Topic: how to generate 3D surface from depth??

The image analyzer can generate 3D surface if we input the image and its depth image. Also, I can save the surface as .ply files. In the .ply files, we can see the generated 3D position(X,Y,Z).

But as we know, Xi(xi; d) = depth(i,j)*(K*R)^(-1)*x(i,j)+C, C=-transpose(R)*T.  K,R,T is the camera's internal and external parameter.(see Mutiple view geometry in computer vision, Hartley and Zisserman)

In image analyzer,we don't provide the camera's parameters. How did it generate the 3D position?  Can you give me some related formulas in details? Or if you use a assumed camera's parameter, can you tell me them accurately?I need them to evaluate my results.

Thank you very much.


Re: how to generate 3D surface from depth??

It is very simple - the depth map pixel values multiplied by the depth map scale factor are simply used as Z value and X,Y is the pixel location. This corresponds to a pinhole camera at infinite distance I would think (i.e. no perspective).

Michael Vinther

> software developer <