Research on TOF camera-based guided binocular fusion technology
Siyuan Gao, Lichao Fang, Xintian Cui · 2021 IEEE 3rd International Conference on Frontiers Technology of Information and Computer (ICFTIC) · 2021
As environment perception becomes increasingly necessary in human-computer interaction tasks, researchers are applying it to specific tasks, such as depth-of-field measurements by depth cameras. In order to improve the accuracy of depth information in-depth maps obtained by current depth cameras, this paper uses the complementary properties of TOF technology and binocular stereo vision technology to fuse TOF and binocular technology, guiding binocular stereo matching by matching TOF parallax maps with binocular parallax map feature points, and discussing the process of guided matching. Among them, the better Census matching algorithm is selected for experiments in the stereo matching process. The binocular matching after TOF guidance by this method significantly improves the matching time and the quality of the matched parallax map.