Repeat Problem 12.6, but use a Bayes classifier (assume Gaussian densities). Indicate clea...

Repeat Problem 12.6, but use a Bayes classifier (assume Gaussian densities). Indicate clearly how you would go about obtaining training samples and how you would use these samples to train the classifier.

12.6 An experiment produces binary images of blobs that are nearly elliptical in shape (see the following figure). The blobs are of three sizes, with the average values of the principal axes of the ellipses being (1.3,0.7), (1.0,0.5), and (0.75,0.25).The dimensions of these axes vary ±10% about their average values. Develop an image processing system capable of rejecting incomplete or overlapping ellipses and then classifying the remaining single ellipses into one of the three size classes given. Show your solution in block diagram form, giving specific details regarding the operation of each block. Solve the classification problem using a minimum distance classifier, indicating clearly how you would go about obtaining training samples and how you would use these samples to train the classifier.