Research on Underwater Navigation Algorithm Based on SIFT Matching Algorithm

Abstract

Autonomous Underwater Vehicle (AUV) is currently used to side scan sonar, doppler velocimetry and inertial navigation devices and other equipment-based integrated navigation program navigation. The side scan sonar can generate the underwater acoustical image in real time and match it with the feature database through the image matching algorithm to filter the navigation data. Aiming at the characteristics of underwater acoustic image data, the method of drawing underwater acoustic image is described. In order to integrate the feature matching of the underwater acoustic image into the integrated navigation, this paper introduces the SIFT feature matching algorithm to improve the characteristics of the underwater acoustic image and introduces the discriminant condition based on the least squares method. After the matching criterion is introduced, the improved SIFT feature matching algorithm is greatly improved when the landmark matching is achieved, and the occurrence of false matching is avoided. Finally, the real experimental data obtained by the side-scan sonar equipment are verified by the algorithm, and the satisfactory results are obtained.