Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

# Compression Method for Gray Scale Image Data by Pseudo Low Pass Filter

IP.com Disclosure Number: IPCOM000106953D
Original Publication Date: 1992-Jan-01
Included in the Prior Art Database: 2005-Mar-21
Document File: 3 page(s) / 64K

IBM

## Related People

Kimoto, H: AUTHOR [+2]

## Abstract

Disclosed is a compression method for gray scale image data which features a predictive coding and a pseudo-low-pass filter. This method codes the multiply of a filter coefficient R (0 < R <= 1) and the differential value between the image data and its prediction value. The smaller R value gives the deeper filtering. The R = 1 means no filtering. The method achieves the less than twice compression time and the equal decompression time to the predictive coding without filtering, and the 1:3 to 1:4 compression ratio against the 1:2 to 1:3 compression ratio by the no filter predictive coding.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Compression Method for Gray Scale Image Data by Pseudo Low Pass Filter

Disclosed is a compression method for gray scale image
data which features a predictive coding and a pseudo-low-pass filter.
This method codes the multiply of a filter coefficient R (0 < R <= 1)
and the differential value between the image data and its prediction
value.  The smaller R value gives the deeper filtering.  The R = 1
means no filtering.  The method achieves the less than twice
compression time and the equal decompression time to the predictive
coding without filtering, and the 1:3 to 1:4 compression ratio
against the 1:2 to 1:3 compression ratio by the no filter predictive
coding.

The effect of the pseudo-low-pass filter appears in Figure 1
that the filtered value follows the original value with some delay.
The delay depends on the coefficient R.

The procedures of the method are as follows:
Compression procedure:
Step 1. Calculate the prediction value Y for the target data Y
by a predictive coding algorithm.
Step 2. Calculate the differential value D for the data Y by the
equation:
D = (Y - Y) * R
where Y, Y, and R are the target data, its prediction, and
the pseudo-low-pass filter coefficient, respectively.
Step 3. Encode the difference D by an entropy coding, such as
the Huffman coding.
Decompression procedure:
Step 1. Calculate the prediction value Y for the filtered data
to be decoded Y' by...