What should be the values given to the input integer , bits size of the integer and the K value for the input image?

What should be the values given to the input integer , bits size of the integer and the K value for an image input for the rice coder program give in this url http://www.mathworks.in/matlabcentral/fileexchange/5359-rice-coder .Also how to write the inverse code of rice-codes ?

Answer 1

[Jan]

This is a very old submission and some users have rated it low. If the author does not tell the details, it is strongly recommended not to use the software. It would be inefficient and prone to errors, if the forum users guess the details.

Therefore the inverse operation is hard to guess also.

I assume programming it from scratch will be more efficient.

Answer 2

[Walter Roberson]

The value of the bit size of the integer is something you determine in advance according to the properties of the distribution of values you want to encode. Rice encoding is suitable for distributions in which many of the values could fit into small integers but every once in a while you need to send something larger. How you determine the appropriate value is something you should be studying as part of your theory.

If it helps any, the "bit size of the integer" is log2 of "M" in the Wikipedia description of Rice encoding.

The "input integer" and the "bits to be sent directly" should be such that "input integer" * 2 ^ "bits size of the integer" + "bits to be sent directly" is the original value that is being encoded.