A New Line Matching Approach for High-Resolution Line Array Remote Sensing Images

In this paper, a new line matching approach for high-resolution array remote sensing images is presented. This establishes the correspondence of straight lines on two by combining multiple constraints. Firstly, three geometric constraints, epipolar, direction and point-line relationship, are used in turn to reduce number candidates. After this, similarity double descriptor distance, determine optimal matches. Finally, co-linearity constraint check one-to-many many-to-one correspondences results. The proposed tested eight representative image patches selected from ZY-3 satellite images, results compared with those state-of-the-art approaches. Experiments demonstrate superiority potential due its higher accuracy greater matches most cases.