Ask Your Question

How to match given video stream with set of images

asked 2014-07-25 04:38:53 -0500

Punith K gravatar image

I want to match video with set of images to identify whether user is yawning or not. Can we try matching Histogram to match pattern with video input.

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2014-07-25 04:40:57 -0500

No you can't ... it will be slow ass as hell. What I suggest is the following approach

  1. Start with a VJ face detector. That reduces the search area.
  2. Using a mouth region detector would even be better, but might be hard to train.
  3. Use a classifier to define what mood is being expressed. Or apply a feature point detector to find mouth corners and define relations for moods.

Good luck!

edit flag offensive delete link more


Thanks #StevenPuttemans for your replay. I am using VJ's face detector to extract the face features. I have extracted each frame from the video and I'm thinking to match those frames with set of yawning images i have in my database.

Or Can I try with pattern matching to match each frame with database I have.

Punith K gravatar imagePunith K ( 2014-07-25 04:50:39 -0500 )edit

Why match images. Create an average feature vector of yawning images. Than apply the same transformation on the new face and do a neighbor matching. The difficult part is creating a meaningfull descriptor. Other way could be to create binary SVM classifiers for each expression. The matching will be way to slow as I see it.

StevenPuttemans gravatar imageStevenPuttemans ( 2014-07-25 04:54:09 -0500 )edit

okay. Do you know any good site to know about SVM classifier creation, I searched almost the entire internet. Please do mention some of the sites useful for beginners.

Punith K gravatar imagePunith K ( 2014-07-25 05:00:37 -0500 )edit

Hmm you just hit a pittfall of OpenCV. The CvSVM module is highly outdated and badly documented. People always suggest using LibSVM to apply the SVM modeling. Check this code block I made for trying to create an SVM model using the OpenCV interface. However it requires a mapping from the HOG descriptor interface to a format that can be used by the CvSVM interface. It produces a model but I have not had the time to perform exhaustive testing. Ignore the fact that I was training a cookie detector ;) This was just a dummy case.

StevenPuttemans gravatar imageStevenPuttemans ( 2014-07-25 05:07:38 -0500 )edit

Thanks man :)

Punith K gravatar imagePunith K ( 2014-07-25 05:16:33 -0500 )edit

Question Tools


Asked: 2014-07-25 04:38:53 -0500

Seen: 194 times

Last updated: Jul 25 '14