Underwater Target Detection by Side-Scan Sonar Based on Yolov7-Attention

Side-scan sonar plays a crucial role in underwater exploration, and autonomous detection of side-scan sonar images is vital for detecting unknown underwater environments. However, due to the complexity of the underwater environment, few highlighted areas of the target, blurred feature details, and difficulty in collecting data from side-scan sonar, achieving high-precision autonomous target recognition in side-scan sonar images is challenging. This article solves this problem by improving the You Only Look Once v7 (YOLOv7) model to achieve high-precision object detection in side-scan sonar images. Firstly, Since the side-scan sonar image has a large useless area, the paper introduces Swin-Transformer for dynamic attention and global modeling, which improves the model’s weight on the target region. Secondly, the Convolutional Block Attention Module (CBAM) is used to further improve feature representation and enhance the accuracy of the neural network model. Experiments on the side-scan sonar image dataset (SSS) show that the improved YOLOv7 model has 85.32% and 53.32% in its average accuracy (mAP0.5) and (mAP0.5:0.95), respectively. These results are 4.65% and 4.98% higher than the YOLOv7 model. The improved YOLOv7 algorithm proposed in this paper has great potential for object detection and recognition of side-scan sonar images.