Study and Application of Service Robot Positioning Technology Based on Visual System

This paper comes up with the dual Otsu threshold method based on four-color block and K-L transform, which can be used to segment the image and extract the objective as well as identify the objective by the color characteristics. This method realizes real-time adjustment of the image segmentation threshold and improves the accuracy, timeliness and robustness of the visual system. Finally, with the service robot as the platform, it designs and preliminarily realizes the visual system of the service robot and achieves good effect in the practical application.