The manufacturing industry has increasingly adopted robotic technologies and automation systems in recent years to enhance production efficiency. In large-scale manufacturing, most of processes are mechanized, while tasks requiring human judgment and flexible motion still depend heavily on manual skills, which are significantly influenced by individual differences, such as physical characteristics, body size and lengths of various body parts, all of which impact the work efficiency. For instance, maintaining inappropriate postures or performing motions misaligned with the worker’s physique can lead to accumulated fatigue and physical strain, which in turn decrease productivity and increase the risk of occupational accidents. In this paper, we propose a physique-aware skilled motion decision system utilizing depth cameras and skeletal estimation. The proposed system captures skilled worker body using a depth camera and extracts the three-dimensional coordinates of their joints through skeletal estimation. Experimental results demonstrate that the proposed system can estimate worker current posture and make motion decisions based on joint positions, angles and trajectories considering individual physique. By optimizing motion paths to target positions, the system effectively supports safe and efficient work execution based on worker motion data.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Physique-Aware Skilled Motion Decision System Based on Depth Camera and Skeletal Estimation

  • Kyohei Wakabayashi,
  • Tetsuya Oda,
  • Hideyuki Shimada,
  • Leonard Barolli

摘要

The manufacturing industry has increasingly adopted robotic technologies and automation systems in recent years to enhance production efficiency. In large-scale manufacturing, most of processes are mechanized, while tasks requiring human judgment and flexible motion still depend heavily on manual skills, which are significantly influenced by individual differences, such as physical characteristics, body size and lengths of various body parts, all of which impact the work efficiency. For instance, maintaining inappropriate postures or performing motions misaligned with the worker’s physique can lead to accumulated fatigue and physical strain, which in turn decrease productivity and increase the risk of occupational accidents. In this paper, we propose a physique-aware skilled motion decision system utilizing depth cameras and skeletal estimation. The proposed system captures skilled worker body using a depth camera and extracts the three-dimensional coordinates of their joints through skeletal estimation. Experimental results demonstrate that the proposed system can estimate worker current posture and make motion decisions based on joint positions, angles and trajectories considering individual physique. By optimizing motion paths to target positions, the system effectively supports safe and efficient work execution based on worker motion data.