USE OF TRANSFER LEARNING AND MACHINE LEARNING FOR DISEASE PRONE DETECTION IN WATERMELON FRUIT

ABSTRACT

Given that consumption has increased as a result of medical advice, watermelon crop disease detection is essential for maintaining its health. The Jupyter IDE, libraries, Kaggle dataset, preprocessing, constructing CNN model, training, and testing steps of a deep learning-based technique were suggested. The pre-selected box setting method used by the CNN model was enhanced and tested repeatedly, yielding an average accuracy of 98.9%. The CNN model beats other machine algorithms, such as SVM, ANN, KNN, and Naive Bayes, in identifying watermelon sickness in a neural environment. More research is required.

Keyword: Watermelon, CNN, Disease Prune, Transfer learning, Machine Learning