Seed Type Identification by Image Analysis - Correlation of Nutrients with Size, Shape and Colour Characteristics of Seeds

Page: 138

I. Smýkalováa, J. Horáčeka, M. Hýblc, M. Pavelekb, M. Bjelkováb, T. Krulikovskáe, and D. Hampeld

a Department of Biotechnology, b Department of Technical Crops, c Department of Grain Legumes, Agritec Plant Research, d National Plant Variety Office, Brno, e Department of Fermentation Chemistry and Bioengineering, Institute of Chemical Technology, Prague

 

Digital image analysis (DIA) has become a powerful technique for characterization and grading of plants in crops. DIA is utilized in determination of cultivar origin, detection of contaminants and in discrimination of new cultivars for registration purposes. Seed images, their digitization, segmentation and evaluation using special software enable rapid sorting of genotypes based on the measured size, shape and colour characteristics of seeds. In the study DIA was used for measurement of characteristics of small flax and large pea seeds followed by seed classification. The size and shape of seeds were evaluated from scanner images due to colour instability in scanning. Colour deviation calculated from the coordinates in colour space was used for colour classification of seeds. Some of the DIA data were correlated with fat nutrients in flax seeds.

 

Full text (PDF)