Detection of Rice Blast Disease (Magnaporthe grisea) Using Different Machine Learning Techniques
Bidisha Chakraborty
Government College of Engineering and Leather Technology, Kolkata, West Bengal, India.
Santanu Banerjee *
Chhatrapati Shahu Ji Maharaj University, Kanpur, Uttar Pradesh, India.
Sanjoy Samanta
Department of Entomology, Uttar Banga Krishi Vishwavidyalaya, Coochbehar, 736165, West Bengal, India.
Udit Debangshi
Department of Agronomy, Institute of Agriculture, Visva-Bharati, Sriniketan, West Bengal, India.
Sandhya V. Yadav
Fergusson College (Autonomous), Pune (MH), 411005, India.
Pravin B. Khaire
Department of Plant Pathology & Microbiology, Mahatma Phule Krishi Vidyapeeth, Rahuri (MH), 413722, India.
Vaibhav B. Shelar
Biological Nitrogen Fixation Lab, College of Agriculture, Pune (MH), 411005, India.
Govardhan D. Bansode
Fergusson College (Autonomous), Pune (MH), 411005, India.
Kiran B. Landage
Biological Nitrogen Fixation Lab, College of Agriculture, Pune (MH), 411005, India.
*Author to whom correspondence should be addressed.
Abstract
Rice is one of the most important staple food crops in the world. Most Asian countries are dependent on rice and huge quantities of rice are grown every year. However, there are many categories of diseases (e.g., blast) which affect rice production and can ultimately lead to huge financial loss to rice growers. Yield loss due to rice blast disease about 10 to 30 percent annually and under favourable condition, this disease can destroy the rice plant within 15 to 20 days and cause yield loss up to 100%.Therefore to ensure better quality, quantity and better productivity early disease detection should be done so that the right amount of pesticides can be as administered at right time to curb the infection. Nowadays Machine Learning has been integrated into the agriculture sector. The aim of this review paper is to identify which Machine Learning algorithms work best in rice blast disease detection. The algorithms reviewed here include Naive Bayes, LSTM RNN, Random Forest Classifiers, Support Vector Machines, K Means, Decision Tree and Convolutional Neural Networks. This review paper also covers the future scope of improvement of some Machine Learning algorithms like Naive Bayes and Recurrent Neural Networks.
Keywords: Rice, Magnaporthe grisea, detection, machine learning, algorisms