Biotechnological Advances in Crop Disease and Pest Detection: Image Segmentation Using G-R Component and K-means

Abstract

This paper addresses the challenge of detecting crop diseases and pests against complex backgrounds by introducing an innovative image segmentation method that leverages the G-R component and K-means clustering algorithm. Initially, crop image data is preprocessed using the G-R component to enhance the contrast between healthy and affected areas of the crop. Subsequently, a median threshold transformation prepares the images for the final segmentation phase, where K-means clustering effectively isolates regions indicative of pest and disease presence. The proposed method demonstrates significant potential for application in agricultural biotechnology, particularly in enhancing the accuracy and efficiency of pest and disease identification in crops. Despite its promise, the initial trials revealed limitations due to the small size of the database, which did not adequately represent the variability of diseases and pests across different regions. To address this, we propose expanding the database to include a wider range of conditions and integrating advanced image segmentation and recognition technologies. Enhancing the database and employing more sophisticated algorithms are expected to improve diagnostic capabilities and, consequently, contribute to increased crop yields. This study underscores the critical role of biotechnological innovations in advancing agricultural practices, particularly through the application of image processing techniques to support better crop management and disease control strategies.