Feature Pyramid U-Net with Attention for Semantic Segmentation of Forward-Looking Sonar Images

Abstract

Forward-looking sonar is a technique widely used for underwater detection. However, most sonar images have underwater noise and low resolution due to their acoustic properties. In recent years, the semantic segmentation model U-Net has shown excellent segmentation performance, and it has great potential in forward-looking sonar image segmentation. However, forward-looking sonar images are affected by noise, which prevents the existing U-Net model from segmenting small objects effectively. Therefore, this study presents a forward-looking sonar semantic segmentation model called Feature Pyramid U-Net with Attention (FPUA). This model uses residual blocks to improve the training depth of the network. To improve the segmentation accuracy of the network for small objects, a feature pyramid module combined with an attention structure is introduced. This improves the model’s ability to learn deep semantic and shallow detail information. First, the proposed model is compared against other deep learning models and on two datasets, of which one was collected in a tank environment and the other was collected in a real marine environment. To further test the validity of the model, a real forward-looking sonar system was devised and employed in the lake trials. The results show that the proposed model performs better than the other models for small-object and few-sample classes and that it is competitive in semantic segmentation of forward-looking sonar images.

Fast Semantic Segmentation of Remote Sensing Images Using a Network That Integrates Global and Local Information

“…MANet [19] adopts a novel dual-attention mechanism and a new cross-dimensional interactive attention feature fusion module, which enhances feature extraction and fusion. Based on U-Net, FPUA [20] uses a feature pyramid module combined with an attention structure to improve the segmentation accuracy of the network for small objects. To enhance the regional integrity of images and thus reduce misclassification, RE-Net [21] introduces the regional information into the base network.…”

SACuP: Sonar Image Augmentation with Cut and Paste Based DataBank for Semantic Segmentation

“…One way to improve segmentation is morphological reconstruction combined with the level set method (MRLSM) [28], which improves upon the level set method [29] to perform sonar image segmentation. Deep learning-based methods include feature pyramid U-Net with attention (FPUA) [30], which features an improved semantic segmentation network architecture. Recently, following the introduction of the segment anything model (SAM) [31], there have been attempts to apply SAM to sonar [32].…”