Pieces Identification in the Chess System of Dual-Robot Coordination Based on Vision

With the research and development of chess robot and machine vision, chess robot with visual function has recently received an increasing interest in the community. This paper introduces the chess system of dual-robot coordination based on vision and the process of visual system structure, gives the coordinates translation from computer and image coordinates to actual coordinates. And in the process of pieces identification, it firstly gives the character segmentation method and training process based on BP neural network, and then identifies the pieces. All the work contributes to the chess system of dual-robot coordination based on vision.