イチゴの品質評価に関する画像処理の研究(第1報)カラー画像処理による果実表層の打ち傷の検出

概要

論文の詳細を見る
This paper presents a study on quality estimation of strawberry fruit using color image processing. A CCD color camera was used to capture the images of the fruits placed in an image-capturing box. Bruises on the fruit were detected by analyzing color models, namely RGB and L*a*b*. Color model analysis was able to discriminate between bruised and non-bruised surfaces. However variation in RGB components alone was not sufficient to detect bruises. The L*a*b* model was found to be stable and effective. Specifically, a* level was able to show the characteristics of the bruises. Thus, a* level of the L*a*b* model could be used to represent the bruises on strawberry surfaces. The storage of strawberries at a low temperature of 2°C and 90% relative humidity was found to inhibit further engraving of bruised areas as compared to those stored at normal room condition of 18°C and 65% relative humidity. Hence, the freshness of fruit was preserved, when they were stored under the above controlled low temperature condition.
日本生物環境工学会の論文