Identifikasi Kematangan Cabai Rawit Menggunakan Metode Momen Warna Dan SVM

Abstract

This study focuses on the identification of the ripeness of bird's eye chili (Capsicum annuum L.) using the color moment method and Support vector machine (SVM). Bird's eye chili holds significant economic value in the agricultural industry, but the rapid and accurate identification of ripeness remains a challenge. The color moment method is employed to extract features from images of bird's eye chili, while SVM is used for classification based on the extracted features. This research expands the use of data and optimizes SVM parameters to enhance the accuracy of identification. The results indicate that the classification model using the SVM method successfully classifies the ripeness levels of chili with an accuracy of 82%, categorized as good. Specifically, it achieved a precision of 66.67%, recall of 83.33%, and F-measure of 74.07% for the "ripe" category. The "unripe" category achieved perfect precision and recall, both at 100%, with an F-measure of 100%. Meanwhile, the "half-ripe" category achieved a precision of 80.00%, recall of 77.42%, and F-measure of 78.73%. However, some misclassifications were attributed to color dominance issues in certain images. Therefore, this study contributes to the development of more accurate and efficient techniques for chili ripeness identification. Recommendations for future research involve further evaluation with a broader test dataset, improvements in classification methods, and the consideration of additional features to enhance classification accuracy. This research offers both scientific and practical benefits in image processing, machine learning, and the agricultural industry