Manuscript received March 4, 2026; revised March 25, 2026; accepted June 4, 2026; published October 23, 2026.

Abstract—Plant diseases pose a major challenge to the world's agricultural productivity and food security. It makes sense to address this challenge through the development of accurate, efficient and automated disease detection techniques. The study presents a hybridized method of automated procedures based on particle swarm optimization-adjusted support vector machine-along-with sophisticated image processing to the automated detection and classification of the Armenian cucumber leaf diseases. Proposed methodology uses a successive series of seven steps for the process of image acquisition, noise removal by mean filtering, contrast enhancement using Contrast Limited Adaptive Histogram Equalization (CLAHE), image segmentation using Fuzzy C-Means (FCM), feature extraction by using Principal Component Analysis (PCA) and Disease classification using Particle Swarm Optimization-Support Vector Machine (PSO-SVM), Back Propagation Neural Network (BPNN) and random forest. An experiment was carried out on 440 Armenian cucumber leaf images (260 diseased, 180 healthy), exploited for testing purposes, where the PSO-SVM classifier offers optimum classification performance 97.3% accuracy, 96.8% sensitivity, 98.1% specificity, 97.5% precision and 96.8% recall, outperforming conventional machine learning methods considerably. An analysis of the available deep learning-based methods shows that the methodology is cheaper, highly accurate, and scalable in plant disease classification. With optimized feature selection and robust image pre-processing along with the use of classifiers based on machine learning, a reliable plant health is proposed. The research work considerably contributes to smart farming in farming in agricultural sector. It deploys an automated and low-cost solution for early disease detection in the field. As opposed to deep learning models, which are computationally expensive, the PSO-SVM model is a lightweight and efficient model that can be applied in real-time agricultural models. The model has an accuracy of 97.3%, sensitivity of 96.8%, and specificity of 98.1%, which proves that the model can be used effectively in practice.

Cite: Fida'a Al-Quran, Malik Jawarneh, Omar Isam Al Mrayat, Ghaith M. Jaradat, Said Mohammad Ahmad Al-Tahat, and We'am Adel Talafha, "A Hybrid PSO-SVM Approach for Automated Armenian Cucumber Leaf Disease Detection Using Advanced Image Processing Techniques," Journal of Image and Graphics, Vol. 14, No. 5, pp. 929-942, 2026.

Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).