Color Analysis of Coffee Cherry Image Using Image Processing and Machine Learning

Abstract

This research aimed to analyze and compare the color values of coffee cherries at different ripening stages, including unripe, ripe, and overripe stages, using histogram analysis and color segmentation with K-means clustering in the RGB color system. The results showed that histogram analysis highlighted the significance of the green color in the unripe stage, which decreased significantly in the ripe and overripe stages, reflecting the degradation of chlorophyll. Meanwhile, the red color increased in the ripe stage and decreased in the overripe stage, corresponding to pigment changes in the coffee cherries. The blue color exhibited minimal changes but contributed to the balance of the analysis. For color segmentation using K-means clustering, the number of clusters (k) significantly impacted the segmentation results. A value of k=3 was suitable for analyzing primary colors, while k=4 provided more detailed shades, enhancing the segmentation of different tones, such as lighter and darker greens in the unripe stage or purple tones in the ripe and overripe stages. These techniques effectively identified the ripeness levels of coffee cherries and could be applied to automated sorting systems in the coffee industry, improving sorting processes with greater accuracy.