Perbaikan Kontras Citra Dengan Ekualisasi Histogram Dan Gaussian Pada Klasifikasi Semangka

Abstract

The Maturity of watermelons can be recognized based on watermelon skin textures. The similarity of watermelon skin texture to the background of this research was carried out. The introduction of skin texture can be extracted first-order statistical features. With the application of histogram equalization, it is expected that the statistical feature calculation process is more accurate. The use of a Gaussian filter is used to improve the image of the histogram equalization result. Support vector machine (SVM) is one method that can be used as a classification method. In this research, the statistical trait parameters used are mean, variance, skewness, kurtosis, entropy. The test was carried out using 100 images consisting of 70 training data and 30 test data. Based on the results of classification testing with SVM, 26 data were verified correctly. Accuracy results obtained were 86.66%.