Project 18: Character classification
Input: an easily processable representation of character images. Each image is labelled with it's ASCII code.
Operation: Manually separate the set of characters into a 'Training set' and a 'Test set'. Create an algorithm to learn size invariant features of characters in the Training set, and check your results with the Test set. (Hint: use a simple matrix-matching feature set with a Nearest Neighbour classifier.)
Output: Recognition accuracy statistics.
Coding: C/C++ (a console application, no GUI)