Image target detection method combining lightweight attention mechanism and YOLOv3 network

The invention discloses an image target detection method combining a lightweight attention mechanism and a YOLOv3 network. The method comprises a training process of a target detection algorithm of the lightweight attention mechanism and the YOLOv3 network, the algorithm combines the lightweight attention mechanism and the YOLOv3 network to improve the feature extraction capability, and a depth separable convolution module is combined into the YOLOv3 network to improve the feature extraction capability. The efficiency of the algorithm is improved, the detection precision is further improved, the multi-scale fusion method is used in the traditional YOLOv3 network, the feature extraction capability of the model is improved, the performance of the model is further improved. The lightweight attention mechanism, the depth separable convolution and the multi-scale fusion method are combined into the YOLOv3 network. The target detection method with the high recognition degree is designed, the task of target detectio