Parametric approach for prior knowledge-base 3D reconstruction

Within the framework of Heritage documenting, digital reconstruction has stimulated, since years, the development and implementation of many techniques and software. Indeed, laser scanning and advances in imaging and photography are democratizing access to data and improving the ability to share and interpret Heritage information over the world. 3D acquisition knew a significant progress thanks to the introduction of laser sensors, but is also having problems related to exploiting the acquired data. We propose a methodological approach of 3D reconstruction, based on prior knowledge whose computer interpretation serves to structure the acquired data. It is about parametric interpretations describing the object and the morphological relations of its parts, allowing the generation of consistent and reusable models. This reconstitution process we want generalizable is explicated through an example; the finality not being the modeling in itself but the description of its different phases.