Deep-sea net-coat aquaculture is an important development direction of marine aquaculture at present, so the net-coat breakage detection technology is of great significance. This paper proposes an image detection technology based on machine vision, firstly using an improved pyramid fusion algorithm to preprocess the collected images; after that, combined with the Faster R-CNN method of deep learning, a large number of net-suit image data is put into the convolutional neural network for training, and finally the trained model is applied to the actual underwater net-suit detection test, and the experiment shows that the proposed The experiments show that the detection method proposed in this paper has significant improvement in both accuracy and speed.