A Practical Framework for Automatic 3D Reconstruction of Clothing with RGB-D Cameras

摘要

Accurate 3D modelling of real clothing is of great significance for many applications. However, most existing methods of acquiring these 3D models are too expensive and non-automatic. In this paper, we propose a practical framework for the automatic 3D reconstruction of clothing by scanning real clothes using a motorized rotation stage and three RGB-D cameras (Intel RealSense Depth Camera D415). The framework consists of three parts, including data acquisition module, point cloud process module and model reconstruction module. Users can easily use it to reconstruct clothes because it works automatically and quickly. Experimental results show that the proposed system and method can obtain 3D models of clothing with satisfied 3D mesh and texture. This research provides manufacturers a fast and economical way to create 3D clothing models for applications like virtual reality display and online shopping etc.