Power operation site safety protection equipment wearing state detection method based on YOLO-MobileNet

The invention belongs to the technical field of target detection, and particularly discloses a power operation site safety protection equipment wearing state detection method based on YOLO-MobileNet, and the method comprises the steps: constructing a data set, and carrying out the marking of a data set image; a safety protection equipment wearing state detection network based on the YOLO-MobileNet is constructed; and training and testing the network model. According to the method, a bottleneck module forming a MobileNetv2 is improved, an improved MobileNetv2 feature extraction network is constructed by using the improved bottleneck module and an original bottleneck module, a backbone feature extraction network CSPDarkNet53 of YOLOv4 is replaced by the feature extraction network, and meanwhile, a part of a standard convolutional network of the YOLOv4 is replaced by a deep separable convolutional network, so that the feature extraction efficiency of the MobileNetv2 is improved. The detection precision of the sa