Applied Research on Target Detection and Tracking of a Multi-beam Sonar

The paper describes the working principle of the multi-beam sonar with electronic pitching beams vertically, develops offshore oil platform security, as well as the method of drawing sonar image. In order to detect and track the moving target, the multi-period image is preprocessed using gray level transformation, custom kernels convolution, OST threshold segmentation and binary image morphology. Combining with the characteristics of detecting target, this paper gets the lines in multi-period image and constructs the feature set. Then it conducts clustering and confidence evaluation, and gets the target trajectory. The results of sonar range trial on lake show that the methods proposed is effective to detect and track targets from multi-beam sonar image.