Object Recognition and Grasping for Baxter Dual-Arm Robot Based on YOLOv8_OBB

Robots play a crucial role in the field of intelligent manufacturing, offering flexibility, safety, and efficiency, and are widely used in industrial production. Collaborative robots, such as the Baxter dual-arm robot, have a significant advantage in the international market. However, there is still room for improvement in resource utilization and cost-effectiveness for tasks like object recognition and grasping. To address the issues of poor expressiveness, instability, low precision in traditional teaching methods, and the high cost of depth-camera-based approaches, this paper proposes an object grasping method that integrates YOLOv8_OBB with infrared technology. This method utilizes the infrared laser on the Baxter robot's wrist for distance measurement, integrates the YOLOv8_OBB model for precise object recognition and pose estimation, and employs inverse kinematics to complete the robotic arm's accurate grasping task. The approach lays a technical foundation for further research in dual-arm collaborative assembly tasks. The research results demonstrate that the proposed method offers good stability and practicality, providing a potential direction for the development of robotic vision and grasping technologies.