A Deep Learning based-Convolutional Network Model for Wheat Leaf Disease Identification

Damage to cereal crops can be serious when this pathogenic fungus is present. Their cultural influences are diverse. Though it is difficult to eradicate an entire illness on a large scale, one effective technique is to monitor crop fields for signs of the disease so that preventative measures can be taken. Digital image analysis for disease identification, with the option to collect them in the field using handheld equipment, is an effective control strategy. Here, we present a technique that can detect wheat seedlings for five different fungal, either singly or in combination, and also predicts the disease stage. The proposed model has classified 1972 photos of wheat fungal illnesses into several databases. Unique disease labels are present in over 80% of the images in the collection (including seedlings), while over 13% have suggestions for healthy plants and 6% have labels for various ailments. This set was created using a method that used an image hashing algorithm to reduce the issue of training data homogeneity. An Efficient Net-the illness diagnosis algorithm utilizes based convolutional neural network. The most accurate networks were those that trained using image style augmentation and transfer (0.9897).