UAV Remote Sensing Image Mosaic Technology Combined with Improved SPHP Algorithm

The paper proposes an improved algorithm, for the problem that SPHP algorithm is not suitable for the UAV Remote sensing image mosaic algorithm because it is susceptible to geomorphic factors, resulting in deformation or ghosting. The algorithm firstly extracts feature points by combining SURF algorithm and Harris algorithm. The obtained feature points are coarsely matched by KNN algorithm, then matched by PROSAC algorithm. Finally, the weight coefficient is introduced to calculate the spatial transformation model of the image overlap region. This model replaces the original spatial model of SPHP algorithm. The ghosting of the image overlap area is reduced, and the stitched image is caused to have a smaller deformation. The final result shows that this improved SPHP algorithm can effectively remove the ghost image of the stitched image and generate better stitching results.