Melon Surface Color and Texture Analysis for Estimation of Soluble Solids Content and Firmness

Abstract

th , 2012; Revised: August 29 th , 2012; Accepted: August 31 th , 2012 Purpose: The net rind pattern and color of melon surface are important f or a high market value of melon fruits. The development of the net and color are closely related to the changes in shape, size, and maturing. Therefore, the net and color characteristics can be used indicators for assessment of melon quality. The goal of this study was to investigate the possibility of estimating melon soluble solids content (SSC) and firmness by analyzing the net and color characteristics of fruit surface. Methods: The true color images of melon surface obtained at fruit equat or were analyzed with 18 color features and 9 texture features. The partial least squares (PLS) method was used to estimate SSC and firmness in melons using their color and texture features. Results: In sensing melon SSC, the coefficients of determination of validation (Rv 2 ) of the prediction models using the color and texture features were 0.84 (root mean square error of validation, RMSEV: 1.92 °Brix) and 0.96 (RMSEV: 0.60 °Brix), respectively. The Rv 2 values of the models for predicting melon firmness using the color and texture features were 0.64 (RMSEV: 4.62 N) and 0.79 (RMSEV: 2.99 N), respectively. Conclusions: In general, the texture features were more useful for estimating melon internal quality than the color features. However, to strengthen the usefulness of the color and texture features of melon surface for estimation of melon quality, additional experiments with more fruit samples need to be conducted.