EEDL-based Detection and Classification of Apple Foliar Leaf Disease

Apple farmers constantly grapple with the challenge of increasing their yield and safeguarding apple trees from diseases. The prevalence of diseases and pests significantly hampers apple production, resulting in substantial financial losses for the industry annually Detecting apple leaf diseases swiftly and accurately is crucial for effectively handling and curbing these issues within orchards. Specifically, advancements in computer vision methods that utilize deep learning have opened avenues for identifying and understanding these diseases at an early stage directly on the leaves Here, the EEDL(Enhanced Efficient Deep Learning)model is proposed for the detection and classification of five types of apple foliar diseases. In the EEDL model, we modify and fine-tune the Efficient Net model variant B0 by adding four additional layers. These layers consist of the augmentation layer, dense layer, dropout layer and the final output classifier dense layer. The AFD-7C dataset(8000 images are gathered in the real filed environment of apple research Centre and apple orchards of Himachal Pradesh. 9037 images are obtained from the Plant Village public dataset) is utilized in the experimental procedure and achieves an accuracy of 97.82%.