GCN Embedding Swin-Unet for Forest Remote Sensing Image Semantic Segmentation

Forest resources are among the most important ecosystems on the earth. The semantic segmentation and accurate positioning of ground objects in forest remote sensing (RS) imagery are crucial to the emergency treatment of forest natural disasters, especially forest fires. Currently, most existing methods for image semantic segmentation are built upon convolutional neural network (CNN). Nevertheless, these techniques face difficulties in directly accessing global contextual information and accurately detecting geometric transformations within the image’s target regions. This limitation stems from the inherent locality of convolution operations, which are restricted to processing data structured in Euclidean space and confined to square-shaped regions.Inspired by the Graph Convolution Network (GCN) with robust capabilities in processing irregular and complex targets, as well as Swin Transformers renowned for exceptional global context modeling, we present an innovative semantic segmentation framework for forest remote sensing imagery termed GSwin-Unet. This framework embeds GCN model into Swin-Unet architecture, and for the first time apply the method of combining GCN and Transformer in the domain of forest RS imagery analysis. GSwin-Unet features an innovative parallel dual-encoder architecture of GCN and Swin transformer. First, we integrate the Zero-DCE (Zero-Reference Deep Curve Estimation) algorithm into GSwin-Unet to enhance forest RS image feature representation. Second, a feature aggregation module (FAM) is proposed to bridge the dual encoders by fusing GCN-derived local aggregated features with Swin transformer-extracted features. Our study demonstrates that the GSwin-Unet significantly improves performance on the Forest Remote Sensing Dataset and exhibits good adaptability on GID dataset.