NSF Award
1448572
The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase I project is in addressing one of the most common problems associated with ordering clothing and shoes online, which is the inability to know how things will fit. The technology being developed is a virtual footwear fitting system (VFFS) utilizing a user's own foot geometry. The virtual fitting experience will give size recommendations based on comfort levels and support levels for different shoe models. This will reduce the 35% return rate of online footwear retailers and thereby reduce their estimated $1.5 billion in annual losses in the US alone. Reducing the risk factor associated with online ordering will help people benefit fully from the convenience and large selection of online retailers. The reduction in return rate will reduce the carbon footprint of some 30 million annual returns. The VFFS will be available to the masses and will create a large exposure for 3D modeling fields such as computer aided design, finite element analysis and computer vision. 3D scanning with smartphones has endless applications, and the virtual fitting technology will help make computer vision principles more adapted for those applications. Virtual fitting opens the doors for mass customization, which is the future of footwear and garments. The technology will help bring our society closer to a world where things are tailored for us, convenient and affordable.<br/><br/><br/><br/>This project focuses on the development of a virtual footwear fitting system to reduce the return rate of online retail. The system consists of a foot scanner, shoe scanner and fitting algorithm. Online shoppers will use their smartphones to scan their feet and obtain a 3D model. A photogrammetry algorithm is being developed for this specific application. It utilizes feature recognition of the foot along with smartphone acceleration outputs to construct a 3D model with 1 mm accuracy in different image capturing environments. An adaptive inner volume (AIV) scanner is being developed to scan the inner volume of shoes that is combined with an outer scan to give a 3D shoe model with 0.5 mm accuracy along with material stiffness. The fitting algorithm will compare foot and shoe scans to perform a comprehensive fitting analysis giving online shoppers a way to know what shoe size they need and how comfortable their potential purchase will be. By the end of this project, a basic virtual fitting platform will be developed. It will allow a user to scan his or her foot with an iPhone and obtain a size recommendation for a specific shoe within ±0.5 size of his or her personal choice based on comfort levels, support levels and fit preferences.
