Subband coding method with wavelet transform for high efficiency video signal compression

Claim 1

A subband coding method for dividing a digital video signal into a plurality of frequency bands in vertical and horizontal directions in a spatial frequency region comprising the steps of: executing two-dimensional multilayer (n-layer) wavelet transform on each processing unit of said digital video signal, extracting data from said plurality of frequency bands in proportion to a number of data in each frequency band, rearranging extracted data into a plurality of data blocks each consisting of 2"×2" pieces of data, and scanning the extracted data in said data blocks by scanning from data extracted from a layer consisting of lowest frequency bands to data extracted from a layer consisting of highest frequency bands and by horizontally scanning data extracted from an LH band consisting of high frequency components in a vertical direction and low frequency components in a horizontal direction, then vertically scanning data extracted from an HL band consisting of low frequency components in said vertical direction and high frequency components in said horizontal direction, and then scanning data extracted from an HH band consisting of high frequency components in said vertical and horizontal directions to generate a data train and accomplish run-length coding.