On Wed, May 1, 2013 at 7:45 PM, Stéfan van der Walt <stefan(a)sun.ac.za>wrote:
> On Thu, May 2, 2013 at 2:38 AM, Ankit Agrawal <aaaagrawal(a)gmail.com>
> > Can you point me to that code? Thanks.
> I see Rosten himself now has NumPy code available:
I'm not sure I'd call that NumPy code. I think the code is auto generated.
My guess is that it would be simpler to start fresh.
I have just made a pull request with wu's circle generation (one of the
It has nothing to do with the proposed project, but now I can build skimage
and understand how to commit my changes.
Circle images are attached to this post.
вторник, 30 апреля 2013 г., 4:33:48 UTC+4 пользователь Stefan van der Walt
> Hi Fedor
> > My name is Fedor Morozov and I am a 2nd year computer science student in
> > Moscow State University.
> Thanks for getting in touch!
> > What do you think about such project? I guess, it's rather ambitious,
> yet it
> > is diverse, interesting
> > and will keep me from slacking. If some of the mentioned ideas get
> > projects, they can be
> > adopted during summer.
> Nathan Faggian has done some work towards registration in pyimreg, and
> what we discovered is that it is fairly hard to come up with a good
> API. Perhaps you can have a look at that code and give us your
> feedback? Registration would be a very useful feature to have in
> Perhaps we could start with something simple: compare Harris features
> with a robust matching algorithm such as RANSAC to find alignment with
> a simple error measure (there is a PR for RANSAC by @ahojnnes). Once
> aligned, perform Laplacian or other blending. If that works, one can
> focus on adding other feature detectors, multi-resolution matching,
> different error measures, optimization based approaches, etc. There
> may be some code in https://github.com/stefanv/supreme that we can
> re-use, or we can simply start from scratch.
> Eventually, I am also interested in non-rigid registration,
> registration of voxel data, etc.
> Registration / blending is an immense topic that can very easily cover
> an entire GSoC.
I have made the first draft of my proposal<https://github.com/scikit-image/scikit-image/wiki/GSoC-2013-Ankit-Agrawal-I…>and uploaded it on our github wiki. I would like all the mentors and
community members to review it and comment for any possible improvements
I have some of personal comments on it :
1] I feel that implementing all the four features is slightly ambitious and
implementing three of them would be the best. After skimming through
papers, this is the order of difficulty in terms of implementation :
BRIEF < STAR < FREAK < BRISK(most difficult)
2] As mentioned in the BRISK paper<http://www.robots.ox.ac.uk/~vgg/rg/papers/brisk.pdf>,
BRISK detector makes use of FAST detector<http://web.eecs.umich.edu/~silvio/teaching/EECS598_2010/slides/11_16_Hao.pdf>in the process. If it comes down to implementing 3 instead of 4 features,
it would be best that I implement FAST as the 4th feature instead of BRISK,
which would relax the timeline a bit and will give me more time for testing
I suggest all members to review and share their opinions on it as
soon as possible so that I can make the changes and upload it on
gsoc-melange site. Thank you for your time.
Communication and Signal Processing,