Visual servo control of the hexapod robot with obstacle avoidance

Abstract

This paper is to design a visual servo control for a hexapod robot with obstacle avoidance. The implementation of the motion control for the hexapod robot using the inverse kinematics and visual recognition system is used to achieve the trajectory tracking with obstacle avoidance. The control structure is composed of three parts: a tilt camera based on the concepts of mechanical geometry, visual servo control systems, and motion dynamics for trajectory tracking and obstacle avoidance. First, the depth between the obstacle and robot based on the proportion of the size of the area is constructed. Then, the image processing is used to identify whether there are any obstacles in the front, and make it as a feedback to the servo control system. For image recognition, we use OpenCV to process environment to the grayscale and binarization, filter noise through erosion and dilation, and then fill all of the contours using Sobel edge detection, and finally calculate the area and compare that with each other. Finally, some experiments for a hexapod robot with obstacle avoidance are used to validate the performance of the proposed control scheme.