Textile and Apparel Science

Color histogram analysis of virtual garment fit image for automated fit evaluation

---

Abstract

Clothing fit is a major factor influencing the customer's purchase decision, but it is difficult to directly evaluate the physical fit of the garment online. Using virtual fitting allows us to assess specific fit issues for different body parts. However, evaluating garment fit through virtual simulation typically relies on expert judgment, limiting its scalability. To overcome this, the study quantitatively analyzes fit images using histogram algorithms, enabling mass customization by assessing fit across various body types. Clothing patterns were draped on human models of various body types, and garment fit images were represented by strain maps and distance maps. The front, side, and rear views of the wearer were then captured and used for analysis. This method aims to enhance customization in clothing development by providing a rapid, objective assessment of similarity among numerous images, reducing dependency on expert evaluation.

Keywords: garment fit analysis, virtual fitting, histogram algorithms, body types, image similarity, clothing development, color maps, bhattacharyya method

How to Cite:

Heo, S., Kang, Y., Kim, S. & Oh, J., (2024) “Color histogram analysis of virtual garment fit image for automated fit evaluation”, International Textile and Apparel Association Annual Conference Proceedings 80(1). doi: https://doi.org/10.31274/itaa.17286