Practical works on preservation of cultural heritage monuments require comprehensive information about current monument condition. Moreover, tracking changes in monument condition over time is of special significance. This task is usually fulfilled by monitoring condition of deteriorated cultural heritage objects and storing the monitoring data in the form of detailed atlases of the object. Such monitoring activities are expensive as they involve a team of qualified experts in chemistry, biology, physics, restoration. Each expert gathers information about monument condition on site, within his field of expertise.

The proposed technique allows monitoring monument condition by means of a regular digital photography of monuments followed by a computer-aided analysis of the colour images. The output electronic atlas is produced as a result of processing digital image series, using a special computer-aided system. This approach significantly reduces monitoring costs, facilitates and speeds up creation of electronic atlases of cultural heritage objects.

The system works as follows. The input data are series of digital colour images of a deteriorated monument. First, the images are processed in order to ensure the adjustment of shooting conditions and image unification in terms of colour. Then, special computer algorithms are applied to analyze colour parameters of the images and detect deterioration areas due to biodegradation and mechanical damage. In fact, the colour parameters of the detected deteriorated areas are compared to sample colour parameters of specific destructors stored in a reference database which is created beforehand as a result of assessment of sample images of decomposers by experts in appropriate fields.

Therefore, the colour parameters of images of deteriorated monument surfaces can indicate specific destructors. In case of biodeterioration, computerized analysis of colour images leads to  specific names and aggression parameters of microorganisms, moss, lichens, etc.   In case of mechanical destruction, the image colour parameters are indicative of shears and cracks.  Finally, the computer program creates an electronic atlas for the monument which comprises deteriorated and undeteriorated areas. This atlas is then stored in a larger database of monument atlases. Later, an updated atlas can be produced following further shooting of the object. Comparison of the new atlas with the previous one(s) allows an effective and reliable detection of new deterioration areas over time. 

The technique was tested by the authors in 2006-2008 in St.-Petersburg (Russia) and in Khersonesos (Ukraine). The pilot objects were the rapakivi granite head at Sergievka Estate (and ancient marble and limestone monuments. The results are promising, although further research is needed in order to enrich the databases with more data on other destructors.

The reported project follows up the research findings presented at the III ECAI Conference held in May 2007 in Moscow, Russia.