Artificial Intelligence Techniques for Early Plant Leaf Disease Detection: A Comprehensive Review

Abstract

The agricultural area is still a major provider to many countries' economy, but diseases that continuously infect plants represent continuous threats to agriculture and cause massive losses to the country's economy and it is a main threat to the food security. There are a large range of diseases that attack the plants which have its symptoms and pathogen categories. These diseases control the yield and quality of plants and early detection can reduce disease severity and protect crops. Traditionally, farmers use their own experience or hire experts for identification of diseases in their crops but this method requires a thorough knowledge of plant pathogens and take a lot of time and is also prone to being mistaken with a high error rate. Now a day’s artificial intelligent techniques are introduced widely to identify plants diseases in a short time and with low error rate and high accuracy. This paper reviews the problem of plant diseases detection and recognition, it overviews the main diseases which can infect the plants, divides them into categories and discusses the symptoms of each disease. After that, it overviews more than fifty current states of the arts related to plants diseases detection using machine learning and image processing algorithms, deep learning algorithms and Internet of Things, comparing between a lot of them in aspects of accuracy and time complexity. Finally, it overviews the main challenges facing researchers in this aspect, and based on these challenges, we propose a plant diseases detection framework using deep learning algorithms.