Karst rocky desertification on the Qinghai–Tibet Plateau poses a severe threat to the region’s fragile ecosystem. Accordingly, the rapid and accurate delineation of plateau karst landforms is essential for monitoring ecological degradation and guiding restoration strategies. However, automatic recognition of these landforms in remote sensing imagery is hindered by challenges such as blurred boundaries, fragmented targets, and poor intra-region consistency. To address these issues, we propose the Region Affinity Propagation Network (RAP-Net). This framework enhances intra-region consistency, edge sensitivity, and multi-scale context fusion through its core modules: Region Affinity Propagation (RAP), High-Frequency Multi-Scale Attention (HFMSA), and Global–Local Cross Attention (GLCA). In addition, we constructed the Plateau Karst Landform Dataset (PKLD), a high-resolution remote sensing dataset specifically tailored for this task, which provides a standardized benchmark for future studies. On the PKLD, RAP-Net surpasses eight state-of-the-art methods, achieving 3.69–10.31% higher IoU and 3.88–14.28% higher Recall, thereby demonstrating significant improvements in boundary delineation and structural completeness. Moreover, in a cross-regional generalization test on the Mount Genyen area, RAP-Net—trained solely on PKLD without fine-tuning—achieved 2.38% and 1.94% higher IoU and F1-scores, respectively, than the Swin Transformer, confirming its robustness and generalizability in complex, unseen environments.

Landslides are characterized by their suddenness and destructive power, making rapid and accurate identification crucial for emergency rescue and disaster assessment in affected areas. To address the challenges of limited landslide samples and data complexity, a landslide identification sample library was constructed using high-resolution remote sensing imagery combined with field validation. An innovative Dual-Coded Segmentation Network (DS Net), which realizes dynamic alignment and deep fusion of local details and global context, image features and domain knowledge through the multi-attention mechanism of Prior Knowledge Integration (PKI) module and Cross-Feature Aggregation (CFA) module, significantly improves the landslide detection accuracy and reliability. To objectively evaluate the performance of the DS Net model, four efficient semantic segmentation models—SegFormer, SegNeXt, FeedFormer, and U-MixFormer—were selected for comparison. The results demonstrate that DS Net achieves superior performance (overall accuracy = 0.926, precision = 0.884, recall = 0.879, and F1-score = 0.882), with metrics that are 3.5–7.1% higher than the other models. These findings confirm that DS Net effectively improves the accuracy and efficiency of landslide identification, providing a critical scientific basis for landslide prevention and mitigation.

Accurate semantic segmentation of complex landforms in remote sensing imagery is hindered by pronounced intra-class heterogeneity, blurred boundaries, and irregular geomorphic structures. To overcome these challenges, this study presents LENet (Landforms Expert Segmentation Net), a novel segmentation network that combines axial semantic modeling with deformation-aware compensation. LENet follows an encoder–decoder framework, where the decoder integrates three key modules: the Expert Enhancement Block (EEBlock) for capturing long-range dependencies along axial directions; the Feature Expert Compensator (FEC) employing deformable convolutions with channel–spatial decoupled weights to emphasize ambiguous intra-class regions; and the Cross-Sparse Attention (CSA) mechanism that suppresses background noise via multi-rate sparsity masks and enhances intra-class consistency through cosine-similarity weighting. Experiments conducted on the PKLD plateau karst and GVLM landslide datasets demonstrate that LENet achieves IoU scores of 70.39% and 80.95% and Recall values of 83.33% and 91.38%, surpassing eight state-of-the-art methods. These results confirm that LENet effectively balances global contextual understanding and local detail refinement, providing a robust and accurate solution for complex landform segmentation in remote sensing imagery.