Plant disease diagnosis using image processing techniques -A review on machine and deep learning approaches

Abstract

Diagnosis and monitoring of plant diseases at early plant growth stages are crucial to minimizing the disease dissemination and it also facilitates effective plant protection measures. Thus, the automatic and real-time system for the diagnosis of plant diseases is very much needed in the current era of agricultural information. Several techniques and/or approaches of image processing have been studied to address the challenges in the diagnosis of plant diseases through acquired images. Of them, conventional machine learning and advanced deep learning (particularly CNN) approaches are nowadays explored considerably and have demonstrated promising and relatively an accurate classification than those of conventional approaches. The present literature review aims to present and discuss the potential applications of these techniques.