Characterization of sesame varieties through image analysis

Abstract

The identification and discrimination of crop varieties is very crucial for their breeding, variety registration, seed production and certification. Oligogenic morphological characters of seeds are distinct as well as stable and hence can be used for identification of varieties. Digital image analysis is an alternative to the manual classification of biological seed by integrating an image acquisition device and a computer. Data collected on sesame seed characters using grain scanner has significant differences for every observation. The sesame variety, SVPR 1 recorded the highest seed surface area (3.92mm2), perimeter (7.88mm) and length (3.32mm) when compared to all other varieties. Cluster analysis revealed that the varieties could be grouped into two major clusters in which CO 1, TMV 3, TMV 4, TMV 5, TMV 7 formed one cluster whereas the other varieties were grouped under another cluster, which showed that the genotypes in one cluster had similarity in most of the parameters and also its parentage. Thus, image analysis helps in discriminating the morphological variation in seeds related to genotype and its evolution.