Side-scan sonar image target identification method

Abstract

The invention relates to a side-scan sonar image target recognition method, and the method comprises the steps: collecting continuous side-scan sonar seabed images, and obtaining a plurality of original sonar images which comprise a to-be-recognized target; performing gray scale binarization processing on the original sonar image; performing modeling processing on noise in the sonar image after gray scale binarization processing, and enhancing the sonar image in a multi-scale geometric transformation domain; carrying out edge detection on the enhanced sonar image by utilizing a Canny operator;filtering a non-target edge in the sonar image after edge detection according to a side-scan directional image threshold filtering method, and reserving a detection target edge; and identifying the target after edge detection. According to the technical scheme provided by the invention, the edge detection capability and the identification efficiency of the side-scan sonar image under a complex noise background are greatly improved through image enhancement and denoising in a multi-scale geometric change domain and in combination with an edge detection technology of side-scan directivity.

Description

technical field [0001] The invention relates to the technical field of target recognition, in particular to a method for target recognition of side scan sonar images. Background technique [0002] Ocean acoustic target detection technology is crucial for safeguarding national sovereignty, safeguarding the safety of the national marine environment, and promoting ocean exploration and development. As the most ideal tool for submarine mapping, side-scan sonar has a high working frequency and relatively strong resolution, which can clearly reflect the topography of the seabed and the wreckage of ships and planes on the seabed. Since the 1960s, the side-scan sonar system has It has been widely used in the fields of seabed target detection, seabed mapping and seabed exploration, and marine resource development. The existing research on side-scan sonar image processing mainly includes image enhancement, geometric correction, noise reduction and restoration and other preprocessing,...

AI Technical Summary

Problems solved by technology

[0003] When the front side scan sonar detects the target, the background interference includes reverberation, ocean environment noise and aircraft noise, etc., which cause the echo signal of the target reflection characteristics to be distorted, and the reflection on the image is obvious speckle noise phenomenon, with prominent noise, The edges are blurred and the texture is weak. At the same time, due to the influence of imaging methods, system parameter settings and other factors, side scan sonar imaging has obvious far and near brightness changes and geometric distortion.

Images