AGRICULTURAL PLANT DISEASE DETECTION AND IDENTIFICATION

Since agricultural crops farming takes considerations from various variables such as environment, soil and sunlight, existence of diseases cannot be avoided. The main objective of this proposed work is to develop an Embedded System for plant disease identification using Convolutional Neural Network to identify which of the diseases is present on the agricultural crop plants. The system can obtain accurate perception of plant disease in agriculture field and then transmit the same to users. Results are observed in mobile app and displayed on mobile screen. The system considered different environmental parameters like Temperature, Humidity and Soil Moisture and used Raspberry Pi as Processor, and sensors for sensing environmental conditions. Study developed the innovative solution that provides efficient disease detection in agricultural crop plants based on Lenet-5 Architecture. The model is trained with batch size of 150 epochs. The system is designed to identify the main diseases namely Pepper bell bacteria spot, Early and late blight of potato and tomato, Tomato target spot, Tomato mosaic virus, Tomato yellow leaf curl virus, Tomato bacterial spot, Tomato black mold, Tomato septoria leaf spot and Tomato spider mites two spotted spider mites. We have trained a deep convolutional neural network on a dataset of diseased and healthy agricultural crop plant leaves under controlled conditions to identify diseases or absence thereof