Design of an Automated Plant Disease Detection System Using Digital Image Processing and Neural Networks

International Journal of Computer Science and Technology | Vol. 10, No. 7, July 2008 | pp. 49–56

DOI: 10.46882/2008/IJCST/107016

Research Article

Title: Design of an Automated Plant Disease Detection System Using Digital Image Processing and Neural Networks

Names of Authors: Chidi N. Okafor¹, Blessing C. Eze²

Authors’ Affiliations: ¹Department of Electronic Engineering, University of Nigeria, Nsukka, Nigeria; ²Department of Agronomy, Michael Okpara University of Agriculture, Umudike, Nigeria

Abstract: Early detection of agricultural crop diseases is imperative to minimizing yield losses and ensuring global food security. Manual inspection methods are error-prone, highly subjective, and impractical across massive agricultural landscapes. This paper describes the design and testing of an automated leaf disease detection engine tailored for cash crops. The framework uses color segmentation via K-means clustering within the L*a*b* color space to isolate necrotic lesions from healthy leaf tissue. Spatial and statistical properties are extracted using a Gray-Level Co-occurrence Matrix (GLCM), which feeds into a Multi-Layer Perceptron (MLP) neural network trained backpropagation. The system was evaluated on a dataset of 1,200 digital leaf images encompassing healthy, black spot, and rust-infected specimens. The automated diagnostic engine achieved a classification accuracy rate of 94.6% across all test sets. The processing pipeline completed feature extraction and neural inference within an average duration of 720 ± 25 milliseconds per leaf sample, demonstrating high suitability for edge-deployment on mobile diagnostic devices.

Keywords: Image processing, K-means clustering, neural networks, plant disease detection, pattern recognition, precision agriculture

Manuscript Timeline: Received April 9, 2008; Revised May 22, 2008; Accepted June 14, 2008; Published July 7, 2008

Citation: Okafor, C. N., & Eze, C. C. (2008). Design of an Automated Plant Disease Detection System Using Digital Image Processing and Neural Networks. International Journal of Computer Science and Technology, 10(7), 49–56.

Subscribe to read the full article: https://www.internationalscholarsjournals.org/subscribe-to-read