Image processing for maturity classification of tomato using otsu and manhattan distance methods

Currently, image processing-based systems have been widely applied in various fields, one of which is agriculture. The system can be used to classify fruit maturity. Tomato the agricultural products consumed by community. Therefore, requirement for ripe tomatoes increases. In this work, classification method based on processing grading maturity level tomato was developed distinguish into three classes: unripe, half-ripe, and ripe. Classification carried out skin color tomato. required five main processes; initially, detection region interest (ROI) using Otsu followed conversion RGB HSV space. Afterward, segmentation with thresholding S channel space implemented. Subsequently, extraction mean, median, max, min features each from YIQ space; therefore, a total number 12 produced. Finally, K-nearest neighbor (KNN) Manhattan distance values k = 1, 3, 5, 7, 9. dataset consists 90 images (30 raws, 30 half-ripes, ripes), where divided two types, including 54 as training data 36 testing data. evaluation results were able achieve highest accuracy value 0.9722.