Penerapan image processing untuk mengetahui tingkat kematangan buah kopi menggunakan algoritma k-nearest neighbor (knn)

Abstract

Indonesia, as one of the world’s largest coffee producers, requires a more objective method for selecting coffee cherries, since manual assessment is still subjective and inefficient. This study develops an automatic classification system for coffee ripeness using image processing with the HSV color model and the KNearest Neighbor (KNN) algorithm. A total of 300 coffee cherry images (unripe, half-ripe, ripe) were processed through HSV feature extraction, and classification with KNN (K=3) under three data split scenarios: 90:10, 80:20, and 70:30. Evaluation metrics included accuracy, precision, recall, and F1-score. The best performance was achieved at the 90:10 split with 90% accuracy, while 80:20 and 70:30 produced 83.33% and 81.11%. The half-ripe class was classified most consistently, while errors occurred more frequently in the ripe class. This study demonstrates that combining HSV and KNN is effective for classifying coffee ripeness and has potential to replace manual methods to improve harvest quality.