Chokanan Mango Sweetness Determination Using HSB Color Space

Abstract

The objective of this research is to determine Chokanan mango sweetness by using color properties. Keyence machine vision system was used to capture mango images in HSB color space. Meanwhile, Digital AR2008 Abbe refract meter was used to obtain the value of sweetness. The process was divided into two major stages, the training stage and the testing stage. One hundred and eighty Chokanan mango images were used in this research. Fifty percent of the images were used during training stage and the other 50% were used during testing stage. In the training stage, the hue, saturation and brightness bands were analysed by using linear regression analysis. Based on the result, it has been shown that hue gave the highest value of correlation (-0.92). It also gave the lowest value of standard deviation when compared to the other bands. Therefore, it has been used as a model to determine the sweetness of Chokanan mango images. The results showed that the developed model can determine sweetness at Index 1 and Index 2 with 100% of accuracy and Index 3 with 87% of accuracy. In average, it can be used to determine the sweetness of Chokanan mango with 95.67% of accuracy.