Up todate code that works pretty well
The code: Download
So far the code displays pictures that have been filtered by the blob and then the contourProcessor applied. WIth the model overlayed in red, which is what the model percieves to be the person.(Hoi correct if im wrong)
currently all it does it display this and does not use it in a useful way, some means of calculating what direction the person is heading in needs to be created. To give the robot as well as means of grabbing the images into CImg format.
but other than that we are getting there.
Also talking to Matt in CS, said that the lists in cpp are very inefficient and it might be worth converting them back to vectors if performance is too slow.
Anthony
19 Nov 2006 19:29
| 
Comments (1)
| 
Report a problem
Loading…
Found a bug in contour processor – the shiftmodel method doesn’t work because you’re copying the pixel object and then changing it. Fixed it, I’ll post it soon.
20 Nov 2006, 02:24
Add a comment
You are not allowed to comment on this entry as it has restricted commenting permissions.