Integration of aerial and ground images has been proved as an efficient approach to enhance the surface reconstruction in urban environments. However, as the first step, the feature point matching between aerial and ground images is remarkably difficult, due to the large differences in viewpoint and illumination conditions. Previous studies based on geometry-aware image rectification have alleviated this problem, but the performance and convenience of this strategy are still limited by several flaws, e.g. quadratic image pairs, segregated extraction of descriptors and occlusions. To address these problems, we propose a novel approach: leveraging photogrammetric mesh models for aerial-ground image matching. The methods have linear time complexity with regard to the number of images. It explicitly handles low overlap using multi-view images. The proposed methods can be directly injected into off-the-shelf structure-from-motion (SFM) and multi-view stereo (MVS) solutions. First, aerial and ground images are reconstructed separately and initially co-registered through weak georeferencing data. Second, aerial models are rendered to the initial ground views, in which color, depth and normal images are obtained. Then, feature matching between synthesized and ground images are conducted through descriptor searching and geometry-constrained outlier removal. Finally, oriented 3D patches are formulated using the synthesized depth and normal images and the correspondences are propagated to the aerial views through patch-based matching. Experimental evaluations using five datasets reveal satisfactory performance of the proposed methods in aerial-ground image matching, which succeeds in all of the ten challenging pairs compared to only three for the second best. In addition, incorporation of existing SFM and MVS solutions enables more complete reconstruction results, with better internal stability.

航空影像和地面影像的集成被证明是增强城市环境中地面重建的有效方法。然而，作为第一步，由于视点和照明条件的巨大差异，航空图像和地面图像之间的特征点匹配非常困难。先前基于几何感知图像校正的研究已经缓解了该问题，但是该策略的性能和便利性仍然受到一些缺陷的限制，例如二次图像对，描述符和遮挡的分离提取。为了解决这些问题，我们提出了一种新颖的方法：利用摄影测量网格模型进行空地图像匹配。该方法关于图像数量具有线性时间复杂度。它使用多视图图像显式处理低重叠。所提出的方法可以直接注入现成的运动结构（SFM）和多视图立体声（MVS）解决方案中。首先，分别重建航空图像和地面图像，并通过弱地理参考数据进行初始配准。其次，将航空模型渲染到初始地面视图，在其中获取颜色，深度和法线图像。然后，通过描述符搜索和受几何约束的异常值去除，进行合成图像和地面图像之间的特征匹配。最后，使用合成的深度和法线图像来制定定向的3D补丁，并通过基于补丁的匹配将对应关系传播到鸟瞰图。使用五个数据集进行的实验评估显示了所提出方法在航空地面图像匹配中的令人满意的性能，在十对具有挑战性的对中都取得了成功，而第二对只有三对。此外，结合现有的SFM和MVS解决方案可实现更完整的重建结果，并具有更好的内部稳定性。在所有十对具有挑战性的对中都取得了成功，而第二名仅获得了三对。此外，结合现有的SFM和MVS解决方案可实现更完整的重建结果，并具有更好的内部稳定性。在所有十对具有挑战性的对中都取得了成功，而第二名仅获得了三对。此外，结合现有的SFM和MVS解决方案可实现更完整的重建结果，并具有更好的内部稳定性。