Predicting Apple Plant Diseases in Orchards Using Machine Learning and Deep Learning Algorithms

• Published in: SN computer science Vol. 5; no. 6; p. 700
• Main Authors: Ahmed, Imtiaz, Yadav, Pramod Kumar
• Format: Journal Article
• Language: English
• Published: Singapore Springer Nature Singapore 01.08.2024; Springer Nature B.V
• Subjects: Agribusiness; Agricultural production; Agriculture; Algorithms; Apples; Artificial neural networks; Automation; Color imagery; Computer Imaging; Computer Science; Computer Systems Organization and Communication Networks; Crop diseases; Crops; Data Structures and Information Theory; Datasets; Deep learning; Defoliation; Disease prevention; Effectiveness; Illnesses; Image quality; Infections; Information Systems and Communication Service; Machine learning; Neural networks; Original Research; Pattern Recognition and Graphics; Plant diseases; Remote sensing; Security for Communication and Computing Application; Signs and symptoms; Software Engineering/Programming and Operating Systems; Vision; Deep learning; Convolutional neural network; Apple disease classification; Machine learning
• ISSN: 2661-8907; 2662-995X; 2661-8907
• DOI: 10.1007/s42979-024-02959-2

Apple cultivation in the Kashmir Valley is a cornerstone of the region’s agriculture, contributing significantly to the economy through substantial annual apple exports. This study explores the application of machine learning and deep learning algorithms for predicting apple plant diseases in orchards. By leveraging advanced computational techniques, the research aims to enhance early detection and diagnosis of diseases, thereby enabling proactive disease management. The study utilizes a dataset comprising diverse environmental and plant health factors to train and validate the models. Key highlights include the comparative analysis of machine learning and deep learning approaches, the identification of optimal feature sets, and the assessment of model performance. The findings contribute to the development of efficient and accurate tools for precision agriculture, facilitating timely intervention and sustainable orchard management. The apple industry in Kashmir faces a significant challenge due to the prevalence of various diseases affecting apple trees. One prominent disease that adversely impacts apple yields in the region is the Apple Scab, caused by the fungus Venturia inadequacies. Apple Scab is characterized by dark, scaly lesions on leaves, fruit, and twigs, leading to defoliation and reduced fruit quality. The disease thrives in cool and humid conditions, which are prevalent in the Kashmir Valley. This study addresses the limitations of traditional, labor-intensive, and time-consuming laboratory methods for diagnosing apple plant diseases. The goal is to provide an accurate and efficient deep learning-based system for the prompt identification and prediction of foliar diseases in Kashmiri apple plants. Our study begins involves the creation of a dataset annotated by experts containing approximately 10,000 high-quality RGB images that illustrate key symptoms associated with foliar diseases. In the next step, an approach to deep learning that utilizes convolutional neural networks (CNNs) was developed. Comparative analysis five different deep learning algorithms, including Faster R-CNN, showed that the method was effective in detecting apple diseases in real time. The proposed framework, when tested, achieves state-of-the-art results with a remarkable 92% accuracy in identifying apple plant diseases. A new dataset is presented that includes samples of leaves from Kashmiri apple plants that have three different illnesses. The findings hold promise for revolutionizing orchard management practices, ultimately benefiting apple growers and sustaining the thriving apple industry in the Kashmir Valley.