Strawberry Ripeness Identification Using Feature Extraction of RGB and K-Nearest Neighbor

Nowadays, Indonesia has not been playing an active role in fulfilling the demand for strawberries in foreign markets yet. One of the reasons is the low quality of fruit selection that still uses conventional methods. Therefore, a proper method to group strawberries automatically is considered necessary. This research aims to identify the ripeness of strawberries using RGB feature extraction and the K-Nearest Neighbor (k-NN) algorithm. The strawberry image data used in this study is divided into two, namely training data consisting of 30 images, and test data with 20 images which is classified into four categories, i.e., ripe, unripe, raw, and not strawberry. Based on the test results obtained, incorrect classification is discovered happened on the unripe strawberry images due to the tendency of the red or green dominantly but not uniformly distributed. However, the accuracy of the ripeness classification is 85% for value of k used is 7. Therefore, it can be concluded that the system is able to detect the image of the strawberry category as well as the non-strawberry category.