Technology Category
Computer Vision: SegmentationMachine Learning: Learning on the Edge & Model CompressionIntelligent Robots: Multimodal Perception & Sensor Fusion
TEFormer: Texture-Aware and Edge-Guided Transformer for Semantic Segmentation of Urban Remote Sensing Images
🤖 AI Summary
Urban remote sensing image semantic segmentation suffers from classification ambiguity due to subtle texture variations, ill-defined object boundaries, and high spatial structural similarity among land-cover classes. To address these challenges, this paper proposes a texture-aware and edge-guided collaborative Transformer architecture. We introduce a novel texture-aware module to enhance discriminability among visually similar categories, and design an edge-guided three-branch decoder integrated with a feature fusion module to jointly optimize fine-grained texture modeling, multi-scale contextual reasoning, and boundary structure preservation. Extensive experiments on the Potsdam, Vaihingen, and LoveDA benchmarks achieve mIoU scores of 88.57%, 81.46%, and 53.55%, respectively—surpassing state-of-the-art methods by significant margins. These results demonstrate the effectiveness and robustness of our approach in complex urban scenes.
Technology Category
Application Category
📝 Abstract
Semantic segmentation of urban remote sensing images (URSIs) is crucial for applications such as urban planning and environmental monitoring. However, geospatial objects often exhibit subtle texture differences and similar spatial structures, which can easily lead to semantic ambiguity and misclassification. Moreover, challenges such as irregular object shapes, blurred boundaries, and overlapping spatial distributions of semantic objects contribute to complex and diverse edge morphologies, further complicating accurate segmentation. To tackle these issues, we propose a texture-aware and edge-guided Transformer (TEFormer) that integrates texture awareness and edge-guidance mechanisms for semantic segmentation of URSIs. In the encoder, a texture-aware module (TaM) is designed to capture fine-grained texture differences between visually similar categories to enhance semantic discrimination. Then, an edge-guided tri-branch decoder (Eg3Head) is constructed to preserve local edges and details for multiscale context-awareness. Finally, an edge-guided feature fusion module (EgFFM) is to fuse contextual and detail information with edge information to realize refined semantic segmentation. Extensive experiments show that TEFormer achieves mIoU of 88.57%, 81.46%, and 53.55% on the Potsdam, Vaihingen, and LoveDA datasets, respectively, shows the effectiveness in URSI semantic segmentation.
Problem
Research questions and friction points this paper is trying to address.
Addresses semantic ambiguity in urban remote sensing image segmentation
Handles complex edge morphologies from irregular object boundaries
Resolves subtle texture differences between visually similar categories
Innovation
Methods, ideas, or system contributions that make the work stand out.
Texture-aware module captures fine-grained differences
Edge-guided tri-branch decoder preserves local details
Edge-guided feature fusion integrates contextual information
🔎 Similar Papers
