Novel data registration techniques for art diagnostics and 3D heritage visualization

In Cultural Heritage applications, as often in the medical field, diagnoses have to be accomplished by comparing data obtained with different techniques and sensors. However, the acquired data are often misaligned and in order to exactly superimpose them, they need to be correctly registered among each other. In this contribution we present an automatic featurebased registration technique to align image data. Moreover, since 2D images acquired with various devices are generally of easier interpretation when mapped onto a 3D model or produce nicer and photo-realistic visualization results, we propose an automated method to map 2D images onto 3D surfaces. Results are presented and discussed.