Establishment of maturity index level of ‘Pisang Berangan’ based on extracted color features at 3 different maturity stages

Abstract

Banana is one of the food sources which contain a lot of vitamins that are beneficial to the human body. It is a popular tropical fruit that has high market value all over the world. Currently, banana is sorted manually by the fruit sellers or suppliers from color inspection of the peel color. This process is relatively slow and time-consuming, higher cost and labor demanding. The major purpose of this study is to establish the maturity index level of local banana variety, Pisang Berangan based on the extracted color features at 3 different maturity stages by using vision machine system. In this study, 30 banana samples were used in image acquisition. The images were grabbed in 2 different views: top and side views daily using a CCD camera set up in a black box. All of the acquired images were processed using several image processing procedure to enhance the image quality and separate sample image with the background image. Image processing procedure such as IMAQ Threshold was carried out in the LabVIEW programming software. Color features data such as RGB, L*a*b* and HSV were extracted based on the acquired image. All of the extracted data were analyzed using Qualitative Data Analysis (QDA). From the analysis, a maturity index level chart of Pisang Berangan is produced to assist the sellers and suppliers in classifying the banana according to the corresponding ripening stages for a better postharvest management procedure.