Abstract—In this paper we propose an improved procedure for Chronic Wound (CW) assessment. First, the optimal threshold for segmenting the wound area from the background is selected. This is done using a combination of Otsu’s method and Particle Swarm Optimization (PSO). The color features of the image are then extracted using K-means clustering to monitor the progression of wound healing. Fifty CW images from the Medetec medical image database were analyzed to compare the efficiency of the proposed method with that of the traditional Otsu method. The proposed method achieved improved image segmentation in 46% of cases, while in 16% quality was unchanged and in 38% quality was degraded. We also demonstrated the application of K-means clustering to classification of wound tissue. This allows the progression of tissue repair to be assessed and impaired healing to be identified via a telemedicine system, of particular salience in remote areas where clinical expertise in wound management is lacking.

Index Terms—swarm intelligence, image processing

Cite: Wanyok Atisattapong, Chontida Chansri, Jidapa Somboonbadeebut, and Pakorn Songkaew, "Identifying the Optimal Threshold for Image Segmentation Using PSO and Its Application to Chronic Wound Assessment," Journal of Image and Graphics, Vol. 10, No. 3, pp. 116-121, September 2022.

Copyright © 2022 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.