An Adapted Visual Servo Algorithm for Substation Equipment Inspection Robot

Based on the real-time image capturing and recognition, the robot is responsible to inspect the working status of various equipments in the smart substation. In this paper, an adapted efficient robotics visual servo algorithm is proposed in order to improve the accuracy of capturing the target images. When the robot captures image of the equipment, SIFT method is utilized to match with the template image to verify the captured image containing the region of interesting (ROI) or not. Once the ROI is part or entirety out of the real-time image, the PTZ is controlled to make up for the deficiency by calculating the offset pixels. Then the robot will capture more accurate target image. Farther more, a "high zoom ratio" visual servo solution is also figured out by this proposed algorithm. The experiment results proved that this algorithm improved the efficiency of image capture to the robot. And it is effective to save the computer resource to achieve long battery life when the robot is working the inspection missions. The application of this proposed algorithm will provide facilities for robot to realize 24 hour all-weather inspection in the substation.