Exploring Classification of Rice Leaf Diseases using Machine Learning and Deep Learning

Abstract

Rice is third most commonly grain in the world. Some farmers chose rice cultivation over other crops because of rice's wide range of habitat adaptability and minor agriculture threat as the population slowly grows; by 2050, it is predicted that 14,886 million metric tonnes food will required to meet demand. Agribusiness contributes roughly 17% of India's GDP, and surveys indicate that roughly 70% of the population is either relying directly or indirectly on agriculture. A variety of ailments and infections in plants can be brought on by a variety of factors, including soil characteristics, environmental factors, the choice of undesirable crops, poor manure, and various leaf diseases. Plant diseases have a significant impact on agricultural production. These factors have a direct impact on the country's overall crop production. Continuous plant monitoring is required to prevent disease infection. Early plant disease detection is therefore of utmost importance in agriculture. The main motive of paper is to propose an effective and appropriate method for classifying various rice leaf diseases using deep learning approaches. Initially, classification was accomplished through the use of machine and ensemble learning classifiers. The outcomes were compared to CNN and transfer learning models. InceptionResNetV2 has the highest validation accuracy of 88 percent. According to the comparison, transfer learning models outperform machine learning classifiers.