Damage Detection on Building Surfaces with Multi-spectral Techniques

Multi-spectral imaging methods give the possibility to detect damages, like moisture and natural cover (e.g. moss), on surfaces of building structures. An Infrared Vidicon system and two digital cameras have been tested concerning their applicability for multispectral analysis. The results of the evaluation with multi-spectral image classification are compared. Digital cameras simplify the acquisition of image data. The classification of images taken with digital cameras achieves promising results. Still there are difficulties with the influence of inhomogeneous illumination on the classification results. Therefore, a multi-spectral laser-scanning system was developed, which uses four semiconductor lasers at different wavelengths for the defined illumination of the object surface.