Predicting and Classifying Potato Leaf Disease using K-means Segmentation Techniques and Deep Learning Networks
Md. Ashiqur Rahaman Nishad, Meherabin Akter Mitu, Nusrat Jahan
Abstract
Potato is one of the most cultivated crops. Worldwide potatoes have its own cultivation priority as a staple food. For a successful potato production, we can develop a strong food security system as it is the great source of vitamins and minerals. However, several diseases affect potato production and degrade agricultural development. Therefore, diseases detection in early stage can provide a better solution for a successful crop cultivation. In this study, our aim is to detect and classify potato leaf diseases using deep learning algorithm. We applied K -means clustering segmentation and to increase model's efficacy, numerous data augmentation techniques have been applied on the training data. We have selected VGG16, VGG19, and ResNet50 network model. However, by using VGG16 we achieved 97% accuracy which is the best provided results among three networks. The recommended method outperforms several current methodologies as we compared the performances of the recent models according to relevant parameters.