Design of an Automatic 3D Reconstruction Scanning System in Virtual Reality

Three-dimension reconstruction technology helps designers and developers convert virtual objects from the physical world. However, there are still methodological limitations in collecting scanned objects’ images, such as time-consuming work with lower accuracy by manual scan and the high cost of existing machine scanning. Therefore, this study proposes a new stereoscopic surround scanning system covering the target objects in quantity from multiple angles. Camera equipment (i.e., a mobile phone) with controlled poses is added above the system, and images from multi-directional perspectives are recorded. The 3D reconstruction algorithm only needs numerous images to reconstruct objects from 2D to 3D. Based on the 3D reconstruction algorithm, this research provided two comparative experiments and several methods to evaluate the benefit of the scanning machine. Besides, previous automatic scanners mainly cost hundreds to thousands of dollars, while the new scan station in this study was designed with less than 100${\$}$ hardware budget. Accordingly, applying the whole system can improve the scanning method and reduce labour costs. This design of the scanning system brings new insights to promote the development of reconstruction technology in virtual reality applications.