KLASIFIKASI BIJI PINANG BELAH PADA PENGEMBANGAN MESIN SORTIR PINANG MENGGUNAKAN PENGOLAHAN CITRA DIGITAL

Classification split betel nut is an important step in determining the quality and economic value. So far, Indonesia betel nut exporter perform this stage is visually based on color by human power. This method has several disadvantages, among others: the accuracy is low, inconsistent, and low capacity. Therefore, the digital image processing split betel nut used as an alternative to classify. This study aimed to analyze the results of digital image processing samples split betel nut. The samples used predefined quality is good and bad. Image processing begins with the image capture by the camera, then the binary image is formed with a threshold technique. Furthermore, the image is separated from the background. Then calculate the number of pixels and the calculated value of the various components of the RGB color model based on value. Data analysis is done on the RGB color model, RGB index, and Y'UV. Value color model components compared with each other to be able to distinguish the quality betel nut sides. Basic classification of quality nut that can be used is the color of R, G, B index, Y 'and U. The test results show that the test sample is the greatest accuracy on the U value Y'UV color model that is worth 96%.