Enabling 3D online shopping with affordable depth scanned models

Abstract

Online shopping systems nowadays are constructed with the capability of displaying text and 2D images to the customers. Although it is enough for the purpose of exchanging product information between sellers and buyers, the plain and dull shopping experience is not really appreciated. With the recent advancement of 3D technologies in the web environment, we can enhance such an experience with 3D visualized products. However, the creation of 3D content is a major challenge in realizing a 3D enabled shopping system. The high cost of 3D scanning devices inhibits the popularity of scanned 3D objects on the web and related applications. Therefore, one major targets of our system is to provide a low cost 3D scanning solution suitable for naïve web users. We exploit the Kinect device, which is a low-cost and fast depth sensor with color, in the development of a 3D scanning system. With an integration of a marker-based tracking system to estimate the current view angle of Kinect sensor during depth acquisition, the obtained depth data can reconstruct the 3D model by a preliminary coordinate frame alignment and a registration of all point clouds. The whole process of capturing, storing, uploading, compressing and displaying 3D models of products on web application is done with minimal user involvements. The introduction of attractive and fashionable 3D product on the web will gain significant attention from customers and evolve habits and traditions in e-commerce, especially for C2C business, in which sellers can promote their goods for sale with scanned 3D models.