Published Paper Details:

This paper presents a novel approach to detect purple blotch disease in crops using Convolutional Neural Networks (CNNs). Purple blotch disease is a fungal infection that affects crops like onions, garlic, leeks, and shallots, leading to reduced yields and economic losses. Our method involves training CNN models on a comprehensive dataset of high-resolution images depicting onion leaves with and without purple blotch lesions. Data annotation is performed to label the affected areas, and preprocessing techniques such as data augmentation and normalization are applied to enhance model robustness. Transfer learning is employed to fine-tune pre-trained CNN architectures, improving classification performance. Experimental results demonstrate the efficacy of our CNN-based approach in accurately identifying purple blotch lesions on onion leaves. Furthermore, the trained models exhibit robustness when applied to real-world images captured under varying conditions. Precision and recall metrics are evaluated to assess model performance, and hyperparameter tuning is conducted to optimize accuracy. Our research contributes to advancing agricultural technology by providing a reliable and efficient method for early detection of purple blotch disease, thereby enhancing crop productivity and sustainability.