| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 263674 | Energy and Buildings | 2012 | 8 Pages |
This paper describes a new approach to multi-sensor registration of infrared images and 3D-laser scanner models, based on the extraction of common features in the IR image and the range image obtained from a laser-scanner 3D-point cloud. The workflow developed in this paper allows the automatic registration of two different sensors with completely different characteristics, including fields of view, spatial resolution and spatial position. As a result of the workflow, thermographic and geometric data are combined in a thermographic 3D-model and its projection as an orthothermogram, where thermal defects can be visually detected and precisely located, and heat losses can be quantified given the availability of surface measurements. This approach is applied to different types of constructions and building materials, showing its efficiency and appropriateness for building inspection and energy studies.
► Automatic thermographic texture of 3D point clouds. ► Orthothermogram to perform thermal and geometric measurements. ► Detection and precise location of thermal defects in 3D model. ► Energy evaluation through isotherms and surface measurements. ► Quantification of heat losses in orthothermogram.
