A new image segmentation method based on HSI color space for biped soccer robot
Honge Ren, Qiubo Zhong · 2008
Many researchers now are engaging themselves in the research on vision system for humanoid soccer robot. By making a deep analysis on the HSI color space, a better algorithm is presented which is judging the main feature of current pixel dynamically according to an intensity (I value) related formula in this paper. Fuzzy K-means clustering method is adopted in the image segmentation after all the similar pixels are found. This research is a good foundation for the future work of more complex vision based motion planning of humanoid soccer robot.