Photogrammetric meshes and 3D points cloud reconstruction: A genetic algorithm optimization procedure

Virtual reconstruction of heritage is one of the most interesting and innovative tool for preservation and keeping of historical, architectural and artistic memory of many sites that are in danger of disappearing. Find the best way to present an object in virtual reality is necessary for reasons linked to technology itself. In particular, the rendering of heavy object, in terms of details and meshes, influences the presentation of the whole virtual scene. Different researches have shown the onset of problems such as sickness due to an incorrect construction and representation of virtual scenes. In this paper we propose a 3D points cloud reconstruction method based on an optimization procedure by using genetic algorithm to improve the mesh obtained by low cost acquisition devices. The improved photogrammetric technique could be used to build virtual scenario by inexpensive devices (i.e. smartphone), without the cost and computational complexity of expensive devices.