manifolds. Finally, the assumption that MS patches and PAN patches form the same geometric manifolds in two distinct spaces, is cast on geometric groups to formulate geometric multi-manifold embedding for estimating high resolution MS image patches. Some experiments are taken on datasets acquired by different satellites. The experimental results demonstrate that our proposed method can obtain better fusion results than its counterparts in terms of visual results and quantitative evaluations.

  • Land cover mapping at very high resolution with rotation equivariant CNNs: Towards small yet accurate models
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-02-19
    Diego Marcos, Michele Volpi, Benjamin Kellenberger, Devis Tuia

    In remote sensing images, the absolute orientation of objects is arbitrary. Depending on an object’s orientation and on a sensor’s flight path, objects of the same semantic class can be observed in different orientations in the same image. Equivariance to rotation, in this context understood as responding with a rotated semantic label map when subject to a rotation of the input image, is therefore a very desirable feature, in particular for high capacity models, such as Convolutional Neural Networks (CNNs). If rotation equivariance is encoded in the network, the model is confronted with a simpler task and does not need to learn specific (and redundant) weights to address rotated versions of the same object class. In this work we propose a CNN architecture called Rotation Equivariant Vector Field Network (RotEqNet) to encode rotation equivariance in the network itself. By using rotating convolutions as building blocks and passing only the values corresponding to the maximally activating orientation throughout the network in the form of orientation encoding vector fields, RotEqNet treats rotated versions of the same object with the same filter bank and therefore achieves state-of-the-art performances even when using very small architectures trained from scratch. We test RotEqNet in two challenging sub-decimeter resolution semantic labeling problems, and show that we can perform better than a standard CNN while requiring one order of magnitude less parameters.

  • Slavery from Space: Demonstrating the role for satellite remote sensing to inform evidence-based action related to UN SDG number 8
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-03-02
    Doreen S. Boyd, Bethany Jackson, Jessica Wardlaw, Giles M. Foody, Stuart Marsh, Kevin Bales

    The most recent Global Slavery Index estimates that there are 40.3 million people enslaved globally. The UN’s Agenda 2030 for Sustainable Development Goal number 8, section 8.7 specifically refers to the issue of forced labour: ending modern slavery and human trafficking, including child labour, in all forms by 2025. Although there is a global political commitment to ending slavery, one of the biggest barriers to doing so is having reliable and timely, spatially explicit and scalable data on slavery activity. The lack of these data compromises evidence-based action and policy formulation. Thus, to meet the challenge of ending modern slavery new and innovative approaches, with an emphasis on efficient use of resources (including financial) are needed. This paper demonstrates the fundamental role of remote sensing as a source of evidence. We provide an estimate of the number of brick kilns across the ‘Brick Belt’ that runs across south Asia. This is important because these brick kilns are known sites of modern-day slavery. This paper reports the first rigorous estimate of the number of brick kilns present and does so using a robust method that can be easily adopted by key agencies for evidence-based action (i.e. NGOs, etc.) and is based on freely available and accessible remotely sensed data. From this estimate we can not only calculate the scale of the slavery problem in the Brick Belt, but also calculate the impact of slavery beyond that of the enslaved people themselves, on, for example, environmental change and impacts on ecosystem services – this links to other Sustainable Development Goals. As the process of achieving key Sustainable Development Goal targets will show, there are global benefits to ending slavery - this will mean a better world for everyone: safer, greener, more prosperous, and more equal. This is termed here a Freedom Dividend.

  • Building instance classification using street view images
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-03-02
    Jian Kang, Marco Körner, Yuanyuan Wang, Hannes Taubenböck, Xiao Xiang Zhu
  • Remote sensing monitoring of the impact of a major mining wastewater disaster on the turbidity of the Doce River plume off the eastern Brazilian coast
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-03-02
    Natalia Rudorff, Conrado M. Rudorff, Milton Kampel, Gustavo Ortiz

    In November 2015, the Doce River basin was struck with Brazil’s worst environmental disaster: the collapse of an iron ore mining waste water dam delivering about 60 Mm3 of contaminated mud that traveled from its headwaters to the Atlantic Ocean. We analyzed the impact of the tailings dam failure on the turbidity of the Doce River plume and coastal waters based on remote sensing methods applied to Landsat and MODIS-Aqua imagery. Turbidity maps were obtained using a semi-analytical algorithm that uses a two band (i.e., red and near infrared) selection scheme to optimize the retrieval for a wide range of turbidity levels. The dam failure occurred in the context of a severe hydrological drought, but despite the abnormally low streamflow the turbidity front of the Doce River plume at the coast peaked to levels above 1000 FNU after the disaster. With subsequent rainfall-runoff events, the deposited mud within the river-estuary system continued to be washed out and cause high turbidity levels with extensive plumes over the coastal zone. The high turbidity (>20 FNU) plume front reached 11 km off the coast towards the southern shelf, the moderate turbidity plume (10–20 FNU) reached 39 km, and the low turbidity plume (<5 FNU) exceeded 75 km. Changes on the dominant wind pattern from NE to S, inverted the river plume direction towards the northern shelf, but the high turbidity plume stayed near the mouth (<5 km). The mud deposited from the disaster in the inner shelf remained a source of fine sediment resuspension for the coastal front turbidity during 2016 and 2017, with the intensification of coastal resuspension processes during the winter season. The combination of higher spatial and temporal resolutions of Landsat and MODIS-Aqua imagery, respectively, with auxiliary in situ stream flow and turbidity measurements allowed disentangling the impacts caused by the tailings dam failure from the natural sources of suspended sediment that influence the variability of coastal water turbidity.

  • Geometrically stable tracking for depth images based 3D reconstruction on mobile devices
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-03-14
    Yangdong Liu, Wei Gao, Zhanyi Hu

    With the development of hardwares such as mobile devices and portable depth cameras, on-line 3D reconstruction on the mobile devices with depth streams as input turns to be possible and promising. Most existing systems use volumetric representation methods to fuse the depth images and use ICP algorithm to estimate the poses of cameras. However, ICP tracker suffers from large drift in scenes containing insufficient geometric information. To deal with this problem, we propose a stability based sampling method which select different number of point-pairs in different windows according to their geometric stability. In addition, we fuse the ICP tracker with the IMU information through an analysis of the condition number. Then we apply the stability based sampling method to the spatially hashed volumetric representation. Qualitative and quantitative evaluations of tracking accuracy and 3D reconstruction results show that our method outperforms the current state-of-the-art systems, especially in scenes lacking sufficient geometric information. In total, our method achieves frame rates 20 Hz on an Apple iPad Air 2 and 200 Hz on a Nvidia GeForce GTX 1060 GPU.

  • Early assessment of crop yield from remotely sensed water stress and solar radiation data
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-03-20
    Mauro E. Holzman, Facundo Carmona, Raúl Rivas, Raquel Niclòs
  • Refinement of LiDAR point clouds using a super voxel based approach
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-03-21
    Minglei Li, Changming Sun

    We propose a new approach for automatic refinement of unorganized point clouds captured by LiDAR scanning systems. Given a point cloud, our method first abstracts the input data into super voxels via over segmentations, and then builds a K-nearest neighbor graph on these voxel nodes. Abstracting into voxel representation provides a means to generate an elastic wireframe over the original data. An iterative resampling method is then introduced to project resampling points to all potential surfaces considering repulsion constraints from both interior and exterior of voxels. Our point consolidation process contributes to accurate normal estimation, uniform point distribution, and sufficient sampling density. Experiments and comparisons have demonstrated that the proposed method is effective on point clouds from a variety of datasets.

  • Monitoring thirty years of small water reservoirs proliferation in the southern Brazilian Amazon with Landsat time series
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-03-23
    Damien Arvor, Felipe R.G. Daher, Dominique Briand, Simon Dufour, Anne-Julia Rollet, Margareth Simões, Rodrigo P.D. Ferraz

    The recent decoupling of agricultural production and deforestation in the southern Amazon has been made possible thanks to (1) the adoption of intensive agricultural practices, including irrigation, and (2) the diversification of economic activities, including fish farming. Whereas this new agricultural model has brought out positive results to contain deforestation, it also implied new pressures on the environment, and especially on water resources. Many small artificial water reservoirs have been built with different uses, e.g. crop irrigation, energy generation, fish farming or livestock watering. In this paper, we introduce a method to automatically map small water bodies based on time series of Landsat images. The method was tested in the municipality of Sorriso (state of Mato Grosso, Brazil). The statistical results (Overall Accuracy = 0.872; Kappa index = 0.745) validated the efficiency of the methodology although the spatial resolution of Landsat images limited the detection of very small and linear reservoirs. In Sorriso, we estimated that the cumulated area and the number of small water reservoirs increased more than tenfold (from 153 to 1707 ha) and fivefold (86 to 522), respectively, between 1985 and 2015. We discuss the numerous socio-environmental implications raised by the cumulated impacts of these proliferating small reservoirs. We conclude that integrated whole-landscape approaches are necessary to assess how anthropized hydrosystems can counteract or exacerbate the socio-environmental impacts of deforestation and intensive agriculture.

  • Deep fusion of multi-view and multimodal representation of ALS point cloud for 3D terrain scene recognition
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-03-26
    Nannan Qin, Xiangyun Hu, Hengming Dai

    Terrain scene category is useful not only for some geographical or environmental researches, but also for choosing suitable algorithms or proper parameters of the algorithms for several point cloud processing tasks to achieve better performance. However, there are few studies in point cloud processing focusing on terrain scene classification at present. In this paper, a novel deep learning framework for 3D terrain scene recognition using 2D representation of sparse point cloud is proposed. The framework has two key components. (1) Initially, several suitable discriminative low-level local features are extracted from airborne laser scanning point cloud, and 3D terrain scene is encoded into multi-view and multimodal 2D representation. (2) A two-level fusion network embedded with feature- and decision-level fusion strategy is designed to fully exploit the 2D representation of 3D terrain scene, which can be trained end-to-end. Experiment results show that our method achieves an overall accuracy of 96.70% and a kappa coefficient of 0.96 in recognizing nine categories of terrain scene point clouds. Extensive design choices of the underlying framework are tested, and other typical methods from literature for related research are compared.

  • Physical resources assessment in a semi-arid watershed: An integrated methodology for sustainable land use planning
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-03-27
    Karuppusamy Balasubramani

    The study demonstrates the application of geospatial technologies to evaluate physical resources of semi-arid watersheds and presents a comprehensive methodology applicable elsewhere. The selected Andipatti watershed, located in Theni district in the State of Tamil Nadu (India), is known for agricultural activities; however, haphazard planning, management practices and inadequate investments result in land and water resource degradation. Since most of the agricultural lands in developing countries are similar to these conditions, the present study is attempted as a case to develop a framework to assess the land and water resources potential, utilisation level and land suitability for agriculture; and to evolve better management strategies. The physical characteristics of the watershed were studied based on in-situ, remotely sensed and secondary data sources. Thematic layers were generated with the combination of remote sensing, image processing and GIS techniques. In order to characterize and quantify the watershed based on soil erosion and surface runoff rates, the revised universal soil loss equation (RUSLE) and natural resources conservation services curve number (NRCS-CN) were utilized. Data on water levels and geochemistry of water samples, collected from 36 dug wells were also utilized for this study. Sodium adsorption ratio (SAR) and electrical conductivity, as formulated by the US Salinity Laboratory (USSL) were utilized to examine the suitability of groundwater for irrigation purpose. The storie index has been used to assess the productivity of land using profile and textural characteristics of the soil. Keeping Food and Agricultural Organisation (FAO) guidelines as a reference, as many as 727 homogenous micro-land units were prepared. The physical land qualities and characteristics of each land unit were compared with the requirements of 13 major crops of the study area and suitable crops for each unit were identified. The individual suitability classes of all crops were compared using logical analysis and suitability crops for each land unit were determined under irrigated and rain-fed conditions. In order to integrate the results of these analyses and to suggest sustainable agricultural development measures, the study area was divided into 44 micro-watersheds. The information on land productivity, groundwater quality and existing land use/land cover patterns of the watershed were used to calculate land potential-utilisation index and groundwater potential-utilisation ratio for all micro-watersheds. All the results of land and water resources assessment were compared and a proposed land use map was prepared. The findings suggest strategies for coping with sustainable agricultural practices for the present study area and provide an integrated methodology for future assessments elsewhere, especially in the developing countries.

  • Prioritized multi-view stereo depth map generation using confidence prediction
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-03-31
    Christian Mostegel, Friedrich Fraundorfer, Horst Bischof
  • PatternNet: A benchmark dataset for performance evaluation of remote sensing image retrieval
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-25
    Weixun Zhou, Shawn Newsam, Congmin Li, Zhenfeng Shao

    Benchmark datasets are critical for developing, evaluating, and comparing remote sensing image retrieval (RSIR) approaches. However, current benchmark datasets are deficient in that (1) they were originally collected for land use/land cover classification instead of RSIR; (2) they are relatively small in terms of the number of classes as well as the number of images per class which makes them unsuitable for developing deep learning based approaches; and (3) they are not appropriate for RSIR due to the large amount of background present in the images. These limitations restrict the development of novel approaches for RSIR, particularly those based on deep learning which require large amounts of training data. We therefore present a new large-scale remote sensing dataset termed “PatternNet” that was collected specifically for RSIR. PatternNet was collected from high-resolution imagery and contains 38 classes with 800 images per class. Significantly, PatternNet’s large scale makes it suitable for developing novel, deep learning based approaches for RSIR. We use PatternNet to evaluate the performance of over 35 RSIR methods ranging from traditional handcrafted feature based methods to recent, deep learning based ones. These results serve as a baseline for future research on RSIR.

  • A probabilistic graphical model for the classification of mobile LiDAR point clouds
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-24
    Zhizhong Kang, Juntao Yang

    Mobile Light Detection And Ranging (LiDAR) point clouds have the characteristics of complex and incomplete scenes, uneven point density and noises, which raises great challenges for automatically interpreting 3D scene. Aiming at the problem of 3D point cloud classification, we propose a probabilistic graphical model for automatic classification of mobile LiDAR point clouds in this paper. First, the super-voxels are generated as primitives based on the similar geometric and radiometric properties. Second, we construct point-based multi-scale visual features that are used to describe the texture information at various scales. Third, the topic model is used to analyze the semantic correlations among points within super-voxels to establish the semantic representation, which is finally fed into the proposed probabilistic graphical model. The proposed model combines Bayesian network and Markov random fields to obtain locally continuous and globally optimal classification results. To evaluate the effectiveness and the robustness of the proposed method, experiments were conducted using mobile LiDAR point clouds for three types of street scenes. Experimental results demonstrate that our proposed model is efficient and robust for extracting vehicles, buildings, street trees and pole-like objects, with overall accuracies of 98.17%, 97.41% and 96.81% respectively. Moreover, compared with other existing methods, our proposed model can provide higher classification correctness, specifically for small objects such as cars and pole-like objects.

  • Seasonal and interannual assessment of cloud cover and atmospheric constituents across the Amazon (2000–2015): Insights for remote sensing and climate analysis
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-24
    Vitor S. Martins, Evlyn M.L.M. Novo, Alexei Lyapustin, Luiz E.O.C. Aragão, Saulo R. Freitas, Claudio C.F. Barbosa

    The quantitative assessment of cloud cover and atmospheric constituents improves our ability to exploit the climate feedback into the Amazon basin. In the 21st century, three droughts have already occurred in the Amazonia (e.g. 2005, 2010, 2015), inducing regional changes in the seasonal patterns of atmospheric constituents. In addition to climate, the atmospheric dynamic and attenuation properties are long-term challenges for satellite-based remote sensing of this ecosystem: high cloudiness, abundant water vapor content and biomass burning season. Therefore, while climatology analysis supports the understanding of atmospheric variability and trends, it also offers valuable insights for remote sensing applications. In this study, we evaluate the seasonal and interannual variability of cloud cover and atmospheric constituents (aerosol loading, water vapor and ozone content) over the Amazon basin, with focus on both climate analysis and remote sensing implications. We take the advantage of new atmosphere daily products at 1 km resolution derived from Multi-Angle Implementation for Atmospheric Correction (MAIAC) algorithm developed for Moderate Resolution Imaging Spectroradiometer (MODIS) data. An intercomparison of Aerosol Robotic Network (AERONET) and MAIAC aerosol optical depth (AOD) and columnar water vapor (CWV) showed quantitative information with a correlation coefficient higher than 0.81. Our results show distinct regional patterns of cloud cover across the Amazon basin: northwestern region presets a persistent cloud cover (>80%) throughout the year, while low cloud cover (0–20%) occurs in the southern Amazon during the dry season. The cloud-free period in the southern Amazon is followed by an increase in the atmospheric burden due to fire emissions. Our results reveal that AOD records are changing in terms of area and intensity. During the 2005 and 2010 droughts, the positive AOD anomalies (δ > 0.1) occurred over 39.03% (240.3 million ha) and 27.14% (165.99 million ha) of total basin in the SON season, respectively. In contrast, the recent 2015 drought occurred towards the end of year (October through December) and these anomalies were observed over 23.72% (145 million ha) affecting areas in the central and eastern Amazon – unlike previous droughts. The water vapor presents high concentration values (4.0–5.0 g cm−2) in the wet season (DJF), while we observed a strong spatial gradient from northwestern to southeastern of the basin during the dry season. In addition, we also found a positive trend of water vapor content (∼0.3 g/cm2) between 2000 and 2015. The total ozone typically varies between 220 and 270 DU, and it has a seasonal change of ∼25–35 DU from wet season to dry season caused by large emissions of ozone precursors and long-range transport. Finally, while this study contributes to climatological analysis of atmospheric constituents, the remote sensing users can also understand the regional constraints caused by atmospheric attenuation, such as high aerosol loading and cloud obstacles for surface observations.

  • Toward better boundary preserved supervoxel segmentation for 3D point clouds ☆
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-23
    Yangbin Lin, Cheng Wang, Dawei Zhai, Wei Li, Jonathan Li

    Supervoxels provide a more natural and compact representation of three dimensional point clouds, and enable the operations to be performed on regions rather than on the scattered points. Many state-of-the-art supervoxel segmentation methods adopt fixed resolution for each supervoxel, and rely on the initialization of seed points. As a result, they may not preserve well the boundaries of the point cloud with a non-uniform density. In this paper, we present a simple but effective supervoxel segmentation method for point clouds, which formalizes supervoxel segmentation as a subset selection problem. We develop an heuristic algorithm that utilizes local information to efficiently solve the subset selection problem. The proposed method can produce supervoxels with adaptive resolutions, and dose not rely the selection of seed points. The method is fully tested on three publicly available point cloud segmentation benchmarks, which cover the major point cloud types. The experimental results show that compared with the state-of-the-art supervoxel segmentation methods, the supervoxels extracted using our method preserve the object boundaries and small structures more effectively, which is reflected in a higher boundary recall and lower under-segmentation error.

  • Migration monitoring of Ascia monuste (Lepidoptera) and Schistocerca cancellata (Orthoptera) in Argentina using RMA1 weather radar
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-23
    Denis A. Poffo, Hernán M. Beccacece, Giorgio M. Caranti, Raúl A. Comes, María E. Drewniak, Agustín Martina, Adriana I. Zapata, Andres Rodriguez, Jorge N. Saffe

    The meteorological polarimetric radar RMA1 located in the city of Córdoba was used for a nonconventional phenomenon detection. Massive migrations of both Ascia monuste during early summer of 2015 and Schistocerca cancellata during late winter of 2017 were characterized by means of polarimetric variables (correlation factor ρhv ρ hv and reflectivity factor ZH Z H ). The butterfly swarms show a pulsating behavior as a consequence of biological needs. The highest altitude detected was 2400 m msl. The correlation factor confirms the biological characteristic of the echo. The locust swarm migration shows a different pattern in several ways. First, it has a more uniform aspect regarding its displacement. Second, the locusts were observed to attain altitudes of 1700 m msl. Third, the correlation coefficient for the locust case showed regions with high values, which are different from the low value areas. It is concluded that radar observations of insect species may result in useful biological criteria for the government to asses areas that need to be protected for agricultural production.

  • A multi-scale fully convolutional network for semantic labeling of 3D point clouds
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-16
    Mohammed Yousefhussien, David J. Kelbe, Emmett J. Ientilucci, Carl Salvaggio

    When classifying point clouds, a large amount of time is devoted to the process of engineering a reliable set of features which are then passed to a classifier of choice. Generally, such features – usually derived from the 3D-covariance matrix – are computed using the surrounding neighborhood of points. While these features capture local information, the process is usually time-consuming and requires the application at multiple scales combined with contextual methods in order to adequately describe the diversity of objects within a scene. In this paper we present a novel 1D-fully convolutional network that consumes terrain-normalized points directly with the corresponding spectral data (if available) to generate point-wise labeling while implicitly learning contextual features in an end-to-end fashion. This unique approach allows us to operate on unordered point sets with varying densities, without relying on expensive hand-crafted features; thus reducing the time needed for testing by an order of magnitude over existing approaches. Our method uses only the 3D-coordinates and three corresponding spectral features for each point. Spectral features may either be extracted from 2D-georeferenced images, as shown here for Light Detection and Ranging (LiDAR) point clouds, or extracted directly for passive-derived point clouds, i.e. from multiple-view imagery. We train our network by splitting the data into square regions and use a pooling layer that respects the permutation-invariance of the input points. Evaluated using the ISPRS 3D Semantic Labeling Contest, our method scored second place with an overall accuracy of 81.6%. We ranked third place with a mean F1-score of 63.32%, surpassing the F1-score of the method with highest accuracy by 1.69%. In addition to labeling 3D-point clouds, we also show that our method can be easily extended to 2D-semantic segmentation tasks, with promising initial results.

  • Three-dimensional building façade segmentation and opening area detection from point clouds
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-09
    S.M. Iman Zolanvari, Debra F. Laefer, Atteyeh S. Natanzi

    Laser scanning generates a point cloud from which geometries can be extracted, but most methods struggle to do this automatically, especially for the entirety of an architecturally complex building (as opposed to that of a single façade). To address this issue, this paper introduces the Improved Slicing Method (ISM), an innovative and computationally-efficient method for three-dimensional building segmentation. The method is also able to detect opening boundaries even on roofs (e.g. chimneys), as well as a building’s overall outer boundaries using a local density analysis technique. The proposed procedure is validated by its application to two architecturally complex, historic brick buildings. Accuracies of at least 86% were achieved, with computational times as little as 0.53 s for detecting features from a data set of 5.0 million points. The accuracy more than rivalled the current state of the art, while being up to six times faster and with the further advantage of requiring no manual intervention or reliance on a priori information.

  • Ancient Chinese architecture 3D preservation by merging ground and aerial point clouds
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-08
    Xiang Gao, Shuhan Shen, Yang Zhou, Hainan Cui, Lingjie Zhu, Zhanyi Hu

    Ancient Chinese architecture 3D digitalization and documentation is a challenging task for the image based modeling community due to its architectural complexity and structural delicacy. Currently, an effective approach to ancient Chinese architecture 3D reconstruction is to merge the two point clouds, separately obtained from ground and aerial images by the SfM technique. There are two understanding issues should be specially addressed: (1) it is difficult to find the point matches between the images from different sources due to their remarkable variations in viewpoint and scale; (2) due to the inevitable drift phenomenon in any SfM reconstruction process, the resulting two point clouds are no longer strictly related by a single similarity transformation as it should be theoretically. To address these two issues, a new point cloud merging method is proposed in this work. Our method has the following characteristics: (1) the images are matched by leveraging sparse mesh based image synthesis; (2) the putative point matches are filtered by geometrical consistency check and geometrical model verification; and (3) the two point clouds are merged via bundle adjustment by linking the ground-to-aerial tracks. Extensive experiments show that our method outperforms many of the state-of-the-art approaches in terms of ground-to-aerial image matching and point cloud merging.

  • Automatic 3D reconstruction of electrical substation scene from LiDAR point cloud
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-08
    Qiaoyun Wu, Hongbin Yang, Mingqiang Wei, Oussama Remil, Bo Wang, Jun Wang

    3D reconstruction of a large-scale electrical substation scene (ESS) is fundamental to navigation, information inquiry, and supervisory control of 3D scenes. However, automatic reconstruction of ESS from a raw LiDAR point cloud is challenging due to its incompleteness, noise and anisotropy in density. We propose an automatic and efficient approach to reconstruct ESSs, by mapping raw LiDAR data to our well-established electrical device database (EDD). We derive a flexible and hierarchical representation of the ESS automatically by exploring the internal topology of the corresponding LiDAR data, followed by extracting various devices from the ESS. For each device, a quality mesh model is retrieved in the EDD, based on the proposed object descriptor that can balance descriptiveness, robustness and efficiency. With the high-level representation of the ESS, we map all retrieved models into raw data to achieve a high-fidelity scene reconstruction. Extensive experiments on large and complex ESSs modeling demonstrate the efficiency and accuracy of the proposed method.

  • Multi-scale object detection in remote sensing imagery with convolutional neural networks
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-02
    Zhipeng Deng, Hao Sun, Shilin Zhou, Juanping Zhao, Lin Lei, Huanxin Zou

    Automatic detection of multi-class objects in remote sensing images is a fundamental but challenging problem faced for remote sensing image analysis. Traditional methods are based on hand-crafted or shallow-learning-based features with limited representation power. Recently, deep learning algorithms, especially Faster region based convolutional neural networks (FRCN), has shown their much stronger detection power in computer vision field. However, several challenges limit the applications of FRCN in multi-class objects detection from remote sensing images: (1) Objects often appear at very different scales in remote sensing images, and FRCN with a fixed receptive field cannot match the scale variability of different objects; (2) Objects in large-scale remote sensing images are relatively small in size and densely peaked, and FRCN has poor localization performance with small objects; (3) Manual annotation is generally expensive and the available manual annotation of objects for training FRCN are not sufficient in number. To address these problems, this paper proposes a unified and effective method for simultaneously detecting multi-class objects in remote sensing images with large scales variability. Firstly, we redesign the feature extractor by adopting Concatenated ReLU and Inception module, which can increases the variety of receptive field size. Then, the detection is preformed by two sub-networks: a multi-scale object proposal network (MS-OPN) for object-like region generation from several intermediate layers, whose receptive fields match different object scales, and an accurate object detection network (AODN) for object detection based on fused feature maps, which combines several feature maps that enables small and densely packed objects to produce stronger response. For large-scale remote sensing images with limited manual annotations, we use cropped image blocks for training and augment them with re-scalings and rotations. The quantitative comparison results on the challenging NWPU VHR-10 data set, aircraft data set, Aerial-Vehicle data set and SAR-Ship data set show that our method is more accurate than existing algorithms and is effective for multi-modal remote sensing images.

  • Algorithms for semantic segmentation of multispectral remote sensing imagery using deep learning
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-04-30
    Ronald Kemker, Carl Salvaggio, Christopher Kanan

    Deep convolutional neural networks (DCNNs) have been used to achieve state-of-the-art performance on many computer vision tasks (e.g., object recognition, object detection, semantic segmentation) thanks to a large repository of annotated image data. Large labeled datasets for other sensor modalities, e.g., multispectral imagery (MSI), are not available due to the large cost and manpower required. In this paper, we adapt state-of-the-art DCNN frameworks in computer vision for semantic segmentation for MSI imagery. To overcome label scarcity for MSI data, we substitute real MSI for generated synthetic MSI in order to initialize a DCNN framework. We evaluate our network initialization scheme on the new RIT-18 dataset that we present in this paper. This dataset contains very-high resolution MSI collected by an unmanned aircraft system. The models initialized with synthetic imagery were less prone to over-fitting and provide a state-of-the-art baseline for future work.

  • L0-regularization-based skeleton optimization from consecutive point sets of kinetic human body
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-05-01
    Yong Zhang, Bowei Shen, Shaofan Wang, Dehui Kong, Baocai Yin

    Human skeleton extraction is essential for shape abstraction, estimation and analysis. However, it is difficult to implement with the existence of sparse data or noise and the shortage of connectivity within point clouds. To tackle this problem, we propose L0-regularization-based skeleton optimization method from consecutive point sets of kinetic human body. We firstly give an initial reconstruction of a dense point cloud from multi-view human motion images, and extract L1 L 1 -medial skeleton from each point set individually, and then partition all skeleton points into semantic components, from which the partitioned point set is then sampled into skeleton sequence. By further observing that consecutive frames reflecting same body actions may present similar moving trajectories, we build geometric correlations spatiotemporally between adjacent frames. To be specific, our method proposes a temporal constraint and a spatial constraint, where the first constraint considers not only the correlations between each frame and the others, but also the correlations between adjacent frames, and the second one depicts the correlation within the same skeleton block and within the joint points that between different blocks to prevent the non-equidistant distribution of the skeleton points. By integrating the above spatio-temporal constraints, we establish a sparse optimization model and apply L0 optimization to all point sets of different frames. Experimental results show that our method can recover missing skeleton points, correct outliers in skeletons and smooth skeletons in the process of movement while retaining the action features of these skeletons.

  • Semantic line framework-based indoor building modeling using backpacked laser scanning point cloud
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-04-25
    Cheng Wang, Shiwei Hou, Chenglu Wen, Zheng Gong, Qing Li, Xiaotian Sun, Jonathan Li

    Indoor building models are essential in many indoor applications. These models are composed of the primitives of the buildings, such as the ceilings, floors, walls, windows, and doors, but not the movable objects in the indoor spaces, such as furniture. This paper presents, for indoor environments, a novel semantic line framework-based modeling building method using backpacked laser scanning point cloud data. The proposed method first semantically labels the raw point clouds into the walls, ceiling, floor, and other objects. Then line structures are extracted from the labeled points to achieve an initial description of the building line framework. To optimize the detected line structures caused by furniture occlusion, a conditional Generative Adversarial Nets (cGAN) deep learning model is constructed. The line framework optimization model includes structure completion, extrusion removal, and regularization. The result of optimization is also derived from a quality evaluation of the point cloud. Thus, the data collection and building model representation become a united task-driven loop. The proposed method eventually outputs a semantic line framework model and provides a layout for the interior of the building. Experiments show that the proposed method effectively extracts the line framework from different indoor scenes.

  • Monitoring Andean high altitude wetlands in central Chile with seasonal optical data: A comparison between Worldview-2 and Sentinel-2 imagery
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2018-04-13
    Rocío A. Araya-López, Javier Lopatin, Fabian E. Fassnacht, H. Jaime Hernández

    In the Maipo watershed, situated in central Chile, mining activities are impacting high altitude Andean wetlands through the consumption and exploitation of water and land. As wetlands are vulnerable and particularly susceptible to changes of water supply, alterations and modifications in the hydrological regime have direct effects on their ecophysiological condition and vegetation cover. The aim of this study was to evaluate the potential of Worldview-2 and Sentinel-2 sensors to identify and map Andean wetlands through the use of the one-class classifier Bias support vector machines (BSVM), and then to estimate soil moisture content of the identified wetlands during snow-free summer using partial least square regression. The results obtained in this research showed that the combination of remote sensing data and a small sample of ground reference measurements enables to map Andean high altitude wetlands with high accuracies. BSVM was capable to classify the meadow areas with an overall accuracy of over ∼78% for both sensors. Our results also indicate that it is feasible to map surface soil moisture with optical remote sensing data and simple regression approaches in the examined environment. Surface soil moisture estimates reached r2 values of up to 0.58, and normalized mean square errors of 19% using Sentinel-2 data, while Worldview-2 estimates resulted in non-satisfying results. The presented approach is particularly valuable for monitoring high-mountain wetland areas with limited accessibility such as in the Andes.

  • Skipping the real world: Classification of PolSAR images without explicit feature extraction
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-12-13
    Ronny Hänsch, Olaf Hellwich

    The typical processing chain for pixel-wise classification from PolSAR images starts with an optional preprocessing step (e.g. speckle reduction), continues with extracting features projecting the complex-valued data into the real domain (e.g. by polarimetric decompositions) which are then used as input for a machine-learning based classifier, and ends in an optional postprocessing (e.g. label smoothing). The extracted features are usually hand-crafted as well as preselected and represent (a somewhat arbitrary) projection from the complex to the real domain in order to fit the requirements of standard machine-learning approaches such as Support Vector Machines or Artificial Neural Networks. This paper proposes to adapt the internal node tests of Random Forests to work directly on the complex-valued PolSAR data, which makes any explicit feature extraction obsolete. This approach leads to a classification framework with a significantly decreased computation time and memory footprint since no image features have to be computed and stored beforehand. The experimental results on one fully-polarimetric and one dual-polarimetric dataset show that, despite the simpler approach, accuracy can be maintained (decreased by only less than 2% 2 % for the fully-polarimetric dataset) or even improved (increased by roughly 9% 9 % for the dual-polarimetric dataset).

  • Beyond RGB: Very high resolution urban remote sensing with multimodal deep networks
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-11-23
    Nicolas Audebert, Bertrand Le Saux, Sébastien Lefèvre

    In this work, we investigate various methods to deal with semantic labeling of very high resolution multi-modal remote sensing data. Especially, we study how deep fully convolutional networks can be adapted to deal with multi-modal and multi-scale remote sensing data for semantic labeling. Our contributions are threefold: (a) we present an efficient multi-scale approach to leverage both a large spatial context and the high resolution data, (b) we investigate early and late fusion of Lidar and multispectral data, (c) we validate our methods on two public datasets with state-of-the-art results. Our results indicate that late fusion make it possible to recover errors steaming from ambiguous data, while early fusion allows for better joint-feature learning but at the cost of higher sensitivity to missing data.

  • Near real-time shadow detection and removal in aerial motion imagery application
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-11-22
    Guilherme F. Silva, Grace B. Carneiro, Ricardo Doth, Leonardo A. Amaral, Dario F.G. de Azevedo

    This work presents a method to automatically detect and remove shadows in urban aerial images and its application in an aerospace remote monitoring system requiring near real-time processing. Our detection method generates shadow masks and is accelerated by GPU programming. To obtain the shadow masks, we converted images from RGB to CIELCh model, calculated a modified Specthem ratio, and applied multilevel thresholding. Morphological operations were used to reduce shadow mask noise. The shadow masks are used in the process of removing shadows from the original images using the illumination ratio of the shadow/non-shadow regions. We obtained shadow detection accuracy of around 93% and shadow removal results comparable to the state-of-the-art while maintaining execution time under real-time constraints.

  • Object Scene Flow
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-11-22
    Moritz Menze, Christian Heipke, Andreas Geiger

    This work investigates the estimation of dense three-dimensional motion fields, commonly referred to as scene flow. While great progress has been made in recent years, large displacements and adverse imaging conditions as observed in natural outdoor environments are still very challenging for current approaches to reconstruction and motion estimation. In this paper, we propose a unified random field model which reasons jointly about 3D scene flow as well as the location, shape and motion of vehicles in the observed scene. We formulate the problem as the task of decomposing the scene into a small number of rigidly moving objects sharing the same motion parameters. Thus, our formulation effectively introduces long-range spatial dependencies which commonly employed local rigidity priors are lacking. Our inference algorithm then estimates the association of image segments and object hypotheses together with their three-dimensional shape and motion. We demonstrate the potential of the proposed approach by introducing a novel challenging scene flow benchmark which allows for a thorough comparison of the proposed scene flow approach with respect to various baseline models. In contrast to previous benchmarks, our evaluation is the first to provide stereo and optical flow ground truth for dynamic real-world urban scenes at large scale. Our experiments reveal that rigid motion segmentation can be utilized as an effective regularizer for the scene flow problem, improving upon existing two-frame scene flow methods. At the same time, our method yields plausible object segmentations without requiring an explicitly trained recognition model for a specific object class.

  • Landmark based localization in urban environment
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-09-28
    Xiaozhi Qu, Bahman Soheilian, Nicolas Paparoditis

    A landmark based localization with uncertainty analysis based on cameras and geo-referenced landmarks is presented in this paper. The system is developed to adapt different camera configurations for six degree-of-freedom pose estimation. Local bundle adjustment is applied for optimization and the geo-referenced landmarks are integrated to reduce the drift. In particular, the uncertainty analysis is taken into account. On the one hand, we estimate the uncertainties of poses to predict the precision of localization. On the other hand, uncertainty propagation is considered for matching, tracking and landmark registering. The proposed method is evaluated on both KITTI benchmark and the data acquired by a mobile mapping system. In our experiments, decimeter level accuracy can be reached.

  • A hybrid MLP-CNN classifier for very fine resolution remotely sensed image classification
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-08-02
    Ce Zhang, Xin Pan, Huapeng Li, Andy Gardiner, Isabel Sargent, Jonathon Hare, Peter M. Atkinson

    The contextual-based convolutional neural network (CNN) with deep architecture and pixel-based multilayer perceptron (MLP) with shallow structure are well-recognized neural network algorithms, representing the state-of-the-art deep learning method and the classical non-parametric machine learning approach, respectively. The two algorithms, which have very different behaviours, were integrated in a concise and effective way using a rule-based decision fusion approach for the classification of very fine spatial resolution (VFSR) remotely sensed imagery. The decision fusion rules, designed primarily based on the classification confidence of the CNN, reflect the generally complementary patterns of the individual classifiers. In consequence, the proposed ensemble classifier MLP-CNN harvests the complementary results acquired from the CNN based on deep spatial feature representation and from the MLP based on spectral discrimination. Meanwhile, limitations of the CNN due to the adoption of convolutional filters such as the uncertainty in object boundary partition and loss of useful fine spatial resolution detail were compensated. The effectiveness of the ensemble MLP-CNN classifier was tested in both urban and rural areas using aerial photography together with an additional satellite sensor dataset. The MLP-CNN classifier achieved promising performance, consistently outperforming the pixel-based MLP, spectral and textural-based MLP, and the contextual-based CNN in terms of classification accuracy. This research paves the way to effectively address the complicated problem of VFSR image classification.

  • Visual object tracking by correlation filters and online learning
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-07-29
    Xin Zhang, Gui-Song Xia, Qikai Lu, Weiming Shen, Liangpei Zhang

    Due to the complexity of background scenarios and the variation of target appearance, it is difficult to achieve high accuracy and fast speed for object tracking. Currently, correlation filters based trackers (CFTs) show promising performance in object tracking. The CFTs estimate the target’s position by correlation filters with different kinds of features. However, most of CFTs can hardly re-detect the target in the case of long-term tracking drifts. In this paper, a feature integration object tracker named correlation filters and online learning (CFOL) is proposed. CFOL estimates the target’s position and its corresponding correlation score using the same discriminative correlation filter with multi-features. To reduce tracking drifts, a new sampling and updating strategy for online learning is proposed. Experiments conducted on 51 image sequences demonstrate that the proposed algorithm is superior to the state-of-the-art approaches.

  • Learning a constrained conditional random field for enhanced segmentation of fallen trees in ALS point clouds
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-04-09
    Przemyslaw Polewski, Wei Yao, Marco Heurich, Peter Krzystek, Uwe Stilla

    In this study, we present a method for improving the quality of automatic single fallen tree stem segmentation in ALS data by applying a specialized constrained conditional random field (CRF). The entire processing pipeline is composed of two steps. First, short stem segments of equal length are detected and a subset of them is selected for further processing, while in the second step the chosen segments are merged to form entire trees. The first step is accomplished using the specialized CRF defined on the space of segment labelings, capable of finding segment candidates which are easier to merge subsequently. To achieve this, the CRF considers not only the features of every candidate individually, but incorporates pairwise spatial interactions between adjacent segments into the model. In particular, pairwise interactions include a collinearity/angular deviation probability which is learned from training data as well as the ratio of spatial overlap, whereas unary potentials encode a learned probabilistic model of the laser point distribution around each segment. Each of these components enters the CRF energy with its own balance factor. To process previously unseen data, we first calculate the subset of segments for merging on a grid of balance factors by minimizing the CRF energy. Then, we perform the merging and rank the balance configurations according to the quality of their resulting merged trees, obtained from a learned tree appearance model. The final result is derived from the top-ranked configuration. We tested our approach on 5 plots from the Bavarian Forest National Park using reference data acquired in a field inventory. Compared to our previous segment selection method without pairwise interactions, an increase in detection correctness and completeness of up to 7 and 9 percentage points, respectively, was observed.

  • Construction of pixel-level resolution DEMs from monocular images by shape and albedo from shading constrained with low-resolution DEM
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-03-27
    Bo Wu, Wai Chung Liu, Arne Grumpe, Christian Wöhler

    Lunar Digital Elevation Model (DEM) is important for lunar successful landing and exploration missions. Lunar DEMs are typically generated by photogrammetry or laser altimetry approaches. Photogrammetric methods require multiple stereo images of the region of interest and it may not be applicable in cases where stereo coverage is not available. In contrast, reflectance based shape reconstruction techniques, such as shape from shading (SfS) and shape and albedo from shading (SAfS), apply monocular images to generate DEMs with pixel-level resolution. We present a novel hierarchical SAfS method that refines a lower-resolution DEM to pixel-level resolution given a monocular image with known light source. We also estimate the corresponding pixel-wise albedo map in the process and based on that to regularize the shape reconstruction with pixel-level resolution based on the low-resolution DEM. In this study, a Lunar–Lambertian reflectance model is applied to estimate the albedo map. Experiments were carried out using monocular images from the Lunar Reconnaissance Orbiter Narrow Angle Camera (LRO NAC), with spatial resolution of 0.5–1.5 m per pixel, constrained by the Selenological and Engineering Explorer and LRO Elevation Model (SLDEM), with spatial resolution of 60 m. The results indicate that local details are well recovered by the proposed algorithm with plausible albedo estimation. The low-frequency topographic consistency depends on the quality of low-resolution DEM and the resolution difference between the image and the low-resolution DEM.

  • Disaster damage detection through synergistic use of deep learning and 3D point cloud features derived from very high resolution oblique aerial images, and multiple-kernel-learning
    ISPRS J. Photogramm. Remote Sens. (IF 5.994) Pub Date : 2017-03-09
    Anand Vetrivel, Markus Gerke, Norman Kerle, Francesco Nex, George Vosselman

    Oblique aerial images offer views of both building roofs and façades, and thus have been recognized as a potential source to detect severe building damages caused by destructive disaster events such as earthquakes. Therefore, they represent an important source of information for first responders or other stakeholders involved in the post-disaster response process. Several automated methods based on supervised learning have already been demonstrated for damage detection using oblique airborne images. However, they often do not generalize well when data from new unseen sites need to be processed, hampering their practical use. Reasons for this limitation include image and scene characteristics, though the most prominent one relates to the image features being used for training the classifier. Recently features based on deep learning approaches, such as convolutional neural networks (CNNs), have been shown to be more effective than conventional hand-crafted features, and have become the state-of-the-art in many domains, including remote sensing. Moreover, often oblique images are captured with high block overlap, facilitating the generation of dense 3D point clouds – an ideal source to derive geometric characteristics. We hypothesized that the use of CNN features, either independently or in combination with 3D point cloud features, would yield improved performance in damage detection. To this end we used CNN and 3D features, both independently and in combination, using images from manned and unmanned aerial platforms over several geographic locations that vary significantly in terms of image and scene characteristics. A multiple-kernel-learning framework, an effective way for integrating features from different modalities, was used for combining the two sets of features for classification. The results are encouraging: while CNN features produced an average classification accuracy of about 91%, the integration of 3D point cloud features led to an additional improvement of about 3% (i.e. an average classification accuracy of 94%). The significance of 3D point cloud features becomes more evident in the model transferability scenario (i.e., training and testing samples from different sites that vary slightly in the aforementioned characteristics), where the integration of CNN and 3D point cloud features significantly improved the model transferability accuracy up to a maximum of 7% compared with the accuracy achieved by CNN features alone. Overall, an average accuracy of 85% was achieved for the model transferability scenario across all experiments. Our main conclusion is that such an approach qualifies for practical use.

Some contents have been Reproduced with permission of the American Chemical Society.
Some contents have been Reproduced by permission of The Royal Society of Chemistry.
化学 • 材料 期刊列表