A Review on Plant Diseases Classification Methods Using Machine Learning Algorithms

Plants are one of nature that are subjected to diseases. Even though it is a natural occurrence, plant diseases are the major problems in plantation industries. It can cause a huge loss in production. Some of plant diseases can be examined by our naked eyes and some of them cannot. The traditional way to detect plant diseases requires detailed monitoring and takes a lot of time especially in large farm. This paper presents several methods using different machine learning algorithms to detect different plant diseases on the different plants to achieve highest possible accuracy. Furthermore, an analytical review of various images processing method for recognizing plant diseases from captured images of plants is presented. Image processing techniques consist of image acquisition, image pre-processing, image segmentation, feature extraction, and classification. All these methods are explained in detail in this paper will be useful for farmers to gather information from the data given to take early action from getting worse. This paper also compares different techniques used by other researchers that consider various features and classifier in terms of their proficiency to enhance the classification methods’ accuracy ratios. From the analysis that we made; it shows the most used method with highest accuracy is based on convolutional neural networks (CNN).