InfraDiffusion: zero-shot depth map restoration with diffusion models and prompted segmentation from sparse infrastructure point clouds
This addresses automated inspection challenges for masonry assets like bridges and tunnels in low-light environments, though it is incremental as it adapts existing diffusion models to a specific domain.
The paper tackles the problem of fine-grained brick-level segmentation in masonry infrastructure using sparse point clouds by proposing InfraDiffusion, a zero-shot framework that restores depth maps with diffusion models, resulting in significant improvements in segmentation accuracy without task-specific training.
Point clouds are widely used for infrastructure monitoring by providing geometric information, where segmentation is required for downstream tasks such as defect detection. Existing research has automated semantic segmentation of structural components, while brick-level segmentation (identifying defects such as spalling and mortar loss) has been primarily conducted from RGB images. However, acquiring high-resolution images is impractical in low-light environments like masonry tunnels. Point clouds, though robust to dim lighting, are typically unstructured, sparse, and noisy, limiting fine-grained segmentation. We present InfraDiffusion, a zero-shot framework that projects masonry point clouds into depth maps using virtual cameras and restores them by adapting the Denoising Diffusion Null-space Model (DDNM). Without task-specific training, InfraDiffusion enhances visual clarity and geometric consistency of depth maps. Experiments on masonry bridge and tunnel point cloud datasets show significant improvements in brick-level segmentation using the Segment Anything Model (SAM), underscoring its potential for automated inspection of masonry assets. Our code and data is available at https://github.com/Jingyixiong/InfraDiffusion-official-implement.