What I am looking to do is take a video input, divide it in to a grid, say 8 x 8 (but i would like the option to change the resolution).
Then I want to detect movement within the cells, so i intend to convert the image to black and white and do some frame differencing to hopefully get a 'bang' from each cell that senses movement.
I have tired a scissors and glue technique which does work but is a little clumsy and hammers my processor!
I also have had a look at cyclops but Im struggling to get it to work?!
I just wondered if any one has and advice for me or can suggest an object (maybe CV.JIT....) so I don't spend to much time going down dead ends!