AI Based Automatic Crop Disease Detection System

India is a land of Agriculture. Agriculture contributes 17% to the total Gross Domestic Product (GDP). Hence, it is an important sector of Indian economy. India ranked second in farm outputs worldwide. Each crop is prone to particular diseases that will affect the quantity and quality of the yield potential. Average yield loss for most of the important food crops is due to crop diseases, which contributes about 42% of the crop failure. In many cases, the whole crop production is destroyed due to crop diseases. Crop production is affected by number of diseases worldwide. Timely detection of diseases will allow to monitorand imlement control measureswith greaterefficiency. The main aim of this project is to design an AI-Based disease detection system that detects the type of disease present in tomato leaf by clicking the images of various leaves through camera and spray the respectivepesticide to the diseased partof the plant. Deep Convolutional neural networks (D-CNN) and transfer learning techniques are used to detect and classify the disease. Three transfer-learning models i.e., AlexNet, InceptionV3 and ResNet152V2 were used for training and classification and the results were compared which gave the accuracy of 95%, 97% and 99% respectively.