We are currently doing out thesis work about OCR or optical character recognition. Just want to ask something about diving a single image into multiple images or templates.
What we are trying to achieve is that we want to segment or divide each characters (total of 16) from a single image into multiple (16) image, for further recognition. Below is the image of the proposed segmented characters.
The solution we thought was to find contours or outlines of each characters. From these contours, we would draw a square boundary around the characters to distinguish them from the background. This square boundaries will then be the basis of division into multiple images.
The question is how can we do that in terms of coding? Is that solution possible?
