全景影像匹配与定位技术
2017-2018，武汉立得公司合作项目
In this project, we study a challenging case to find correspondents from wide baseline panoramic images with large geometric distortions from sphere projection and significant errors from multi-camera rig geometry. First, we deduce the camera model and epipolar model of a multi-camera rig system. Second, epipolar errors are analysed to determine the searching area for pixelwise matching. A low-cost laser scanner is alternatively used to constrain the depth of an object. Third, several classic feature descriptors are introduced to template matching and evaluated on the multi-view panoramic images. we propose a template matching method combining a fast version of SIFT descriptor that experimentally achieves the best performance considering both accuracy and efficiency, by beating other feature descriptors and the most recent template matching methods.
