Abstract
In chickpeas, the seed size is a critical phenotype that needs to be evaluated carefully during variety development. The sieve analysis used for determining seed size distribution in legumes is labor intensive and time consuming method. An image-based method for sizing chickpeas seeds was developed in this study. Samples from a total of 72 plots from two different locations were harvested and seed size was analyzed. The results show that seed size calculated from image-based method was highly correlated to the ground-truth data, with a correlation coefficient of 0.90. The image processing technique provides rapid evaluation of seed size for phenotyping chickpeas and the method also can be adapted for similar seed types.
