Research of Robot Color Logo Orientation Based on Hough Transform

This paper discusses the soccer robot orientation method which is based on sight, and locates soccer robot by using method based on color shape as well as establishes threshold segmentation, edge detection and the model of Hough transform for circle detection with an emphasis in threshold segmentation and Hough transform. Several methods are used for threshold segmentation, evident difference of images processed by different methods exists, this paper proposes Otsu self-adaptive threshold algorithm, which performs well in precision and anti-interference, as main method for threshold segmentation. Considering of many defects of Hough transform used for circular object detection, this paper proposes Hough transform based on gradient, which gets desirable result.