Face Detection and Recognition Based on Deep Learning in the Monitoring Environment

In the construction of smart cities and public safety, the number of cameras has increased dramatically, the artificial video management cannot meet the demand of urban construction. Therefore, intelligent video monitoring technology has become a research hotspot. In the actual video monitoring,the monitoring environment is very complex, Face recognition in video is often difficulty. This paper research the face recognition problem of monitoring video in the city, propose a mothed of detection and recognition method based on deep learning. The detection part adopts the fast speeding YOLO2 algorithm, the recognition part adopts high accuracy ResNet algorithm. Using WIDERFACE face detection database as training data set, Use the CASIA_Webface database for validation experiments. The experimental data show that the mothod YOLO2 algorithm and ResNet algorithm can complete detection and recognition of human face in monitoring video. the result is better. To further verify the recognition effect, the data set collected by the actual camera was tested. the test results show that the real time and accuracy of the system can meet the needs of practical engineering application.