Partial least square
discriminant analysis (PLS-DA) was applied to a total of 48 image features calculated from 5 MR images of
144 processing tomatoes to predict the tomato maturity. The model with 4 latent variables captures 70%
of the variation in tomato maturity. The classification accuracies of the PLS-DA model were around 90%
for green, breaker-light red, and red maturity stages. The Variable Importance in Projection coefficient
of the 48 image features in the model indicated that the diffusion weighted image and spin echo image
with higher T2 weighting were most important for tomato maturity classification