Codebook search methods
View Patent ↗A method of compressing an image frame composed of an array of pixels in the form of digital signals comprises a two stage codebook search. In the first stage the pixelated image frame and the pixelated codebook patches are transformed to frequency domain coefficients and a pattern comparison is made between the coefficients of the image patch and the codebook patches to identify a short list of possible match codebook patches by discarding incorrectly matching patterns. In the second stage the image frame is compared with the short list of codebook patches by any desired method, e.g. by pixel comparisons, to select the best matching codebook patch.
1. A codebook search method comprising the steps of:
a) providing a plurality of codebook patches;
b) providing at least one image patch;
c) transforming by means of a lossless transform the plurality of codebook patches and the at least one image patch into sets of frequency domain codebook coefficients and a set(s) of frequency domain image coefficients, respectively;
d) for the/each image patch selecting a shortlete (ie. shortlist) of possible match codebook patches by means of a pattern comparison of + and − signs of the image coefficients of the/each image patch with the codebook coefficients; and
e) undertaking a further comparison of the image patch(es) against the shortlete codebook patches to select a substantially best match, this further comparison being effected by any desired method.
2. A method as claimed in claim 1 , wherein the number of image coefficients used in the pattern comparison of step d) is, at least initially, equal in number to the number of pixels in a patch.
3. A method as claimed in claim 2 , wherein the number of image coefficients required to make the pattern comparison of step d) is equal in number to those most significant in the frequency spectrum representative of the codebook coefficients.
4. A method as claimed in any preceding claim, wherein the transform of step c) is computed for each of the codebook patches and the codebook coefficients are stored with the codebook patch to which they apply.
5. A method as claimed in any preceding claim, wherein for each codebook patch a ‘sign string’ is established, comprising a multi-bit pattern in which each bit corresponds to a specific instance of the set of multiple coefficients chosen for examination, each bit in the string being assigned a first binary value if the corresponding coefficient is negative or a second binary value if the corresponding coefficient is positive.
6. A method of compressing an image frame composed of an array of pixels in the form of digital signals, including the step of providing a codebook search method as claimed in any preceding claim.
7. An apparatus for compressing an array of data entries in the form of digital signals, the apparatus comprising a computer apparatus set up to run a program embodying a codebook search method according to any one of claims 1 – 5 .
8. A computer program product carrying a computer program embodied in a computer readable medium adapted to perform the codebook search method according to any one of claims 1 – 5 .