主催: 一般社団法人 日本機械学会
会議名: ロボティクス・メカトロニクス 講演会2020
開催日: 2020/05/27 - 2020/05/30
The purpose of this study is to improve the three-dimensional measurement accuracy in active stereo method. We propose a method to modify calibration parameters by non-linear optimization based on the error on euclidean space. The cost function in the optimization uses geometric constraints calculated based on euclidean coordinates of the feature points on the target. The accuracy of the proposed method was verified by the experiment. As a result, that the measurement accuracy was improved only in parts of the measurement regions was confirmed . By investigating the relationship between the spatial distribution of feature points and the measurement area, that the distribution affected the measurement accuracy was suggested.