Research on Matching and Fusion Methods for Infrared Temperature Data and 3D Point Cloud Data of Substation Equipment

With smart grid development, intelligent substation operation and maintenance (O&M) demands advanced equipment monitoring. Though infrared thermography effectively captures temperatures, its 2D nature limits spatial context and complicates fault localization. This paper studies 2D infrared data and 3D point cloud matching/fusion for substation equipment. We map 2D pixels to 3D coordinates via camera calibration and pose estimation, then propose a modified two-stage registration algorithm: a dynamic asymmetric four-point set (DA-4PCS) for efficient coarse registration and a region-corresponding iterative closest point (RC-ICP) for accurate fine registration. A 3D temperature field is constructed using fused temporal data, enabling dynamic digital twin integration. Experiments show high-precision fusion (200ms mapping), transitioning from 2D to 3D perception and supporting intelligent O&M.