Tomato's Disease Detection by Deep Learning Using CNN

Abstract

Leaf disease have been one of the riskiest impendences to food society, because it cut down the crops yield and compromise the quality. Tomato is one of the most demandable vegetables in Bangladesh. Due to the insect and disease of leaf the yield falls down and cultivating system going through an unclear future. Most of the farmers and agricultural farm are going to off their agricultural profession. It hearts us most. We decided to build a model that can detect the disease earlier using the affected leaf image by CNN with 87.5% accuracy which is cleat how accurate CNN work for image classification. We took 1359 data and 80% use for training and 20% data for testing. The model has not needed any human supervisor for using its any features. The CNN model takes its decision with Convolutional layer, Maxpooling layer, Flatten layer, Dense layer. Difference layer has different work and by a fully supervision all the layer the decision has been taken. This detection includes some phase like image acquirements, pre-processing of the image, data segmentation, classification and feature education.