GPR-HIDiff: A diffusion-based model for horizontal interference suppression in urban underground detection radar profiles

• GPR-HIDiff replaces UNet convolutions with ResBlocks and integrates agent and spatial attention modules. • A cosine θ schedule is adopted in the forward diffusion to ensure smooth noise variance transitions. • GPR-HIDiff outperforms conventional and deep learning-based methods on both simulated and real GPR datasets. • Ablation and mixed-sample experiments demonstrate its robustness and enhanced feature representation. Automated subsurface utility detection systems in construction rely heavily on the quality of ground-penetrating radar (GPR) profiles, which are often degraded by high-amplitude horizontal interference. Existing low-rank decomposition methods lack the intelligence and flexibility required for multi-site data processing and involve labor-intensive parameter tuning, impeding their integration into intelligent construction workflows. To address these challenges, this paper proposes a horizontal interference suppression algorithm based on a diffusion model, termed GPR-HIDiff. The proposed model replaces conventional sequential convolutional operators with ResBlocks throughout the encoder, intermediate layer, and decoder of the UNet architecture, enhancing training stability. Lightweight agent attention modules are embedded between ResBlocks at each level to improve global information modeling capability. A spatial attention mechanism is deployed between the encoder and decoder to achieve adaptive spatial feature optimization. Furthermore, the forward diffusion phase adopts a cos θ schedule-based strategy to ensure a smooth temporal variation of noise variance. A standardized dataset comprising real-world measured samples and finite difference time domain simulation samples of urban road models has also been constructed. The effectiveness of the hybrid dataset, the introduced modules, the robustness analysis, and the cos θ schedule is validated through training with single / mixed datasets, ablation studies, evaluation of metric variations before and after the introduction of different noise levels, and comparative experiments with constant, linear, and cos θ schedules. Experimental results demonstrate that GPR-HIDiff significantly outperforms both traditional methods and state-of-the-art deep learning models on both simulated and real-world test samples. It effectively suppresses horizontal artifacts, preserves target hyperbolic contours, and avoids excessive reduction of target scattering, showcasing its exceptional performance. This method provides a powerful algorithmic foundation for high-resolution GPR imaging and target detection.