Plumpness Recognition and Quantification of Rapeseeds using Computer Vision

The plumpness is an important index of crop seed. However, traditional measurements are time-consuming and labor intensive. The computer vision technology, which may offer more efficient and non-destructive methods for measurement, has recently appeared. But it is very difficult to accurately estimate the plumpness of single seed by the ratio between area and perimeter because of the diversity of rapeseed seed’s size. This paper focused on rapeseed seed plumpness recognition and quantification, based on computer vision. A new method, the coefficient of variation of radius (CVR), was used to estimate seed plumpness. The recognition and quantification model for plumpness in single seed were established by using the fuzzy C-means (FCM) clustering and fuzzy math method. The plumpness of the seed is full if plumpness is greater than or equal to 0.6. Some correlative index are calculated and analyzed to verify the validity of this method. The tests show that there is no correlation between plumpness or plumpness ratio, and 1000-seed weight or equivalence diameter. But there are significantly partial correlation between plumpness or plumpness ratio, 1000-seed weight and equivalence diameter. Finally, plumpness ratio index is significantly different among the 12 varieties rapeseed was determined. With the mean value of plumpness ratio of rapeseed variety, the plumpness degree was plotted 10 grades. The results show that the application of computer vision technology is significantly valid for quantitative determination of plumpness in rapeseed seed.