Abstract:Sonar is widely used in marine water cleaning tasks, so sonar images have become an effective tool for garbage detection and underwater scene analysis. However, it is an extremely difficult task to achieve fully supervised denoising and garbage detection for sonar images. This is because sonar images are weakly annotated samples that are susceptible to noise interference and have no reference clean image. To this end, we propose a sonar image garbage instance segmentation model via global despeckling and dynamic attention graph optimization(GD-DAGO). Specifically, a self-supervised blind spot network denoising structure is presented in this paper. The proposed denoising model overcomes the defects of information loss in the traditional blind spot network structure and performs global awareness speckle suppression for the noise characteristics of sonar images themselves. In addition, a novel dynamic attention structure is employed to improve the target region estimation in the instance segmentation module and does not require supervision beyond image-level category labeling. Finally, in order to enhance the cooperative ability between the two tasks, we adopt a local perceptual loss strategy based on mask proposals guided by the downstream task, so that the whole model takes more into account the characteristics of sonar images and better serves the sonar garbage detection task. Experimental results on ARACATI 2017 and marine-debris-fls-datasets (MDFD) show that the proposed algorithm achieves a performance gain of 0.4218 and 4.2% in terms of denoising effect (ENL) and detection accuracy ( AP25 ), respectively, compared with suboptimal algorithms.

A fast graph based method for object segmentation in sidescan sonar image

Image Stitching and Target Perception for Autonomous Underwater Vehicle-collected Side-Scan Sonar Images

A Robust and Fast Method for Sidescan Sonar Image Segmentation Based on Region Growing

Multi-Scale Fish Segmentation Refinement and Missing Shape Recovery

Sonar Image Coarse-to-Fine Few-Shot Segmentation Based on Object-Shadow Feature Pair Localization and Level Set Method

On-Line Multi-Class Segmentation of Side-Scan Sonar Imagery Using an Autonomous Underwater Vehicle

A Fast Algorithm for SAR Image Segmentation Based on Key Pixels

Underwater sonar image segmentation combining pixel-level and region-level information

A fast region-based image segmentation based on least square method

Fast Graph-Based SAR Image Segmentation Via Simple Superpixels

An Integrated Strategy toward the Extraction of Contour and Region of Sonar Images

When SAM Meets Sonar Images

Underwater Sonar Image Segmentation by a Novel Joint Level Set Model

Novel Method For Sar Image Segmentation With Application To Bridge Detection

Sidescan Sonar Image Target Extraction Method Based on Variable Initial Signed Distance Function-Based Active Contour CV Model

YOLOv5-Based Seabed Sediment Recognition Method for Side-Scan Sonar Imagery

An Approach for Integrating SAR Imagery in Sea-Land Segmentation and Coastline Detection

Efficient Instance Segmentation Method For High-Resolution SAR Imagery

Segmentation of sonar imagery using convolutional neural networks and Markov random field

Sonar Image Garbage Detection via Global Despeckling and Dynamic Attention Graph Optimization

Underwater Image Segmentation Based on Fast Level Set Method.