Analysis and Identification of Pipeline's Work Efficiency Based on OpenPose

Currently, the majority of the industrial assembly lines of domestic and even global enterprises still need manual operation. Especially in some production process which is more complex, manual operation is essential. In such a large-scale manual operation, the efficiency of it is very important. Compared with the method of judging the working efficiency of workers by stopwatch, it is more intuitive and analyzable to analyze the motion curve of workers through the video collected by the camera on the industrial assembly line. This paper uses the video data collected by industrial cameras and constructs a computer vision system that can automatically judge the work efficiency of workers by human posture recognition algorithm based on OpenPose.