A Review on Implementation of Deep Learning Models for Image Based Diseases Detection in Rice Plants

Abstract

Rice is mostly an important part of every Asian meal and its absence may affect the usually followed diet structure. Rice plant disease detection is getting more attention these decades if compared with the last decade and the reason is simply the impending increasing food requirements. Keeping the focus on present trends, this review paper is concentrated to review multiple methodologies that had been attempted to detect disease in rice plants with various proposed models. Several methods and models, implementation and achieved results are also being discussed in this paper. Rice plants' diseases that are most commonly seen are leaf blight, brown spot, leaf blast, and sheath blight. Artificial intelligence and machine learning are getting part of the most require a revolution in the agriculture field. Since the population is growing swiftly and in the future world may confront with food abate. It is necessary to control the graving of rice plants. Machine learning algorithms such as CNN (convolution neural network), segmentation, classifiers, other neural networks, and decision trees had been already taken into action for disease detections. Different algorithms have a different mechanism and so does the outcome. But one factor that may be considered as effective in the outcome of any machine learning model is the number of the dataset being utilized for training and testing the particular model. With upcoming new technologies, harmful radio waves can become causes not only in rice plants but different plants changes in their heath. In addition to literature, some survey papers are also part of these papers. The whole agenda of this review paper is to understand the major outcomes of attempted detection of rice plant disease using new technologies and how it can affect the various aspects of rice plants.