CACHE FRIENDLY METHOD FOR PERFORMING INVERSE DISCRETE WAVELET TRANSFORM

摘要

This invention is a method for inverse Wavelet transform using a breadth- first output data calculation which uses input data to calculate at least one output data for each iteration of a software loop even if the same input data is used in a later iteration for calculating other output data. This reduces data movement between memory and the data processor core thus reducing the possibility of cache misses and memory stalls due to access conflicts. The input data and computed output data are preferably stored as subwords packed within data words in memory. In inverse Wavelet transformation, this method performs vertical spatial frequency expansion and horizontal spatial frequency expansion for each level of Wavelet encoding. This invention arranges data flow providing a more efficient use of memory bandwidth and cache space than other known methods. A method for inverse Wavelet transform comprises the steps of storing input data as subwords packed within data words in a memory; calculating output data in a software loop in a breadth-first fashion by recalling input data (902) and filter coefficients (903, 905) for computation of at least one output data for each iteration of the software loop even if the same input data is used in a later iteration for calculating other output data (904, 906); and storing calculated output data as subwords packed within data words in the memory.