Automatic feature selection and classification for Apple Disease Prediction using CNN

The four most prevalent diseases that affect apple leaves are black rot, cedar rust, scab and leaf blotch. The spread of infection can be stopped and the healthful growth of the apple sector ensured by early detection and correct identification of apple leaf diseases. The Convolutional Neural Network (CNN) achieved a commendable accuracy of 89.42%. In comparison, the utilization of Inception v3 demonstrated a substantial enhancement, yielding an impressive accuracy of 92.50%. These results underscore the superiority of Inception v3 in accurately classifying apple leaf diseases, marking a significant advancement in predictive modeling for agricultural applications. The result is more superiority of MobileNet in classifying apple leaf disease more accurately with accuracy of 96.88%. In comparison of all three algorithms, MobileNet has more accuracy to offer a valuable tool for precise disease detection and informed decision-making for farmers, ultimately contributing to enhanced agricultural productivity and sustainable farming practices.