Analysis Naïve Bayes In Classifying Fruit by Utilizing Hog Feature Extraction

Abstract

Indonesia has abundant natural resources, especially the results of its plantations. Lots of local fruit that can be used starting from the root to the skin of the fruit. Local fruit can be consumed as fresh fruit and can also be processed into drinks and food. This is reflected in the diversity of tropical fruits found in Indonesia. Fruits that are rich in benefits and can be used as medicines such as Apples, Avocados, Apricots, and Bananas. These fruits are often found around us. In Indonesia these fruits are produced and also exported abroad. However, the limited methods and technology used to classify this fruit are interesting things to discuss and become the main focus in this research. This study analyzed using the Naive Bayes algorithm and feature extraction of HOG (Oriented Gradient Histogram) to obtain more effective classification results. The results showed that the collection of fruit using the Naive Bayes method and HOG feature extraction had not yet obtained maximum classification results, only with an accuracy of 56.52%. Keywords – Apple, Avocado, Apricot, Banana, Naive Bayes, HOG.