Ektraksi HSV Dan GLCM Dalam Metode K-Nn Untuk Klasifikasi Tingkat Kematangan Buah Mengkudu

Abstract

Noni fruit is a fruit that is quite famous for its properties as a medicine that can cure diseases. The level of maturity of the noni fruit also affects the effectiveness of the fruit as a medicine. With the current technological developments, research on the classification of noni fruit maturity was carried out. It is intended that the selection of noni fruit can produce more effective results. K-Nearest Neighbor can be implemented to measure the maturity level of noni fruit. In this study, Mean HSV is also used as a color extractor and GLCM as a feature extractor. The dataset is 100 image data, with 80 images divided into training data and 20 test data. Then, the level of maturity is divided into four classes, namely: raw, undercooked, ripe, and rotten. Before the classification process is carried out, the previous image data is pre-processed. The first stage is to change the color of the image to grayscale which is then converted to an HSV image. The second stage is the extraction process using the GLCM method in order to produce the values of Energy, Contrast, Correlation, and Homogeneity. The third stage, followed by the feature extraction process using Mean HSV. After these processes, it is continued with the classification stage using the K-NN method. This study produces an accuracy value of 95% using the value of K = 5, pixel distance = 1, 4, and 8 in the test using a combination of the KNN, GLCM, and HSV algorithms.

Klasifikasi Tingkat Kematangan Buah Tomat menggunakan Computer Vision untuk Smart Agriculture

“…Terdapat banyak penelitian sebelumnya yang melakukan klasifikasi tingkat kematangan buah secara otomatis menggunakan objek dan metode yang berbeda dengan memanfaatkan teknologi Computer Vision dan kecerdasan buatan sehingga meminimalisir kesalahan proses pemilahan kematangan buah ketika dilakukan secara konvensional [3][4] [5][6] [7]. Salah satu algoritma yang dapat melakukan klasifikasi dengan baik dan sudah digunakan beberapa peneliti adalah algoritma Support Vector Machine.…”