Classification of Apple Ripeness Detection System Using Self-Organizing Map (SOM) Method

Apple (Malus Domestica) is one of the most popular types of fruit and is in high demand by the public because of its varied flavors. Apples have many nutrients and various vitamins including healthy fats, carbohydrates, proteins, vitamins and many more. The Apple is one of the apple varieties developed in Batu City, Malang and planted in several areas with suitable agroclimates for apple growth. This research uses Anna apple images as datasets. Various ways can be employed to distinguish Anna apples' maturity, including through color image analysis. But to the naked eye, Anna apples are often difficult to distinguish. This research classifies the maturity of Anna apples based on color analysis with the Self-Organizing Map method. Using Google Colab and Python programming language and datasets from kaggle.com as many as 139 datasets, 46% training data, 54% validation data. The Self-Organizing Map method was chosen because of its ability to recognize visual patterns accurately. The accuracy of the results based on the SOM Method performance evaluation metrics namely Quantization Error, Silhouette Score and Topographic Error. Quantization Error RGB (0.004737) is lower than HSV (0.073178) which indicates RGB's ability is effective in representing data in SOM. Silhouette Score HSV (0.704204) is higher than RGB (0.599846) indicating the ability of HSV is slightly better in grouping objects.