Leveraging the CNN Approach for Predicting Severity Levels of Bacterial Wilt Disease in Potato Leaves

The purpose of this research is to find a solution to the problem of accurate and rapid illness assessment by developing an automated classification system that makes use of cutting-edge deep learning methods. A dataset with a total of 1600 high-resolution photographs of potato leaves that are infected by bacterial wilt disease was gathered and carefully categorized into four groups that represented different degrees of severity. These groups were mild, severe, serious, and deep. To learn complicated patterns and characteristics from the pictures, the CNN architecture used a sequential sequence of convolutional, pooling, and dense layers. After that, the performance of the model was tested with the use of an independent testing subset, and comprehensive metrics like accuracy, precision, recall, and F1-score were produced. This study not only contributes to the development of disease evaluation approaches but also lays the groundwork for both more research and its application in agricultural settings that take place in the real world.