A digital image analysis method for assessment of lentil size traits

•Seed size index was accurately predicted through image analysis with R2>0.99.•Seed length and mass were predicted well by image analysis (R2=0.97 and 0.96 respectively).•Seed Size Distribution was predicted within the 10% error of sieves.•Image analysis method gave more detail and precision on size traits than standard sieving. Image analysis offers a rapid and non-destructive method for a wide range of agricultural applications including grading of grains. In this study, a method was developed for seed size grading by analysis of three-dimensional digital image information on single lentil kernels. Predicted length, single-seed mass, bulk-sample mass and Seed Size Index (SSI) were all highly correlated with the physically measured values giving Lin’s concordance test statistics of 0.98, 0.97, >0.99 and >0.99 respectively. Sieves were found to have a 10% misclassification rate and Seed Size Distribution was predicted within error of the sieving method. Results also indicated that image analysis can give much more detailed and precise descriptions of grain size and shape characteristics than can be practically achieved by manual quality assessment