The aim of this work is to evaluate the applicability of Very High Spatial Resolution (VHSR) images to heritage conservation, by using multi-spectral images with different resolutions to extract information concerning the state of conservation of building roofs. To this end, a hybrid classification method was developed that incorporates uncertainty information in the automatic classification of VHSR images. The developed classification methodology includes the following steps: 1) a soft pixel-based classification; 2) computation of the classification uncertainty; and 3) buildings classification based on decision rules which include the results of the soft pixel-based classification and its uncertainty. The proposed methodology was tested on the classification of the historical downtown centre of Coimbra, in Portugal, and validated with the results obtained from an independent pathology survey performed by the University of Coimbra. The test was carried out with a multi-spectral aerial image and a QuickBird satellite image with spatial resolutions of respectively 0.5 m and 2.4 m.