Klasifikasi Kematangan Buah Manggis dengan Algoritma Support Vector Machine (SVM)

Abstract

This research developed an automatic mangosteen fruit maturity classification system utilizing image processing techniques and machine learning algorithms. The proposed system employed the Support Vector Machine (SVM) classifier with feature extraction based on the Hue, Saturation, and Value (HSV) color space from mangosteen fruit images. A dataset consisting of 140 mangosteen fruit images, with 70 ripe and 70 unripe samples, was constructed. Preprocessing steps, including cropping and resizing, were applied to standardize the image dimensions. The RGB color images were converted to the HSV color, and the mean values of Hue, Saturation, and Value were extracted as features for classification. The SVM algorithm with a linear kernel was trained using these features to discriminate between ripe and unripe mangosteen fruits. Evaluation using a confusion matrix demonstrated the system's high classification accuracy of 96%, with satisfactory precision, and recall for both classes. The proposed system exhibits potential for application in the agricultural industry, enabling automated quality assessment, post-harvest management, and maximizing the commercial value of mangosteen fruits. This technology can assist producers in rapidly and accurately classifying mangosteen fruits.

Keywords: Image Processing, HSV, SVM, Machine learning, Buah manggis