In the food industry, determining fruit maturity is very important to obtain
fruit of good quality. The ripeness of some fruits can be determined by the color of the
skin. Like several other types of fruit, the skin color can be used to determine the ripeness
of Ambarella fruit. However, determining the ripeness of Ambarella fruit is still done
manually by human labor, which is considered time-consuming, tiring, requires a lot of
workers, and can cause inconsistencies. The development of technology such as computer
vision allows the determination of fruit ripeness to be carried out automatically, accurately,
and relatively quickly. This study aims to classify the ripeness of the Ambarella fruit
based on its color feature with the use of machine learning. The color features used are
RGB, HSV, and L*a*b*, with Support Vector Machine (SVM) and K-Nearest Neighbor
(KNN) as the classi�ers. Segmentation was done using the Otsu method. Google Colab,
OpenCV, and scikit-learn are utilized in carrying out the experiments. The performance
result shows that the highest accuracy, precision, recall, and f-measure were achieved by
using SVM on the L*a*b* color feature.