China Engineering Robotics Competition and International Open Championship, 2021

In the vision robot recognition competition (undergraduate), our group successfully implemented robot autonomous line patrol, color recognition, and autonomous shooting functions using advanced robot vision and tracking algorithms. By optimizing the robot’s gait control through visual feedback and careful debugging, we achieved a more stable and faster racing performance. These efforts culminated in our winning the first prize in China.

Bipedal robots have greatly benefited from the advancements in modern artificial intelligence, finding widespread practical applications and becoming a significant area of research. This project focuses on the design, debugging, and optimization of vision-based robot competition projects. Specifically, it addresses the mechanical structures and software systems of bipedal robots to achieve humanoid postures and accomplish tasks such as intelligent identification in competitions.

Through the use of robot vision and tracing algorithms, our robot was able to autonomously patrol, recognize colors, and shoot targets. Based on the visual feedback, we fine-tuned the robot’s gait control to enhance its overall coordination. We further optimized the algorithms for each function, considering the impact of external factors during the actual competition environment.

After rigorous testing and verification, we improved the robot’s anti-interference capability, resulting in a more stable performance, especially for the highly dynamic and maneuverable robot. This success demonstrates the potential of bipedal robots and showcases their growing significance in various aspects of social life and research.

The video introduces our group and the debugging process. Download the video here.

The video showcases the complete flow of a bipedal robot competition. Download the video here.