Previous Entry Share Next Entry
Well, that was okay...
I'm finished on the actual convolutions already. It was pretty simple - to identify the cerebral cortex thing, you just use thresholding to remove all but the brightest parts of the brain, and then use some simple techniques to eliminate all areas smaller than, say, 700 pixels in total. The skull was just a matter of using edge detection, and the third task is the entire brain, which was again a process of thresholding and area-finding.

Now to write the report on it... Ho hum... :o)


Log in

No account? Create an account