Revolutionizing agricultural productivity with automated early leaf disease detection system for smart agriculture applications using IoT platform
摘要
In smart agriculture, identifying plant illnesses early is vital because it prevents diseases from spreading. A novel statistical approach-based plant disease detection system is developed to identify and categorize diseases early. A novel statistical threshold value is utilized to divide apart the sick portion of the leaf, and a co-occurrence matrix with grey levels is employed to gather the properties. The system uses a visual sensor to record leaf images from the field. The system is tested on publicly available datasets and real-time images captured from fields. Classification is carried out using the support vector machine to achieve higher accuracy. Performance metrics like classification and detection accuracy are used to assess the highlighted system’s effectiveness. The suggested method outperforms the existing state-of-the-art methods with an average detection accuracy of 98.43% and classification accuracy of about 99.9%.