12-bit iDCT or DCT?

[unix_guru]

has anyone implemented this yet?

[Nate.H]

Hi,

As of Version 8.0, LabVIEW actually ships with VIs (DCT.vi & Inverse DCT.vi) that will compute the Direct Cosine Transform (DCT) and it's inverse. They can be found in the Functions pallet under Signal Processing>>Transforms. They are not, however, included in the Base LabVIEW development package.

Assuming that you are using these functions for image processing and have the Vision Development Module for LabVIEW, you could use the IMAQ ImagetoArray VI to convert an image to a 2D array that you could then input to the DCT or Inverse DCT VIs.

If the above information does not adequately answer your question, please let me know what exactly it is you are trying to do with the DCT/iDCT we can go from there.

Cheers!
_____________
Nate
Applications Engineer
 

[unix_guru]

hello, look at my posting about the basler 1900 camera. national instruments in a partner with basler. most DCT are 8-bit, when you move into 10-bit or 12-bits it is a different world. if you have a 12-bit DCT and iDCT, i'd like to see the code. anyone can do 8-bit. thanks.

[Nate.H]

Hi unix_guru,

See the attached folder for a modified LabVIEW shipping example. It implements image compression with the DCT and Inverse DCT functions that I mentioned in my previous post. I have changed all the data types to  I16 and all the data types the image array is converted to will maintain this precision. I do not see how it could be truncated to 8-bits using this procedure, as the DCT and Inverse DCT functions take in a double to the DLL call.

There are additional compression examples that ship with LabVIEW. Go to the example finder and search for "compression". This should bring up example results including the use of:

principal component analysis (PCA)
discrete wavelet transform (DWT)
integer wavelet transform (IWT)