Foot shape classification methods based on image processing for shoe manufacturing

In this paper, we propose two approaches for classifications of foot shape using image processing technologies. The main concern of the proposed method is to classify the foot shapes based on the foot index and the toe length. To this end, the proposed method employs YCbCr color space, contour function, and convex hull algorithm to detect the contour and end points on the foot image and calculates the distance between end points on the contour. We analyze the results of foot classifications to investigate how well the proposed method classifies the foot image. Our numerical results show that the proposed method classifies the foot shape among slender, standard, and broader types by calculating the foot index of the image. In addition, the foot image is classified into Egyptian, Squared, and Greek types by calculating the length of a big toe and the second toe when the proposed method is used. This implies that the proposed method provides useful information for manufacturing shoes suitable for the foot shape of customers.