Qingyu Ren


Presentation: 10:45-12:00, Kennedy Union Ballroom



Download Project (753 KB)


Maintaining efficient energy usage in buildings or infrastructure is vital to the sustainable development program. The infrared (IR) thermography technique is widely used in thermoelastic behavior analysis and contactless temperature detection. However, the current thermal information is restricted in 2D space (i.e., image), depth and geometric information of the construction is missing. Geometric information and thermodynamics in the 3D space of building or infrastructure are essential for building energy auditing and infrastructure defect inspection. Though LiDAR is a possible solution, its costs, portability, and difficulty in merging additional data may bring more challenges to users. The aim of this project is to generate 3D RGB-IR point cloud models through monitoring over time, then the model can be used for simulation and energy consumption analysis and further applied in emergency response or city thermal behavior analysis. In recent years with advancements in computer vision, and the improvement of computation, structure from motion became an active method in 3D optical imaging. On the other hand, thermal cameras have low resolution compared with RGB cameras, we propose to map thermal information onto 3D reconstructed model. Efficiently and effectively merging 3D point clouds constructed from RGB images collected by ordinary cameras and corresponding infrared images from a co-calibrated thermal imager. This effort is cutting-edge and crucial as low-cost, accurate, and portable 3D thermal reconstruction solutions have significant potential in building energy auditing and infrastructure inspection/maintenance.

Publication Date


Project Designation

Graduate Research

Primary Advisor

Hui Wang

Primary Advisor's Department

Civil and Environmental Engineering and Engineering Mechanics


Stander Symposium, School of Engineering

Institutional Learning Goals

Scholarship; Practical Wisdom

Thermal 3D Point Cloud Generation and Model Reconstruction for Energy Auditing