VEGETABLE TYPE CLASSIFICATION USING NAIVE BAYES ALGORITHM BASED ON IMAGE PROCESSING

Abstract

There are so many different varieties of vegetables in Indonesia that the sorting procedure presents difficulties. In an effort to expedite the introduction of smart farming in Indonesia, more agricultural assistance techniques will be created. Utilizing the Naive Bayes algorithm is one way that may be used to advance agriculture in Indonesia. Image processing consists of converting RGB images to grayscale images, segmenting images using the thresholding method, collecting image features based on the HSV average value and object area, and classifying pictures using the Naive Bayes algorithm. This research seeks to use image processing technologies to agricultural products, particularly vegetables. The system is comprised of a single picture captured by a digital camera. There were eight varieties of vegetables employed for the picture data, with a total of eighty consisting of 64 training data and 16 test data. Spinach, green chilies, red chilies, chayote, cucumber, eggplant, tomatoes, and carrots were the vegetables categorized. The categorization findings indicate that 87.5 % of the test values produced using this approach are accurate. This study demonstrates that the Naive Bayes method has a high degree of accuracy for the categorization of vegetables based on image processing. It is anticipated that the findings of this study would promote the implementation of smart farming 4.0 in Indonesia.