METHODOLOGY TO CREATE 3D MODELS FOR AUGMENTED REALITY APPLICATIONS USING SCANNED POINT CLOUDS

Abstract

Precise digital documentation of cultural heritage assets is essential for its preservation and protection. This documentation increases the efficiency of scientific studies that are being carried out during the restoration and renovation process. Precise digital documentation makes use of different laser scanners technologies. 3D scanning devices usually provide a large amount of point clouds, which require long post-processing times and large storage space. This paper presents a methodology to obtain simplified 3D models designed to remove redundant points and maintain only representative points, preserving the 3D model aspect and allowing the 3D models to be implemented in different augmented reality application on mobile devices. The 3D mesh optimization methods that have been analyzed and compared are dedicated 3D mesh optimization software (CATIA and Geomagic Studio), open source tools such (Meshlab) and a numerical computing environment (MATLAB). The methodology proposes a split step that can be applied to both assemblies and reconstructed objects. In this step the 3D scan is divided into components (for assemblies) or original parts/restored parts (for restored cultural heritage assets). The efficiency and robustness are demonstrated using different 3D scanned Dacian artifacts.