Enhancing Crop Health by Identifying Leaf Disease with High Accuracy Using CNN

In recent years, plant diseases have become a top concern for farmers and agriculturalists. In this paper, seven plant species like apple, tomato, potato, grape, bell pepper, corn, peach leaf diseases have been found in healthy classes as well. The system uses a pre-trained CNN to extract features from an image of the plant leaf as its input. The dataset for the study is a collection of images of various diseased plant leaves. Back propagation and stochastic gradient descent are used to train many layers of the CNN architecture-convolutional, pooling, and fully connected were applied in this study. The extracted traits are then classified using a fully connected layer to determine whether the leaf is afflicted with disease. When compared to K-Means, SVM, RBF, MLP, BPNN, and other similar algorithms which have an accuracy of less than 90%, the CNN algorithm is one of the most accurate. For the implementation of the leaf disease detection, we make use of Python software. Farmers can use the proposed system to predict disease in the early stage and can give diagnosis to plant diseases, allowing for prompt treatment and control measures.