The Sonar Image Sequence Movement Target Detection Based on Surfacelet Transform at Complex Background

摘要

In order to realize movement target detection and tracking, the sonar image sequence movement target detection based on Surfacelet transform at complex background is proposed. Firstly, sonar image characteristic area extract based on Surfacelet transform, secondly, extract characteristic quantities and confirm correspondence between frames of characteristic quantities, at last, calculate motion parameters and put into motion model to estimate the entire image motion vectors, and complete the moving object detection. The experiments show that the method can complete to extract movement targets correctly.