International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 11 | Nov 2024
p-ISSN: 2395-0072
www.irjet.net
Deep Learning-Based Agricultural Image-Based Leaf Disease Detection in Crops Nidhi K M1, Kannika B R2, Pooja Balaganur3, Anusha M4 1234 Bachelor of Engineering, Information Science and Engineering, Bapuji Institute of Engineering and Technology
Davangere, affiliated to VTU Belagavi, Karnataka, India ---------------------------------------------------------------------***--------------------------------------------------------------------Abstract - One area of artificial intelligence is deep learning. The benefits of feature extraction and autonomous learning have made it a hot issue in academia and business in recent years. Natural language processing, audio processing, and picture and video processing have all made extensive use of it. It has also developed into a hub for agricultural plant protection research, which involves diagnosing plant diseases and evaluating the range of pests. Deep learning may be used to detect plant diseases without the drawbacks of artificial selection of disease spot features, as well as increasing the objectivity of plant disease feature extraction and accelerating technological development and research efficiency. An overview of the state of research on deep learning technology is given in this report. in the area of deep learning using agricultural images. Agriculture, which contributes most to expanding economies and populations, is essential to the availability of high-quality food. Plant diseases have the ability to wipe out species variety and result in large losses in food production. Using precise or automated detection methods for early plant disease diagnosis could improve food production quality and lower losses. Deep learning has significantly increased the recognition accuracy of object detection and picture classification systems in recent years. Therefore, we employed convolutional neural network (CNN)-based pre-trained models in our work to efficiently identify plant illnesses. We concentrated on fine-tuning the hyperparameters of many popular pre-trained models, including Inception V4, DenseNet-121, ResNet-50, and VGG-16. Key Words: CNN, deep learning, transfer learning, leaf disease, and patholog
1.INTRODUCTION Agriculture, being a substantial contributor to the world’s economy, is the key source of food, income, and employment. In India, as in other low- and middle-income countries, where an enormous number of farmers exist, agriculture contributes 18% of the nation’s income and boosts the employment rate to 53% [1]. For the past 3 years, the gross value added (GVA) by agriculture to the country’s total economy has increased from 17.6% to 20.2% [2,3]. This sector provides the highest share of economic growth. Hence, the impact of plant disease and infections from pests on agriculture may affect the world’s economy by reducing the production quality of food. Prophylactic treatments are not effective for the prevention of epidemics and endemics. Early monitoring and proper diagnosis of crop Infected plants typically have noticeable marks or spots on their stems, fruits, leaves, or flowers; more specifically, each infection and pest condition leaves distinct patterns that can be used to diagnose abnormalities; identifying a plant disease requires expertise and manpower; additionally, manual examination when identifying the type of infection of plants is subjective and time-consuming; and, occasionally, the disease identified by farmers or experts may be misleading [4]. Determining the types of plant diseases is a critical issue that requires careful attention to detail. Plant Village is a plant disease dataset released by Pennsylvania State University [17]. There are 38 plant disease classes and 54,305 RGB photos in Plant Village. It includes pictures of fourteen distinct plants. Every plant has a minimum of two 256 × 256 picture classes that depict both healthy and damaged leaves. Figure 1 displays a selection of the dataset's pictures. Numerous studies on plant disease identification have been conducted after the dataset's publication [18–21]. CNN deep learning models are widely used in image-based research. They are effective at extracting simple, low-level features from pictures. Unfortunately, training deep CNN layers is challenging due to the high computational cost. Several researchers have suggested transfer learning-based models to address these problems [22–26]. The VGG-16, ResNet, DenseNet, and Inception models are well-liked transfer learning models [27]. The ImageNet dataset, which comprises several classes, is used to train these models. Because image characteristics like edges and contours are shared by all datasets, these models can be trained on any dataset. Thus, the most appropriate and reliable model for image classification has been determined to be the transfer learning approach[28].
© 2024, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 715