Image Processing in Vision of Chinese Chess Playing Robot Based on SVM

Abstract

Vision system of Chinese chess playing robot is the key part. This work is based on color image processing. Color transformations,thresholding,morphological skeleton and the Hough transform were applied to detect and segment the chesses and form the image. The rotational invariant descriptions were extracted in the recognition step and the SVM(support vector machine) was used to recognize the characters on the chesses. This method is proved to be precise,accurate and has high recognition rates by experiments.