Design the best-FIT footwear using image processing

Abstract

In footwear, the variance in the sizes and shapes of feet of people poses a big problem for manufacturers to design shoes with enough sizes that match most of the potential customers. After deciding on the style, most people have to try on at least a few pair of shoes to find the one that “fits” the most. However, fitness of a shoe is highly dependent on many foot parameters which not fully captured by just shoe sizes. With the advancement of advanced techniques such as 3d scanning, we now can assess feet shape and size accurately [1-3]. However, most of the approaches are expensive, cumbersome and bulky. In this paper, we proposed a simple device equipped with two mirrors placed at a 60-degree angle to the horizontal surface. The system can capture three images at the same time: a top view image at the center and two side views of a foot. We also proposed an effective image analysis to interpret foot measurement data obtained from our setup. The results show that our setup to achieve promising results with maximum absolute difference (MAD) of less than 0.3mm. Keywords: multi-side capture, custom-fit shoes, 3d foot scanning.