Human Behavior Recognition Method Based on Edge Intelligence

Abstract

The increasingly intelligent video surveillance system is the certain result of the gradual maturity of information technology. Human behavior recognition is one of the important tasks in the area of intelligent security monitoring. This paper proposes a human behavior recognition mechanism that uses edge-cloud collaborative computing. Firstly, at the edge node $N_0$ , the video is preprocessed to remove similar frames and the extracted skeleton sequence is expressed in multiple levels. Then the cloud trains the spatial-temporal graph ConvNet model and deploys it to the edge nodes $N_1\sim N_m$ . The edge uses the trained model to complete behavior recognition tasks and uploads the results to the cloud for fusion to obtain the final behavior category. The experimental results prove that the advantages of edge-cloud collaboration have made the model recognition accuracy rate steadily increase by more than 2.2%.