Abstract:Low-order features based on convolution kernel are easy to be distorted when encountering dramatic view angle transformation and atmospheric scattering in remote sensing (RS) images. To address this concern, this article first proposes to operate semantic segmentation of RS images based on the high-order information, which can represent the relative relationship of low-order features and is robust and stable when suffering feature distortion. Besides, semantic decouples have recently been well researched and have achieved significant improvement in image understanding. Thus, in this article, a high-order semantic decoupling network (HSDN) is proposed to disentangle features by semantics based on high-order features. Specifically, HSDN first represents each pixel by calculating the pixel-level affinity as a high-order feature and then clusters these pixels into different semantics. Afterward, an attention-like mask generation module is designed for both intra-semantic and inter-semantic groups, leading to three kinds of masks, including the semantic decoupling mask (SDM), which utilizes each high-order cluster centroid as a mask to compact features intracluster and expand different interclusters, so as to improve semantic disentangle performance to a better extent; semantic enhancement mask (SEM), which records pixel-level relative correlation within a class to sufficiently exploit high-order features and could enhance feature robustness; and boundary supplementary mask (BSM), which aims to process borderline pixels to reduce cluster errors. Finally, by applying masks on pixels both within classes and on borderlines, semantic decoupled features are generated and concatenated to realize segmentation. The quantitative and qualitative experiments are conducted on two large-scale fine-resolution RS image datasets to demonstrate the significant performance of adopting high-order representation. Besides, we also implement numerous experiments to va- idate the effectiveness of the proposed semantic decouple framework in dealing with complicated and distortion-prone RS image segmentation tasks.

Boundary-Aware Dual-Stream Network for VHR Remote Sensing Images Semantic Segmentation

Boundary-enhanced dual-stream network for semantic segmentation of high-resolution remote sensing images

DSHNet: A Semantic Segmentation Model of Remote Sensing Images Based on Dual Stream Hybrid Network

Remote sensing image semantic segmentation via class-guided structural interaction and boundary perception

Hierarchical Self-Attention Embedded Neural Network With Dense Connection for Remote-Sensing Image Semantic Segmentation

DRBANET: A Lightweight Dual-Resolution Network for Semantic Segmentation with Boundary Auxiliary

A Boundary Guided Cross Fusion Approach for Remote Sensing Image Segmentation

BES-Net: Boundary Enhancing Semantic Context Network for High-Resolution Image Semantic Segmentation

SAM-Assisted Remote Sensing Imagery Semantic Segmentation with Object and Boundary Constraints

DHRNet: A Dual-Branch Hybrid Reinforcement Network for Semantic Segmentation of Remote Sensing Images

An Object-Aware Network Embedding Deep Superpixel for Semantic Segmentation of Remote Sensing Images

HRCNet: High-Resolution Context Extraction Network for Semantic Segmentation of Remote Sensing Images

Multi-modal remote sensing image segmentation based on attention-driven dual-branch encoding framework

Multi-Attention-Based Semantic Segmentation Network for Land Cover Remote Sensing Images

Remote Sensing Semantic Segmentation via Boundary Supervision-Aided Multiscale Channelwise Cross Attention Network

Feature-Selection High-Resolution Network With Hypersphere Embedding for Semantic Segmentation of VHR Remote Sensing Images

Object-Enhanced Semantic Segmentation Model for High-Resolution Remote Sensing Images

Multiscale Sea-Land Segmentation Networks for Weak Boundaries

Boundary-Guided Lightweight Semantic Segmentation With Multi-Scale Semantic Context

High-Order Semantic Decoupling Network for Remote Sensing Image Semantic Segmentation