Fusion of thermal images and point clouds for enhanced wall temperature uniformity analysis in building environments

Achieving uniform temperature distribution is essential for optimizing insulation and improving building climate control. In this study, we propose a multi-sensor digitalization approach that fuses indoor thermal images with 3D point cloud data to create a detailed, geometry-aligned visualization of wall temperatures. By mapping 2D thermal information onto an accurately calibrated 3D model, we capture fine-grained temperature variations across interior surfaces. We validate the feasibility of this workflow through a comprehensive case study that details data acquisition, calibration, and fusion procedures. Our classroom case study demonstrated that the thermal point cloud reliably maps wall temperatures across a range of approximately 15 to 35 ^∘C, with an average residual misalignment of 2.44 pixels horizontally and 3.36 pixels vertically. Furthermore, it effectively identifies localized areas of thermal non-uniformity-particularly near windows and doors-which can inform targeted insulation improvements and climate control strategies. These results offer clear guidance for practical insulation adjustments and climate control strategies and provide a robust foundation for developing digital twin models. Ultimately, our approach offers a holistic view of the building thermal environment, leading to enhanced occupant comfort and increased energy efficiency.