Types of Plant Leaf Disease Detection Techniques Using Machine Learning Classification Algorithms

Abstract

Plant diseases are one of the most serious threats to agricultural production worldwide — shortening harvests and, in turn, threatening our food supply. Recognizing these illnesses in an early, high-accuracy manner is vital for their management and mitigation. Traditional approaches to plant disease diagnosis are normally based upon visual inspection by a few professionals but tend to be laborious, subjective, and error-prone. ML has become one of the effective methods for automating plant disease classification and improving the accuracy of analysis in detecting plant disease on leaf images only in the last decade. This article talks about the different machine learning classifiers algorithm used to detect the plant leaf disease. There are many approaches described in this paper, both traditional methods such as decision trees and support vector machines and new techniques with CNNs as well. The article talks about the color, texture and shape analysis features and their importance and feature extraction and its effect on classification accuracy. This contrasts the highs and lows of the algorithms across datasets and diseases. You can find articles of emerging advances such as explainable AI, transfer learning and lot of machine learning driven precision agriculture integrated with Internet of Things topics as well. Finally, this paper presents the research problems of systems available to detect plant leaf disease using machine learning algorithms and highlights some research directions that may be pursued in the future to solve these problems.