Block entropy coding in embedded block coding with optimized truncation image compression

• US 6,658,159 B1
• Filed: 03/17/2000
• Issued: 12/02/2003
• Est. Priority Date: 03/17/2000
• Status: Expired due to Fees

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Taubman, David S.
• Primary Examiner(s): Chen, Wenpeng
• Application Number:
  US09/527,758
• Time in Patent Office: 1,355 Days
• Field of Search: 382/240, 382/238, 382/237, 382/239, 382/246, 382/247, 382/248, 382/232, 341/50, 375/240.03, 375/240.19
• US Class Current: 382/240
• CPC Class Codes: H04N 19/645 by grouping of coefficients into blocks after the transform
• CPC Class Codes: H04N 19/647 using significance based coding, e.g. Embedded Zerotrees of Wavelets [EZW] or Set Partitioning in Hierarchical Trees [SPIHT]

Abstract

Embedded block entropy coding with optimized truncation is useful for image compression schemes in conjunction, for example, with a Wavelet transform, so as to form a bit-stream which can be stored or transmitted with increased efficiency. In the arithmetic coding phase, several opportunities exist for minimizing complexity and improving compression and computational performance as compared to existing systems. One method involves bypassing the arithmetic coding procedure for certain lower significance bit-planes of each code-block. Then, the raw binary digits from these bit-planes can be interleaved into the arithmetically coded bit-stream. As a result, both the average number of arithmetically coded symbols and the maximum number of coding passes per code-block can be significantly reduced.

Claims

1. A method for compressing image data, comprising the steps of:

decomposing the image data into code-blocks of coefficients using a transform, each code-block comprising a plurality of bit-planes from a most significant bit-plane to a least significant bit-plane;

processing bit-planes of coefficient data in the code blocks in multiple coding passes to generate raw bit-plane data;

arithmetically coding a portion of raw bit-plane data to generate arithmetically coded data; and

writing the arithmetically coded data and the raw bit-plane data not arithmetically coded directly into a bit-stream;

wherein raw bit-plane data generated during significance propagation and magnitude refinement coding passes for bit-planes p<

p₀−

K is written directly into the bit-stream, wherein p₀ denotes the most significant bit-plane of a code block in which any sample therein becomes contextually significant during arithmetic coding and K is an integer parameter.

2. The method of claim 1, wherein K=3.

3. The method of claim 1, wherein the method for compressing image data is based on embedded block coding with optimized truncation and employs a wavelet transform.

4. The method of claim 1, wherein arithmetically coded data is interleaved with raw bit-plane data in the bit-stream.

5. A method for compressing image data, comprising the steps of:

decomposing the image data into code-blocks of coefficients using a transform, each code-block comprising a plurality of bit-planes from a most significant bit-plane to a least significant bit-plane;

processing bit-planes of coefficient data in the code blocks in multiple coding passes to generate raw bit-plane data;

arithmetically coding a portion of raw bit-plane data to generate arithmetically coded data; and

writing the arithmetically coded data and the raw bit-plane data not arithmetically coded directly into a bit-stream;

wherein raw bit-plane data generated during a normalization coding pass for bit-planes p<

p₀−

K is arithmetically coded before being written directly into the bit-stream, wherein p₀ denotes the most significant bit-plane of a code block in which any sample therein becomes contextually significant during arithmetic coding and K is an integer parameter.

6. The method of claim 5, wherein K=3.

7. The method of claim 5, wherein the method for compressing image data is based on embedded block coding with optimized truncation and employs a wavelet transform.

8. The method of claim 5, wherein arithmetically coded data is interleaved with raw bit-plane data in the bit-stream.