Question

I am stuck in a computer vision problem. My problem is:

Suppose I've a point in pixel coordinate system which is (u,v) located somewhere on a portion of the screen(monitor).

I want to convert this point to camera coordinate system(from the point of view of a camera origin at (0,0,0)).

How do I convert screen coordinate to camera coordinate?

Was it helpful?

Solution

The fact is that you'll never be able to convert a pixel (2 coordinates) to a 3D space. Because the information is not there!

The good news is that there are some ways to extract 3D info from 2D images.

You need to calibrate your camera, then detect an object with known configuration (let's say a chessboard), and then solve for its position in the 3D world.

Check out these links.

http://computer-vision-talks.com/2011/11/pose-estimation-problem/

cv::solvePnP

http://www.morethantechnical.com/2010/11/10/20-lines-ar-in-opencv-wcode/

Licensed under: CC-BY-SA with attribution
Not affiliated with StackOverflow
scroll top