Research on Human Motion Recognition Based on Computer Vision Technology

In order to improve the accuracy of human motion recognition, computer vision technology is selected as the research tool, and a representation method of human bone and joint is proposed. This method takes the adaptive human motion bone center architecture as the action recognition tool, constructs the human motion action recognition model by calculating the angular acceleration and angular velocity of the neck and hip joints, and uses the model to calculate the action data and generate the recognition results. The experimental results show that the motion recognition accuracy of this algorithm is high, and it can be used as a motion recognition tool.