当前位置: X-MOL 学术IEEE Trans. Geosci. Remote Sens. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Water Body Detection in High-Resolution SAR Images With Cascaded Fully-Convolutional Network and Variable Focal Loss
IEEE Transactions on Geoscience and Remote Sensing ( IF 8.2 ) Pub Date : 2020-06-16 , DOI: 10.1109/tgrs.2020.2999405
Jinsong Zhang , Mengdao Xing , Guang-Cai Sun , Jianlai Chen , Mengya Li , Yihua Hu , Zheng Bao

The water body detection in high-resolution synthetic aperture radar (SAR) images is a challenging task due to the changing interference caused by multiple imaging conditions and complex land backgrounds. Inspired by the excellent adaptability of deep neural networks (DNNs) and the structured modeling capabilities of probabilistic graphical models, the cascaded fully-convolutional network (CFCN) is proposed to improve the performance of water body detection in high-resolution SAR images. First, for the resolution loss caused by convolutions with large stride in traditional convolutional neural network (CNN), the fully-convolutional upsampling pyramid networks (UPNs) are proposed to suppress this loss and realize pixel-wise water body detection. Then considering blurred water boundary, the fully-convolutional conditional random fields (FC-CRFs) are introduced to UPNs, which reduce computational complexity and lead to the automatic learning of Gaussian kernels in CRFs and the higher boundary accuracy. Furthermore, to eliminate the inefficient training caused by imbalanced categorical distribution in the training data set, a novel variable focal loss (VFL) function is proposed, which replaces the constant weighting factor of focal loss with the frequency-dependent factor. The proposed methods can not only improve the pixel accuracy and boundary accuracy but also perform well in detection robustness and speed. Results of GaoFen-3 SAR images are presented to validate the proposed approaches.

中文翻译:

级联全卷积网络和可变焦点损失的高分辨率SAR图像中的水体检测

由于多种成像条件和复杂土地背景引起的干扰变化,高分辨率合成孔径雷达(SAR)图像中的水体检测是一项艰巨的任务。受到深层神经网络(DNN)出色的适应性和概率图形模型的结构化建模能力的启发,提出了级联全卷积网络(CFCN),以提高高分辨率SAR图像中水体检测的性能。首先,针对传统卷积神经网络(CNN)中大步长卷积引起的分辨率损失,提出了全卷积上采样金字塔网络(UPNs)来抑制这种损失并实现像素级水体检测。然后考虑水边界模糊,将全卷积条件随机场(FC-CRF)引入到UPN中,这降低了计算复杂性,并导致CRF中高斯核的自动学习和更高的边界精度。此外,为了消除训练数据集中分类分布不均衡引起的训练效率低下,提出了一种新颖的可变焦点损失(VFL)函数,该函数将焦点损失的恒定加权因子替换为频率相关因子。所提出的方法不仅可以提高像素精度和边界精度,而且在检测鲁棒性和速度方面表现良好。提出了高分3 SAR图像的结果以验证所提出的方法。这降低了计算复杂度,并导致CRF中高斯核的自动学习和更高的边界精度。此外,为了消除训练数据集中分类分布不均衡引起的训练效率低下,提出了一种新颖的可变焦点损失(VFL)函数,该函数将焦点损失的恒定加权因子替换为频率相关因子。所提出的方法不仅可以提高像素精度和边界精度,而且在检测鲁棒性和速度方面表现良好。提出了高分3 SAR图像的结果以验证所提出的方法。这降低了计算复杂度,并导致CRF中高斯核的自动学习和更高的边界精度。此外,为了消除训练数据集中分类分布不均衡引起的训练效率低下,提出了一种新颖的可变焦点损失(VFL)函数,该函数将焦点损失的恒定加权因子替换为频率相关因子。所提出的方法不仅可以提高像素精度和边界精度,而且在检测鲁棒性和速度方面表现良好。提出了高分3 SAR图像的结果以验证所提出的方法。用频率相关因子代替恒定的焦点损失加权因子。所提出的方法不仅可以提高像素精度和边界精度,而且在检测鲁棒性和速度方面表现良好。提出了高分3 SAR图像的结果以验证所提出的方法。用频率相关因子代替恒定的焦点损失加权因子。所提出的方法不仅可以提高像素精度和边界精度,而且在检测鲁棒性和速度方面表现良好。提出了高分3 SAR图像的结果以验证所提出的方法。
更新日期:2020-06-16
down
wechat
bug