Identification of Tomato Ripeness Using RGB Color Image Analysis
Firda Fadri (a*), Kiswara Agung Santoso (a)

a) Jember University
Jalan Kalimantan No.37 Kampus Bumi Tegalboto, Jember 68121 Indonesia
*firdafadri[at]unej.ac.id

Abstract

The identification of tomato fruit ripeness is performed to assess fruit quality. The identification process relies on fruit images using image processing techniques. The research data is leveled into four ripeness levels: unripe, half-ripe, ripe t, and rotten tomatoes. The RGB color space is utilized to classify the test image data. Correlation coefficient values and MSE values are obtained from feature extraction. The highest correlation coefficient for each test data indicates the fruit category, aligning with the classification results as expected. The classification of the test image data with the training image data achieves an accuracy rate of 85%. The average MSE values for each ripeness category are exceptionally small (approaching zero), indicating minimal differences between the test image data and the training image data.

Keywords: Tomato- Image Processing- RGB

Topic: Mathematics and Statistics