Real-Time Target Detection and Tracking System Based on Stereo Camera for Quadruped Robots

As quadruped robots are suitable to be used in unstructured environment, environment perception for them becomes a popular field of research nowadays. This paper presents a novel and effective detection and tracking system based on stereo camera for quadruped robots, which allows a particular user to designate himself as the robot's leader. Firstly, the object detection is carried out by YOLO-V3 method based on the newfangled residual network. Then, the object tracking is performed based on ECO method, which is an improved kernel algorithm. Finally, a IDLA and CCF cascaded pedestrian re-identification method is adopted to identify the navigator among the crowd. In this work, we tackle the key causes behind the problems of computational complexity and over-fitting, with the aim of simultaneously improving both speed and performance. The comprehensive experimental results validate the effectiveness and robustness of our algorithm. In addition, our system runs at a real-time of 30 frames per second.