Original scientific paper

https://doi.org/10.17559/TV-20240301001359

Image Multi-Threshold Segmentation Based on Variable Precision Rough Set and K-L Roughness Particle Swarm Optimization

Zhiyong She ; Information Network Security College of Xinjiang University of Political Science and Law, Tumxuk Xinjiang, 844000, P. R. China; No. 52 Qianhai East Street, Tumushuk City, Xinjiang *
Tao Song ; School of Intelligence Technology, Geely University of China, Chengdu Sichuan, 610000, P. R. China; No. 123, SEC. 2, Chengjian Avenue, Eastern New District, Chengdu City, Sichuan Province
Dongpo Zhang ; Information Network Security College of Xinjiang University of Political Science and Law, Tumxuk Xinjiang, 844000, P. R. China; No. 52 Qianhai East Street, Tumushuk City, Xinjiang
Yueping Feng ; School of Computer Science and Technology, Jilin University, Changchun, 130012, P. R. China; No. 2699 Qianjin Street, Changchun City, Jilin Province

* Corresponding author.

Abstract

This paper proposes an image multi-threshold segmentation algorithm based on variable precision rough sets and K-L roughness particle swarm optimization. The algorithm does not require a priori knowledge outside the image and employs variable precision rough sets to address the uncertainty problem in image segmentation. The optimal segmentation threshold is obtained by combining K-L divergence and roughness, and an improved particle swarm optimization algorithm is used to enhance segmentation efficiency. Experimental results demonstrate that the proposed algorithm effectively solves the uncertainty problem in segmentation and achieves better segmentation performance compared to other algorithms.

Keywords

K-L roughness divergence; multi threshold segmentation; particle swarm optimization algorithm; variable precision rough set

Hrčak ID:

328645

URI

https://hrcak.srce.hr/328645

Publication date:

27.2.2025.

Visits: 565 *