Study on the Side-scan Sonar Image Matching Navigation Based on SURF

In the paper, a method based on the speeded up robust features (SURF) algorithm is presented for the seabed relief image matching navigation. By using of high-resolution features of the side-scan sonar image, the effective algorithm of image processing, SURF, is introduced to side-scan sonar seabed relief image processing for underwater navigation application. According to the characteristics of underwater environment and side-scan image, some improving measures including edge-preserving filtering and narrow strip processing are studied and presented in this paper. The method is used for underwater navigation experiments. The results show that the method can determine the location of the underwater vehicles with the matching accuracy of 1.14 m.