Color image database construction for the strawberry breeding

Color and shape of the strawberry fruit are the important indices which evaluate the quality of commodities and variety. The objectives of this study are to construct the strawberry breeding image database which can preserve the surface fruit color, texture, shape, and the color distribution of the cross section of the strawberry. As an experimental sample, the strawberries of 34 cultivars were used. The chromatic image of strawberry surface and cross section, which acquired using experimental image acquisition system, seems to be sufficiently good, when it is considered that purpose of the image acquisition in this study is the construction of the breeding database. In this study, as the Content-based Image Retrieval (CBIR) technique, color image analysis using Color Histogram and Color Distribution Entropy (CDE) was applied. In conclusion, the image analysis method in this study seems to be effective for evaluation of strawberry appearance, since the experimental result was almost equal to the evaluation by visual observation.