Adaptive Underwater Sonar Image Target Detection

The underwater sonar image detection is a very important key technique. In the field of image detection, the detection of underwater sonar image is special and difficult because of its blurry image and underwater interference. To solve the difficult detection problems of underwater sonar image, the median filter algorithm and the Gamma transformation algorithm are used to preprocess the image, and an adaptive threshold method is designed to binarize the images. In addition, the corrosion expansion algorithm is used to further eliminate the interference of the image. Finally, the k-means algorithm is used to cluster the pixels of the image to achieve the final detection classification. To reduce underwater interference, a novel multi-frame detection algorithm is proposed based on multiple consecutive images and the detection performance is improved.