Detection and Classification of Plant Diseases Using SVR

Abstract

Accurate and fast detection of plant diseases can be a great boon to crop yields. Curbing the complete cost to affordable amount is also a serious concern. The present manual technique for the detection of disease is a time-consuming process and many times farmers with humble background cannot afford it. Thus, an automation is needed to make this hectic process fast and well within budget of farmers with low budget. This paper discusses the monitoring of plant diseases using image processing by taking samples of various leaves. In the initial phase, training dataset is created from the collected and enhanced images. Then, a test dataset is prepared arbitrarily and SVR, CNN is utilized for obtaining the classification results. Identification of leaf diseases is the key for preventing the losses in the yield and quantity of the agriculture product. It is very difficult to monitor the plant diseases manually. Leaf diseases can be detected by image processing technique. CNN algorithm is used for leaf diseases classification. Disease detection and classification involve steps like image pre-processing, image segmentation, feature extraction, classification. To detect the plant diseases and to provide the solutions(pesticides) to recover from the diseases.