Classification of Total Carotene and Quality of Chili Pepper (Capsicum frutescens) Based on Image Analysis Yusuf Hendrawan (a*), Rizki Utami (a), Dimastyaji Yusron Nurseta (a), Daisy (b), Sri Nuryani (a), Dewi Maya Maharani (a), Sandra
a) Department of Agricultural Engineering, Universitas Brawijaya, Jl. Veteran, Malang, ZIP 65145, Indonesia
*yusuf_h[at]ub.ac.id
b)Department of Agricultural Product Technology, Universitas Brawijaya, Jl. Veteran, Malang, ZIP 65145, Indonesia
Abstract
The process of sorting the chili pepper (Capsicum frutescens) is done physically, based on the visual color uniformity of the skin of the human eye. This method is not effective and efficient. This study aims to identify the total content of carotene in chili pepper using the color and textural features approach. Color feature extraction used is the value of RGB, HSV, HSL, XYZ, CMY, CMYK, Lab, LUV, LCH, grey color, and 10 textural features from each color space. The feature extraction results were used to identify the total carotene content by the image analysis method. The image of chili peppers used was 360, consisting of 300 training data and 60 test data. Classification test results with a level of 20% produce the best parameters as an indicator of total carotene chili pepper i.e. hue mean features with a range for quality A (55.04> hue mean > 32.15), quality B (19.80> hue mean> 12.21), quality C (3.55> hue mean> 1.93), and 80% accuracy using the confusion matrix and mean square error (MSE).
Keywords: chili pepper; image analysis; total carotene; quality
