A Color Level Evaluation of Roasted Coffee Beans using Histogram for Image Processing to Grade Separation

Abstract

This research aimed to find out about development and improvement intensity measurements of coffee beans to separate roasted coffee quality. In the analysis process, the histogram was carried out to look at pixels of coffee bean images to determine coffee bean's level of color intensity. Experts looked at coffee beans in color grading in the traditional production way, which took too much time. Therefore, the color intensity of coffee beans system was developed. Colors of roasted coffee beans were compared to determine the intensity that matched the standard color values in light, medium,and dark levels. Lights would appear when the beans matched the colors.

The histogram results showed the color intensity of coffee beans at various levels of the coffee beans.  The mean of the RGB color standard was H. The light color level of the coffee beans was H = 76.513, the medium color level was H = 57.966, and the dark level was H = 41.420.  The color intensity of roasted beans became lighter subsequentially during the roasting process. When coffee beans had stand colors and the process was finished, the signals would come on.  The error system has been determined error value in light color level was 0.502, the error medium color level was 0.344, and the dark level of error value was 0.043