Deep Learning Based Classification of Apple Leaf Diseases Using AlexNet

Authors : Felix Olanrewaju Babalola, Nekabari Isabella Kpai, Önsen Toygar

Pages : 67-74

Doi:10.53070/bbd.1349566

View : 54 | Download : 60

Publication Date : 2023-10-18

Article Type : Review Paper

Abstract :The diagnosis of a disease on the plants is a critical step in avoiding a significant loss of harvest and agricultural product amount. The indications can be found on parts of plants such as fruits, leaves, lesions, and stems. The leaf demonstrates the symptoms by changing, and therefore revealing the spots on it. This disease identification is accomplished through manual inspection for pathogen detection, which might take extra time and cost. Hence, automatic detection of plant diseases can be vital in the agricultural economy. This study proposes the use of a simple deep learning model, AlexNet, for detecting anomalies in apple leaves in order to predict the presence or absence of a disease in a tree correctly. The Convolutional Neural Network model is implemented using the Plant Village dataset, augmented to 12,624 images for proper training. The proposed apple leaf disease categorization system achieves an overall accuracy of 99.56 percent. For comparison of results, a different method, namely Binarized Statistical Image Features (BSIF), is also implemented. Furthermore, the results are juxtaposed against studies using similar state-of-the art approaches.
Keywords : Elma yaprağı hastalıkları, tespit, derin öğrenme