Adaptive edge detection and Histogram color segmentation for centralized vision of soccer robot

Abstract

In this paper an adaptive algorithm to detect arena of a soccer robot, and the position, role and orientation of robots are proposed. Adaptive edge detection using Otsu's method is used in arena detection and adaptive color segmentation using Histogram is utilized in robot detection. The proposed method is shown to be able to detect arena and robots in dynamic environment. The experiment results show that the adaptive edge detection method is able to 100% detect the arena regardless the changes in camera displacement and lighting condition. Furthermore, the histogram color segmentation successfully recognizes every significant color that exists in the arena. In the real time fashion of soccer robot, those methods give an average of 2.295% and 0.656% of error in x-axis and y-axis, respectively.