Bill Wei is a senior conservation scientist at the Netherlands Institute for Cultural Heritage, in Amsterdam. He shares his views on the FINGaRtPRINT concept, a EU funded project geared towards prevention of art theft through the use of a dual fingerprinting technology based on recording surface and colour characteristics.

What drove the research into art fingerprinting?
There were two major driving forces. The first was a desire to find a method to uniquely identify valuable cultural heritage objects without requiring direct contact with the object. All previous or modern methods such as stickers or radio-frequency identification (RFID) chips must be attached to the object. This does not meet museum conservators’ requirements.

The second was a need for a more flexible way of studying the surface two and three dimensional objects.  This would help determine the effects of conservation treatments on their condition and appearance.  Such issue matters when making decisions about how to preserve valuable objects of cultural heritage from the impact of time.

What were the major challenges to overcome?
The main challenges were technical. First, we had to resolve the issue of mounting the fingerprinting measurement system that measured the art piece’s roughness on a robot to allow for fingerprinting of virtually all types of two and three dimensional objects. Then, we had to write software, which could be simply understood by all museum and collection workers as they may not have a technical background. The third challenge was to ensure that the time required to create a fingerprint was kept to a minimum.

Between the two available fingerprinting solutions, which include a robot system and a table model, which is the best for fingerprinting art pieces?
The different technologies available serve different purposes. The table model is used to apply the technology to larger areas.  This can be achieved by automatically measuring several neighbouring areas and stitching the results together. The robot can do this too, but the prototype has some mechanical stability problems that have yet to be solved as the quality of the fingerprint is disturbed by vibrations.  However, a larger fingerprint is normally not necessary.

By contrast, the robot model has been calibrated to scan areas of 1.6 x 1.6 mm because the area should be large enough to be easy to locate by the art piece owner, customs or law enforcement officials. In addition, it should encompass enough details and therefore have a high-enough resolution to be unique. We found that for that purpose, using an objective lens of 10 times magnification was sufficient.

What remains to be achieved before the system can be exploited in museums?
The system needs to be optimised so that it can operate even faster than currently. Another priority is to improve its transportability. What is more, the stability of the robot arm against external vibrations needs to be improved, so that it can measure larger areas, if desired.

Photo credit Adrienne Norman