Method and apparatus for implementing wavelet filters in a digital system
AssigneeMotorola, Inc.
InventorsYolanda Prieto, Jose I. Suarez, Yolanda M. Pirez
Life of this patent
Estimating the term…
Abstract
A digital data system (100) provides 1-D, 2-D and 3-D capability and multi-band channel capability. Improved filter banks are created by generating a filter bank having an analysis portion and synthesis portion and obtaining wavelet coefficients (302) for each portion. The wavelet coefficients are expressed in a format capable of canonical signed digit (CSD) representation, such as integers (302). The canonical signed digit (CSD) representation is controlled by a value, N, selected to control resolution of the CSD coding. Optimized CSD-coded wavelet coefficients are used as filters for data signals (318).
Claims27
A method for decomposing and inverse decomposing data signals in a digital data system, comprising the steps of:
- receiving an input data signal through a decomposition stage of the digital data system;
- generating a series of wavelet coefficients for an analysis filter bank and a synthesis filter bank;
- selecting a value, N, to represent the number of powers-of-two terms to represent the wavelet coefficients;
- Canonical Signal Digit (CSD) coding the wavelet coefficients based on the selected value of N;
- adjusting the value of N depending on a predetermined threshold to obtain optimized CSD-coded wavelet coefficients for the analysis filter bank and the synthesis filter bank;
- filtering the input data signal using the optimized CSD-coded wavelet coefficients of the analysis filter bank in the decomposition stage of the digital data system;
- generating a compressed data signal from the filtered input data signal in a quantization stage of the digital data system;
- coding the compressed data signal into a compressed coded signal in a coding stage of the digital data system;
- decoding the compressed coded signal and generating a compressed decoded signal in an inverse coding stage of the digital data system;
- generating a decompressed data signal from the compressed decoded signal in an inverse quantization stage of the digital data system;
- filtering the decompressed data signal using the optimized CSD-coded wavelet coefficients of the synthesis filter bank in an inverse decomposition stage of the digital data system to generate an output signal substantially similar to the input data signal.
A method of forward and inverse decomposing an input data signal and a compressed data signal in a lossy digital data system, comprising the steps of:
- receiving an input signal through the a decomposition stage of the digital data system;
- generating a series of wavelet coefficients for an analysis filter bank as part of an encoder stage of the digital data system;
- generating a series of wavelet coefficients for a synthesis filter bank as part of the decoder stage of the digital data system;
- representing the wavelet filter coefficients for the analysis filter bank as canonical signal digit (CSD) thereby forming a CSD-coded optimal representation of the analysis filter;
- representing the wavelet filter coefficients for the synthesis filter bank as canonical signal digit (CSD) thereby forming a CSD-coded optimal representation of the synthesis filter;
- selecting a value, Ma, to represent the resulting number of CSD-coded coefficients from the optimal representation for the analysis filter, adaptively adjusting the value of Ma to reduce the number of CSD-coded coefficients from the optimal representation of the analysis filter;
- performing a forward signal decomposition for the input data signal using the CSD-coded analysis filter;
- and performing an inverse signal transformation on the compressed data signal using the CSD-coded synthesis filter to generate an output signal substantially similar to the input data signal.
Dependencies are inferred from the claim text.