A Simulation System for Decision of Camera Angle View and Placement: A Comparison Study of Simulation and Experimental Results

In the manufacturing field, the soldering process of brazing electronic components into circuit boards and the milling process of using cutting tools to remove unnecessary parts from materials are part of the manual labor process. To ensure worker safety and improve product quality in the many tasks performed by humans, studies are promoting the collection and analysis of worker motion data. Non-contact sensing systems consider camera to capture images of the worker and detects the worker posture and dangerous movements by utilizing skeletal estimation and object recognition. However for the analysis, the worker body should be completely within the cameras angle of view. If the worker body is outside the camera angle of view or is only partially captured, the data necessary for analysis will still be lost. To capture the motion of the whole body of a worker, multiple cameras using a network of cameras are used in some cases, while in limited environments it is also important to acquire the motion data by a monocular camera. In this paper, we propose a simulation system for decision of camera angle of view and placement. We compare the simulation end experimental results. The evaluation results we conclude that the proposed system can effectively decide the camera placement and angle.