Combination of k-means with naïve bayes classifier in the process of image classification

Abstract

In this paper, the data which has been tested is the image of bananas extracted based on the colour characteristics and statistics, namely red, green, blue, hue, saturation, value, average, variance, skewness, kurtosis, and energy. Image data is extracted as much as 60 data for training data, and images for test data as many as 105 data. From the 105 image data has been tested, the number of correct test data is 89 data and the number of incorrect test data is 16 data. The total percentage obtained from the test is 85%. The results obtained are a collection of images based on the level of maturity that is mature, medium, or raw.