Fish catching using gazing-GA visual servoing - verification of robustness for lighting condition varieties

摘要

The purpose of this research is to propose a scene recognition method for real-time visual servoing of a manipulator, which has robustness for lighting environments. We have verified the robustness of the proposed method against lighting condition variations and the effectiveness of the method for real-time recognition using a fish catching experiment by visual servoing of a hand-eye robot.