Wearable Detection Application of Protective Equipment for Live Working Based on YOLOv9

In order to solve the problems of relying on manual supervision and low degree of automation in the wearing of safety protective equipment for non-outage workers in the power sector, this paper proposes an automatic detection model based on improved YOLOv9. Firstly, a self-calibrating convolution is designed, which can explicitly expand the field of view of each convolution layer, thus enriching the output features and enhancing the feature extraction ability of the model for targets in complex environments. Secondly, the model down-sampling module is improved. In order to reduce the loss of information caused by the down-sampling stage, a down-sampling module based on wavelet transform is designed to maximize the integrity of information. Finally, through experiments on the self-built live working safety wearable data set, the average accuracy mean is used as the model evaluation index, and compared with the existing target detection algorithm. The experimental results show that the im-proved algorithm proposed in this paper has a significant improvement in detection accuracy and real-time performance.