A Hybrid RNN-CNN Approach with TPI for High-Precision DEM Reconstruction

Digital elevation model (DEM), as the fundamental units of terrain morphology, are crucial for understanding surface processes and land use planning. However, automated classification faces challenges due to inefficient terrain feature extraction from raw LiDAR point clouds and the limitations of traditional methods in capturing fine-scale topographic variations. To address this, we propose a novel hybrid RNN-CNN framework that integrates multi-scale TPI features to enhance both DEM generation. Our approach first models voxelated LiDAR point clouds as spatially ordered sequences, using recurrent neural networks (RNNs) to encode vertical elevation dependencies and convolutional neural networks (CNNs) to extract planar spatial features. By incorporating TPI as a semantic constraint, the model learns to distinguish terrain structures at multiple scales. Residual connections refine feature representations to preserve micro-topographic details during DEM reconstruction. Extensive experiments in the complex terrains of Jiuzhaigou demonstrate that our lightweight hybrid framework not only achieves excellent DEM reconstruction accuracy in both flat areas and complex terrains, but also improves computational efficiency by more than 24.9% on average compared to traditional interpolation methods, making it highly suitable for resource-constrained applications.