An Automated CNN Model Based Approach for Apple Leaf Disease Detection

Abstract

This study presents a deep learning approach for the detection of diseases in apple leaves. The method utilizes convolutional neural networks (CNNs) to classify apple leaf images into diseased and healthy categories. The dataset used for training and testing the CNNs consisted of images of apple leaves infected with various diseases, as well as healthy leaves. The results of the study demonstrate that the proposed deep learning approach is able to accurately detect and classify apple leaf diseases with high accuracy. This approach can potentially be used in precision agriculture to improve crop yields and reduce the use of pesticides. Deep learning algorithms can correctly identify misleading leaf photos, enabling farmers to accurately detect leaf illness and take immediate action in consonance with the disorder. We subsequently compiled data from the nursery and preprocessed it to make sure that it was congruent with the particular model that we adopted in order to categorize each ailment in our study. In order to achieve a superior performance, we later amended and used a CNN model that had become viable with our dataset. This essentially serves as our testimony that 99% of the leaves are tainted.