Multisource remote sensing image registration based on geometric constraints

Abstract

With the continuous development of remote sensing technology, the types of remote sensing image data are more diversified. More spatial information of images can be obtained by multisource fusion. In addition, the complementarity between sensors can effectively overcome the limitations of a single sensor in complex environments. The registration of optical image and SAR image is the key point in multisource image registration. Optical images have high resolution. But are vulnerable to the impact of harsh environments, resulting in the loss of spectral details. SAR images have strong penetrability to vegetation, cloud, ice and snow. But they are interfered by speckle noise. The imaging mechanism of optical image and SAR image is different, and the difference of gray information is large, which may lead to the performance failure of two kinds of image registration. To solve the above problems, in this paper, we propose a method of optical image and SAR image registration based on geometric constraints, which optimizes the feature descriptor locally through the spatial geometric structure characteristics between similar feature points. The experimental results show that the proposed method improves the matching performance compared with several state-of-the-art methods in terms of the matching accuracy.