0

Looking to do a proof of concept, and new to Kinect. I believe this is possible, but trying to gauge difficulty with links to tutorials etc explaining how this may work.

Looking to have the Kinect look at a walkway, and essentially detect people movement. This does not mean Skeletal movement, but essentially "foot traffic". I want to determine the noise of traffic, i.e. are there alot of people walking past, or a few. (Note this does not mean counting, just a rough indication. Can this be just pixel movement etc?)

Secondly, if a person then stops and faces the Kinect, pick them up as a user, and track rudimentary movements.

The second part I'm relatively comfortable with, the first I'm not.

Any help is appreciated is pointing me in the right direction. We are a Microsoft house, so any indication if Microsoft SDK, or OpenKinect is the best path would be great too.

mickyjtwin
  • 4,960
  • 13
  • 58
  • 77
  • I haven't used the MS SDK in a while(since 1.6), but have used OpenNI an can recommend the scene segmentation feature where you can detect a user without skeleton tracking and get the user's centre of mass so you can track movement and in your case traffic. I'm not sure if a similar feature is present in the MS SDK. OpenKinect will not provide something as advanced. With OpenKinect, you will need to use a bit of OpenCV on top for people detection – George Profenza Oct 30 '13 at 23:41
  • @GeorgeProfenza would it matter that I don't care about actual people? I only need to track one person at a time. If there is no person being tracked, then in a simple way, maybe by depth pixels, monitor the change in depth pixels?? or blocks of pixels, if they change within a defined tolerance level, class it as busy, else class it as slow. – mickyjtwin Oct 31 '13 at 00:18
  • people != persons ? :P why use a kinect in the first place - try an opencv background subtractor (like MOG2) -> track the contour centroid position and determine if it's going past the region you're interested (door/corridor/etc.) – George Profenza Oct 31 '13 at 01:14
  • the two states are: nobody interacting (no skeletal recognition) where depending on how many people are walking past (no requirement to know how many, just that there is "traffic" going past the camera; a person stops and starts interacting (kinect skeletal interaction) – mickyjtwin Oct 31 '13 at 03:37
  • another simple solution: basic frame differencing (subtract the previous frame from the current) and count the non black pixels. divide them by the total and you should have a percentage of how much motion there is in the frame (or region of interest) – George Profenza Oct 31 '13 at 10:17

0 Answers0