Ask Your Question

Human height estimation using a calibrated camera

asked 2013-12-05 07:28:14 -0600

Ruckus gravatar image

updated 2013-12-05 08:11:06 -0600

Hello OpenCVers!

I am working on an algorithm to estimate the height of detected people in a video, and I'm stuck.

The part that I have working is the detection of people using the HoG algorithm, so I have a bounding box for every person in the frame. And I have calibrated the camera, so I have my intrinsic and extrinsic camera parameters.

The problem is that now I have a formula for the perspective projection with 2 unknowns: height of the object and the distance from the object to the camera. I am using one mono web camera to detect people so I have no information about the distance from the object to the camera. And the height is what I'm trying to estimate, so I don't have that as well.

I know this problem is solvable if I use a kinect or a stereo camera in order to get the distance, but I'm limited to only one mono web camera.

Does anyone have an idea on how to approach this problem? I have read about using reference objects but I can't figure out how to use them to help my problem.

edit retag flag offensive close merge delete

2 answers

Sort by ยป oldest newest most voted

answered 2013-12-05 08:12:06 -0600

Case1 gravatar image

Hi there,

start off with the article "Single-View Metrology: Algorithms and Applications" from Criminsi. You can find it here The following article seems to implement a possible algorithm: (I did not read it).

This should give you a good starting point to find other articles. If you do not have access to a (digital) library, I recommend using which typically gives me nice results (although not comparable to a professional library).

Hope that helps. Case1

edit flag offensive delete link more

answered 2013-12-05 08:16:11 -0600

bpp gravatar image

If the person to be measured has for example in hand or on the t-shirt a marker of known size it is possible to obtain approximately the height. An other option if you don't move the camera could be to put same marker on the floor at a known distance and say to the people to go with their shoes on those marker. Remember that if you don't know the distance between the camera and the object to be measured there's is an ambiguity problem, passing from 3d world to 2d coordinates you lose distances.

edit flag offensive delete link more

Question Tools


Asked: 2013-12-05 07:28:14 -0600

Seen: 3,434 times

Last updated: Dec 05 '13