Utilizing Artificial Intelligence for the Identification of Plant Species and Detection of Diseases through Deep Learning

Utilizing deep learning for the identification of plant species and the detection of diseases is crucial in mitigating the impact of these diseases on plant growth and crop yield. Early diagnosis is key to preventing losses and enhancing the overall quality and quantity of agricultural products. While various machine learning models have been employed for disease detection, the advent of deep learning has significantly increased the potential for improved accuracy in this research domain. Our proposed system aims to identify both the plant species and diseases affecting the leaves. A test database containing various plant diseases was established to assess the accuracy of our model. Training the classifier with the obtained data allows for more accurate predictions. We employed the GoogleNet model, which incorporates different layers designed to predict diseases effectively. In contrast to the conventional method where farmers rely on visual inspection and their knowledge of plant diseases, our system offers a more efficient and accurate alternative. Large-scale manual inspections are time-consuming, challenging, and often lack precision. Seeking expert consultation incurs substantial costs. To address these challenges, we implemented techniques involving automatic disease detection devices, enhancing accuracy while reducing costs and simplifying the process. This review comprehensively elucidates the utilization of deep learning models for visualizing various plant diseases. Additionally, it highlights research gaps, aiming to bring greater clarity to the early detection of plant diseases, even before symptoms become apparent.