9

I am looking at a hardware based platform for surveillance class of applications. I want to identify faces of people as they appear in the scene.

I don't intend to really apply face recognition (then or later). Also orientation of faces can be assumed straight for practical purposes. Face detected once needs no further processing.

Is there is a fast (efficient) algorithm to detect faces from the live video source?

Dipan Mehta
  • 5,569
  • 2
  • 28
  • 53
  • See [Simple Face-detection Algorithm Based on Minimum Facial Features](http://cat.hfu.edu.tw/~b8403009/TD-006386.pdf) and [Near realtime face detection on the iPhone w/ OpenCV port](http://www.morethantechnical.com/2009/08/09/near-realtime-face-detection-on-the-iphone-w-opencv-port-wcodevideo/). I would also try to find out what they use in digital cameras. Please share your experience when you get it running! – Emre May 18 '12 at 19:25
  • @Emre the reference you gave are quite useful. Why don'y you expand this as answer i would love to upvote that! – Dipan Mehta May 23 '12 at 12:51

2 Answers2

3

We know that the problem is solved because our cell phones and digital cameras can do it. You might be able to find out what algorithm they use by trawling the patent databases.

Computer vision libraries like OpenCV and SimpleCV offer face detection as a standard feature. See for example Near realtime face detection on the iPhone w/ OpenCV port. If you want to DIY see Simple Face-detection Algorithm Based on Minimum Facial Features.

Emre
  • 2,807
  • 15
  • 21
1

I'd suggest Viola-Jones Face detector. Easy to implement and efficient in computation time. I did some face recognition project circa 2009 and this method was the most computation efficient method back then. Original Paper: research.microsoft.com/~viola/Pubs/Detect/violaJones_IJCV.pdf
Basic Knowledge: http://en.wikipedia.org/wiki/Viola%E2%80%93Jones_object_detection_framework

Dipan Mehta
  • 5,569
  • 2
  • 28
  • 53
Hephaestus
  • 146
  • 2