Camera calibration - intrinsic parameters

[renepaulsen]

I need to be able to measure the area of a 3D surface. 

Currently I am using a scalingfactor, where the variable, dn is the output from the 3D camera (where dn represents any value in a 2D array, of 160 by 120 - matching the resolution of the ToF camera). 

Then by multiplying the average scalingfactor to the number of pixels detected, it is possible to measure the area fairley accurate. 
However, I wish to be able to mark the ROI on a 2D image and map to the 2D array, but there seems to be an offset, since the object (in this case a rectangular plate) is not in the same spot. 
When manually selecting the values in the .xls document the area is measured fairley (~7% error), but the error is a lot bigger when selecting the plate in the RGB image (probably due to the offset). 

So I was wondering if a calibration would help? Using intrinsic parameters to account for the distortion. Could that also help me compensating for the offset?

I have attached some vi's. Both should be downloaded - then run the "load info" vi. Select (filepath) for the .txt file, and a filepath for the .jpeg file, run vi and press okay. Now Select area to measure, and press okay again. 
The .xls file can be opened and used for manual inspection. 

[renepaulsen]

Here some more attachments (only 3 files per post?)