Seed morphometric parameters (seed area, length, diameter, circumference, roundness and aspect ratio) of 15 accessions were analysed. Measurements on seeds were taken with help of a Computer-Camera system. Variation among accessions and genetic relationships were estimated. Great variability was observed for all parameters studied. High positive or negative correlation (r=0.726-0.963) was showed by Pearson's correlation analysis for the seed metric data of the 15 accessions. The cluster analysis revealed that the 15 accessions could be grouped into three to five clusters according to the methods (Single or Complete Linkage) used. The principal components analysis showed that the first two components accounted for 99.3% of the variability. Morphometric characterization of seeds is rapid, reproducible, reliable and cost effective. This technique of digital imaging is strongly recommended for researchers from developing country like Benin, because of its efficiency.