Identification of potential areas for building retrofit using thermal digital imagery and CFD models

Monitoring actual energy performance of existing buildings and measuring performance deviations from benchmarked values can help identify the potential areas for building retrofit. However, current building auditing practices include application of energy simulation tools or sensor networks, which can be time-consuming and may only provide system or appliance-level retrofit suggestions. In this paper, a quick and cost-effective approach is presented to recognize potential areas for building upgrade using digital and thermal imagery in addition to Computational Fluid Dynamics (CFD) models. In the proposed method, a dense 3D point cloud model of an existing building is first generated using an image-based 3D reconstruction algorithm. Next, thermal images are fused by a novel thermal mapping algorithm. The expected thermal performance is then simulated by CFD analysis using the dimensions extracted from the reconstructed 3D building model. Finally, the actual and expected thermal performance models are automatically superimposed within a new Energy Performance Augmented Reality (EPAR) environment wherein performance deviations are visualized. The proposed method is validated on several rooms in an existing instructional facility. The experimental results reflect the promise that EPAR models can be used as a building retrofit decision-making tool to facilitate the retrofit process.