Verification of unique cloth handling performance based on 3D recognition accuracy of cloth by dual-eyes cameras with photo-model-based matching

Abstract

Nowadays, innovative robotic technology has been implemented in the garment companies. However, robots have been confronted with difficulties in recognising and handling deformable object such as cloth, string etc., especially if the object is unique. Specifically, the cloth placed in front of the robot is rightly the intended one to be handled and to pick and place (handle) at a designated position automatically are two main problems during 3D cloth recognition and handling performance by a robot. In this paper, model generation method from cloth photograph and model-based matching method (recognition method) utilising Genetic Algorithm (GA) are presented. The proposed system uses dual-eyes cameras to recognise the target cloth and estimate the pose of that cloth for handling. The proposed system is used to verify the 3D handling under predetermining position and orientation range. 100 times handling experiment has been executed, having shown the effectiveness of proposed photo-model-based cloth recognition system.