Encoding and decoding data arrays
Some embodiments of the invention provide a method of performing a Discrete Cosine Transform (“DCT”) encoding or decoding coefficients of a data array by (1) multiplying the coefficients by a scalar value before the encoding or decoding, and then (2) dividing the encoded or decoded coefficients by the scalar value. When used in conjunction with fixed-point arithmetic, this method increases the precision of the encoded and decoded results. In addition, some embodiments provide a method of performing a two-dimensional (2D) Inverse Discrete Cosine Transform (“iDCT”). This method splits a pre-multiplication operation of the iDCT into two or more separate stages. When used in conjunction with fixed-point arithmetic, this splitting increases the precision of the decoded results of the iDCT.
1. A method comprising:
decoding an encoded video stream that has been encoded according to a two-dimensional (2D) transform encoding operation that is separable into two one-dimensional (1D) transform operations, the encoded video stream comprising a plurality of encoded values for a plurality of encoded video images, said decoding comprising:
parsing encoded values out of the data stream and creating a two-dimensional data array that stores the encoded values in a particular scan order, wherein the values in the created data array are encoded in both dimensions of the array;
multiplying each value in the data array by a scalar value, wherein the data array is an array of data values from a video image;
performing a first 1D inverse transform to the data array resulting from the multiplying;
transposing the data array resulting from the first 1D inverse transform;
performing a second 1D inverse transform to the data array resulting from the transposing; and
dividing by the scalar value each value in the data array resulting from the second 1D inverse transform to produce a data array comprising decoded values, the data array comprising decoded values being produced without a second transposing step.
2. The method of claim 1 , wherein the particular scan order is a transposed zig-zag scan order.
3. The method of claim 1 , wherein each of the plurality of the video images comprises one or more macroblocks.
4. The method of claim 3 , wherein each macroblock comprises sub-sections of luminance and chrominance blocks.
5. The method of claim 3 , wherein each macroblock is a block of 16 by 16 pixels.
6. The method of claim 1 , wherein the data array is an 8 by 8 pixel image block.
7. The method of claim 1 , wherein the data array is an n by m pixel image block, wherein n and m are integers.
8. A method comprising:
encoding a video image stored as a data array of coefficients, said encoding based on a two-dimensional (2D) transform operation that is separable into two one-dimensional (1D) transform operations, said encoding comprising:
multiplying each coefficient in the data array by a scalar value;
performing a first 1D transform on the data array resulting from the multiplying;
transposing the data array resulting from the first 1D transform;
performing a second 1D transform on the data array resulting from the transposing;
dividing by the scalar value each coefficient in the data array resulting from the second 1D transform; and
without performing another transposing step, generating a data stream based on the coefficients from the data array resulting from the dividing, wherein the data stream comprises an encoding of said video image.
9. The method of claim 8 further comprising, after performing the second 1D transform and before the dividing, performing a post-multiplication of the each coefficient in the data array resulting from the second 1D transform.
10. The method of claim 8 , wherein the video image comprises one or more macroblocks.
11. The method of claim 10 , wherein each macroblock comprises sub-sections of luminance and chrominance blocks.
12. The method of claim 10 , wherein each macroblock is a block of 16 by 16 pixels.
13. The method of claim 8 , wherein the data array is an 8 by 8 pixel image block.
14. The method of claim 8 , wherein the data array is an n by m pixel image block, wherein n and m are integers.