International Journal of Scientific Research and Engineering Development

The article introduces a lightweight computer vision-based solution to the problem of estimating human foot size and suitable shoe size based on a single monocular image, without making use of deep learning models. The algorithm exploits a combination of classical image processing methods, such as HSV conversion, Gaussian smoothing, segmentation and contourbased measurement methods, and geometric scaling with the help of an object of reference (A4 paper). The proposed method provides a low-resource and interpretable alternative, which is compared to the deep learning methods that need large datasets and high computational resources. The video on a curated dataset displaying experimental evaluation shows an average error in measurements of 3.2, which is competitive. Another measure that the system takes into account is the analysis of asymmetry, since feet are measured separately, which is a weakness of traditional techniques. The outcomes reveal the usefulness and concreteness of the method as well as define possibilities of improvement of the approach in the future.

Ayush Kumar Maheshwari, Akshat Bari, Himesh Kumar,"ShoeMorph: Exploring Classical Computer Vision Techniques for Asymmetry-Aware Foot Size Estimation" International Journal of Scientific Research and Engineering Development, V9(2): Page(2286-2291) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.