Identification of Downy Mildew and Leaf Miner Diseases in Hybrid Cucumber

Agriculture plays a crucial role worldwide. For those with little experience in agriculture, an automated system that can recognize diseases in plants from visual indications would be quite helpful. In countries like India, where the economy heavily relies on agriculture, plant diseases significantly impact the quality and quantity of agricultural products. To ensure global food security and maximize profits for stakeholders, accurate detection and classification of plant diseases are of utmost importance. Previous research has shown that the rising incidence of plant leaf diseases accounts for about 42% of losses in agricultural productivity. Several approaches have been made possible in this area with the introduction of deep learning-based image classification. However, the practicality of implementing these solutions on low-end devices necessitates the development of fast, precise, and computationally efficient systems. To address this, our work proposes a lightweight transfer learning-based approach for detection of two diseases namely Downy Mildew and Leafy Miner in plant leaves of Hybrid Cucumber. This approach incorporates an effective preprocessing method that enhances leaf images resulting in improved classification performance. Our proposed system employs a combined model comprising a pretrained MobileNetV2 architecture and a classifier network to extract features and make accurate predictions.