Advanced CNN Architectures for Automated Identification of Plant Diseases
Bhimanshu Singh
Gurdeep Singh
Dr. Kavitha R
Department of Computing Technologies College of Engineering and Technology SRMInstitute of Science and Technology Kattankulathur, Chengalpattu, Tamilnadu, India-603203
Department of Computing Technologies College of Engineering and Technology SRMInstitute of Science and Technology Kattankulathur, Chengalpattu, Tamilnadu, India-603203
Department of Computing Technologies College of Engineering and Technology SRMInstitute of Science and Technology Kattankulathur, Chengalpattu, Tamilnadu, India-603203
ba4762@srmist.edu.in
gg3846@srmist.edu.in
kavithar14@srmist.edu.in
Abstract—— Medicinal plants are vital to both agriculture and healthcare but are highly susceptible to various leaf diseases, which can significantly reduce yield and quality. Traditional methods for identifying plant diseases rely on visual inspection by farmers or experts, often leading to inaccuracies and delays due to human error, especially in rural or resource-limited areas where expert knowledge is scarce. Climate change is altering disease patterns, making traditional detection methods less reliable over time. project aims to develop an AI-powered system using Convolutional Neural Networks (CNNs) to automatically detect medicinal plant leaf diseases, offering rapid and accurate diagnosis via a user-friendly web application. The system integrates image preprocessing, model training, and real-time treatment recommendations, providing valuable guidance but requiring reliable internet connectivity .However, the system's performance depends on the quality and diversity of the dataset and may face challenges in generalizing across different environments and adapting to local agricultural practices or resources.
Keywords—Plant disease detection, Resnet, Convolutional Neural Network, image processing, agriculture technology, Dense Net
