Full-Perception Robotic Surgery Environment with Anti-Occlusion Global-Local Joint Positioning

Area of Science:

• Robotics
• Surgical Technology
• Computer Vision

Background:

• Robotic surgery involves complex human-robot cooperation with dynamic movement.
• Mutual occlusion between individuals, robots, and devices is a significant challenge.
• Traditional methods lack scene integrity and have restricted workspaces.

Purpose of the Study:

• To propose a fully perception robotic surgery environment.
• To develop a global-local joint positioning framework.
• To enhance positioning accuracy in dynamic occlusion scenarios.

Main Methods:

• A global-local joint positioning framework was developed.
• An improved Kalman filter method was utilized for enhanced positioning.
• A view margin model was employed to evaluate accuracy in dynamic occlusion.

Main Results:

• The proposed method demonstrated superior positioning results compared to classical filtering techniques.
• The system effectively addresses mutual occlusion issues in robotic surgery.
• Improved accuracy was achieved in dynamic and occluded environments.

Conclusions:

• The fully perception robotic surgery environment and joint positioning framework offer a robust solution.
• The improved Kalman filter and view margin model contribute to enhanced accuracy and evaluation.
• This work advances the safety and efficiency of human-robot collaboration in surgery.