Polygonal meshing for 3D stereo video sensor data

Many visually-guided robotic systems rely on stereo video data streams to obtain surface models of environ-mental structure. However stereo video-based 3D point-clouds are noisier than those produced from laser-based scanners and are subject to areas of sparse point informa-tion corresponding to textureless or specular surfaces. This complicates the process of constructing polygonal meshes from these point-clouds. This paper develops an approach to meshing for stereo video surface reconstruction that ad-dresses these issues and, by exploiting the known egomotion of the sensor, obtains surface normal and texture render-ing information. This widens the range of applications for which stereo video reconstruction can be applied and opens the possibility of using stereo video to generate models for real-time rendering applications.