10-10-2014 05:38 AM
P.S.: Do you have access to the raw IR image data?
10-10-2014 05:41 AM
Thanks, I will look into it!
No, I only have access to the processed data i.e. a table, where each value is a representation of the depth of the pixel in question.
10-10-2014 05:42 AM - edited 10-10-2014 05:42 AM
Also, the sensor should be calibrated, so the intrinsic calibration parameters should be located somewhere!
Best regards,
K
10-10-2014 07:39 AM - edited 10-10-2014 07:39 AM
Will the pin hole model also work on surfaces with a curvature?
For instance if I want to measure the cirkumference of a cylinder?
10-10-2014 11:28 AM - edited 10-10-2014 11:37 AM
@renepaulsen wrote:
Will the pin hole model also work on surfaces with a curvature?
For instance if I want to measure the cirkumference of a cylinder?
The camera model is simply a way to map the 3d coordinate onto a 2d image plane. If you account for the distortion, the principal point offset and focal length in both directions, accurate measurements can be performed.
Measuring the circumference (even half circumference) of a cylinder would be impossible with one static 3d camera because of the occlusions and not covering the neccessary measuring area.
Best regards,
K