How do I derive a perspective transformation matrix??

How do I derive a perspective transformation matrix??

Post by Thomas D » Fri, 26 Apr 1991 01:37:55



I have a camera mounted at some [x,y,z] looking at a figure on the
z=0 plane.  Is there a way to derive the perspective transformation
matrix without knowing any of the angles that the camera is mounted at?

I can put a square (or anything else) of known dimensions on the z=0
plane to use for this calibration.

Any hints would be welcome.  Everything that I have ever studied shows
how to create a perspective view but I have not been able to figure
out how to "reverse engineer" the problem.

Thanks a lot,

Tom....


 
 
 

How do I derive a perspective transformation matrix??

Post by Charles D. Hans » Sun, 28 Apr 1991 06:24:31


Tuscani and Faugeras had a paper on camera calibration which gave you
the perprective transform, as well optical centers etc.  Excellent
paper and the techniques work!

Faugeras and Toscani. "The Calibration Problem for Stereo", CVPR '86
Miami Beach, pp:15-20 IEEE, 1986

Chuck....


 
 
 

1. deriving values from transformation matrix

Hi,

It's a long shot but I thought I'd ask anyway. Does anyone know a method
to derive specific rotation, scaling, and positions, from a 16 float
transformation matrix string? I'm attempting to build a transformation
interface dialog for shader manipulation and can build a 16 float string
from a matrix, but can't go back the other way to get individual axis
values.

Any help appreciated!

Brian Perry

2. OpenGL acceleration for Linux?

3. How to derive rotations from a transformation matrix

4. Nikon Super Coolscan 2000 -is it any good?

5. Perspective Transformation Matrices

6. MaskBlt, what's wrong here ?

7. Matrix for Perspective Transformation

8. Mimic Oil Painting Brush Strokes?

9. how to rebuild the matrix transformation perspective

10. Perspective transformation matrix ?

11. Looking for instructions to derive a projection matrix from scratch

12. Deriving World Coords from Transform Matrices

13. Deriving Viewing Matrix