The results achieved show the capability of this technique for
solving the problem of having an accurate and objective tool for
measuring cluster compactness. In addition, the fact that the berry
size quality component was estimated automatically without a
controlled background made this technique very feasible for use
under field conditions. It would be very useful, for example, to
assess grapevine quality components in inter varietal studies (such
as genetic association studies).
Grape clusters were measured in a difficult image acquisition
scenario with the aim of simulating field conditions. Future work
will focus on extrapolating these results into the field scenario.