Intelligent Recognition of Leaf Abnormalities Using Deep Learning Method

Deep Neural Networks (DNNs) have significantly improved the detection of different diseases affecting plants. The impact of unidentified and untreated plant diseases is to reduce crop yield significantly, as they negatively affect the quality and quantity of horticultural and agricultural products. Therefore, early recognition of plant diseases is essential for improving crop quality and reducing production losses. To this end, a modern framework has been developed that incorporates image acquisition and analysis, utilizing DNN models to calculate various evaluation metrics such as Recall, F1 scaling, accuracy, and precision. By diversifying the results, it becomes possible to differentiate between healthy and bacterial plant leaves. The proposed methodology has been tested on two types of plants - pepper, and potato - and can determine the percentage of infection on plant leaves.