HUMAN POSE RECOGNITION BASED ON OPENPOSE AND APPLICATION IN SAFETY DETECTION OF INTELLIGENT FACTORY

In order to improve the safety environment of the intelligent factory, accurately identify the location of various key points of the human body, interpret the body pose, and avoid dangerous actions affecting the safety of the intelligent factory equipment and personal safety. This paper designs a intelligent factory safety detection system. The system uses the OpenPose algorithm to detect the position of the human skeleton in each video frame, and extracts features such as normalized coordinates, human motion speed, and joint motion speed. Then the system uses the deep neural network algorithm DNN for classification. The experimental results show that the use of OpenPose for human pose recognition can effectively improve the recognition accuracy.