Suitable or efficient 'Face Recognition' method for both low and high resolution Image
3 views (last 30 days)
Show older comments
I am working on my fourth year project on Face Detection and Recognition. I am finding it difficult to find a good face recognition method for Low resolution Image as well as high resolution image. I earnestly request you to guide me to find a suitable method and if possible it's algorithm, so that I can conduct my project and submit it in time. Thank you in advance.
1 Comment
tag omda
on 14 Apr 2016
hi, i'm working on the same project now, did you achieved any progress in that project. if you did it well, could you plz help me?
Accepted Answer
Image Analyst
on 28 Aug 2014
For the method, I'd suggest you snap a bunch of pictures of people you know to form a "database" - a collection of images in one folder. Then snap a second set of pictures of those people plus some other people. Run your algorithm on those and see if it correctly identifies the people it should and correctly states "not in database" for those people who are not in the database. Then construct an ROC curve. http://en.wikipedia.org/wiki/ROC_curve
For the algorithm, see what the Computer Vision System Toolbox can do, or else go here http://www.visionbib.com/bibliography/contentspeople.html#Face%20Recognition,%20Detection,%20Tracking,%20Gesture%20Recognition,%20Fingerprints,%20Biometrics for published algorithms. If the resolution is low enough, all algorithms will fail. For example you can't identify someone if all you have is a 16 by 16 photo of their face. We don't really help with algorithms unless it's already built-in to a toolbox as a function.
2 Comments
Image Analyst
on 28 Aug 2014
You would also have to have run the database through the algorithm at the same resolution to see how it behaves with those low res images. There are some algorithms that are fairly scale independent (like SIFT). Not sure what your algorithm does.
More Answers (0)
See Also
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!