The classification of dragon fruit maturity directly by farmers has weaknesses because it is influenced by the subjective factors of these farmers such as fatigue and other physical disorders. This causes the farmers' performance to be not optimal and less through so that the maturity classification becomes inconsistent. For this reason, this study aims to automatically classify dragon fruit maturity becomes more effective and efficient because it is carried out with the same and consistent standards even though in large quantities by utilizing image processing. Dragon fruit maturity level will be classified into 3 classes based on the color features HSV, ripe, half-ripe and unripe. The method used to carry out the classification is the Multi-Class SVM. The data used in this study were 105 data consisting of 90 training data and 15 test data. The classification results of dragon fruit maturity using the Multi-Class SVM is 86.67%.
                        
                        
                        
                        
                            
                                Copyrights © 2023