Research on humanoid robot vision system based on fuzzy image segmentation algorithm with color table

Zhong Qiu-bo Zhao Jie and Tong

Abstract

According to the characteristic of humanoid robot soccer, the vision system embedded in robot is required to quickly and precisely recognize the target object, which facilitates a proper decision made by robot. Thereby, a useful approach which bases on HSI color space, uses the improved ambiguous color classes for color segmentation and applies the pyramid matching method to extract image eigenvalues is proposed in this paper. The end of this paper, a simulation experiment is done to prove validity of the proposed algorithm.