Obstacle extraction method in sonar image, computer device and storage medium

Abstract

The invention discloses an obstacle extraction method in a sonar image, a computer device and a storage medium, and the method comprises the steps of sequentially carrying out the OTSU algorithm processing, median filtering algorithm processing, binarization processing and eight-connected region detection of the sonar image, extracting a foreground in the sonar image, determining the foreground as an obstacle, and the like, wherein an OTSU threshold value used by the OTSU algorithm is determined by an image point location distance and a gain coefficient of the sonar image, and a communication threshold value used by the eight-communication region detection is determined by a maximum communication region area and a communication proportion parameter. According to the invention, multiplicative noise, additive noise and large isolated noise points in the sonar image can be suppressed, an image connected domain of a known degraded image in a conversion domain can be restrained, a sparse structure of the degraded image is repaired, and a complete obstacle target can be extracted from a foreground in a low-quality sonar image obtained under an underwater adverse condition. The invention is widely applied to the technical field of image processing.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a method for extracting obstacles in sonar images, a computer device and a storage medium. Background technique [0002] Underwater obstacle detection is required in the fields of ship unmanned driving, military detection, seabed survey, sunken ship salvage and underwater pipeline detection. Since in the marine environment, sound wave transmission is almost the only method for long-distance information transmission, so sonar images are used to detect obstacles, that is, through image analysis of sonar images, obstacles in sonar images are extracted Determine the location and size of the obstacle. [0003] Due to the complex underwater environment, there are factors such as acoustic attenuation, Doppler and reverberation, resulting in low resolution, low contrast, and poor quality of acoustic imaging. Due to noise and interference, underwater acoustic images cannot effec...

AI Technical Summary

Problems solved by technology

[0003] Due to the complex underwater environment, there are factors such as acoustic attenuation, Doppler and reverberation, resulting in low resolution, low contrast, and poor quality of acoustic imaging. Due to noise and interference, underwater acoustic images cannot effectively identify targets. The edge is incomplete, which brings great difficulties to image segmentation, target recognition and other processing

Images