Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : May 27, 2020 - May 30, 2020
3D Measurements for metallic objects or semi-transparent objects are required; however, they are still a hard problem in the robot vision field. We have developed a novel 3D measurement method that utilizes LTM sparse estimation and epipolar geometry constraint, which can measure these complex lighting objects accurately. In this presentation, we will show our results from the perspectives of computation cost and robustness. Na¨ıve LTM estimation takes a very long time; thus, we have developed computation cost and the number of projection and capturing reduction methods. We also developed a robust LTM sparse estimation method for non-linear projector-camera system behavior.
