Open GoogleCodeExporter opened 8 years ago
Thank you for your time.
As we explained in our paper, our implementation is partial. In particular, no
refinement of the interest points detection is made. Futhermore, the descriptor
of each point doesn't consider the orientation.
That's probably why there are differences between OpenCV SURF and P-Surf in
your point 3.
The library is fully open and it can be extended by everyone. The details are
all included inside the paper.
We are sorry about the paper because it's italian only. If you need help we can
discuss about that.
P-Surf was a project made during our university studies, so at a certain point
we had to conclude.
Thank you again for your message.
Alessandro
Original comment by alessand...@gmail.com
on 4 Oct 2010 at 7:33
Thank you very much for support,
I would like to discuss about my query. I am dealing with high speed image
processing. The camera capture speed is 500fps and 1000fps my desired
computation time should be realtime 1/ 500 or 1/1000 ms.
I implemented SURF algorithm and I found I can not achieve real time processing
at high speed capturing.
Is it possible processing to achieve real time image by using P SURF? What is
the computation time of P SURF?
Please let me know about it.
Thank you,
Sushil
sushil.raut2080@gmail.com
Original comment by sushil.r...@gmail.com
on 25 Apr 2013 at 8:29
Original issue reported on code.google.com by
sachin.f...@gmail.com
on 12 May 2010 at 4:59