Apparatus for and method of transcoding data

• US 8,059,716 B2
• Filed: 09/11/2006
• Issued: 11/15/2011
• Est. Priority Date: 09/10/2005
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd., Seoul National University Industry Foundation (Seoul National University)
• Original Assignee: Samsung Electronics Co. Ltd., Seoul National University Industry Foundation (Seoul National University)
• Inventors: Park, Chan-Yul, Cho, Nam-Ik
• Primary Examiner(s): Czekaj; David
• Assistant Examiner(s): HOLDER, ANNER N
• Application Number:
  US11/518,357
• Publication Number:
  US 20070058720A1
• Time in Patent Office: 1,891 Days
• Field of Search: 375/240.03, 375/240.2, 375/240.24, 375/240.25
• US Class Current: 375/240.12
• CPC Class Codes: H04N 19/40 using video transcoding, i.e. partial or full decoding of a coded input stream followed by re-encoding of the decoded output stream
• CPC Class Codes: H04N 19/48 using compressed domain processing techniques other than decoding, e.g. modification of transform coefficients, variable length coding [VLC] data or run-length data motion estimation in a transform domain H04N19/547; processing of decoded motion vectors H04N19/513
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

An apparatus for and a method of transcoding data in a first format, for example, MPEG (moving picture experts group)-2 data into data in a second format, for example, H.264 data. The apparatus may include a scaling dequantization unit, a coefficient transform unit, and/or a scaling quantization unit. The scaling dequantization unit may dequantize the data in the first format (e.g., MPEG-2 data) to generate a DCT (discrete cosine transform) coefficient. The coefficient transform unit may approximate a 4x4 second data format (e.g., H.264) integer transform matrix to a 4x4 DCT matrix and transforms the DCT coefficient into an integer transform coefficient using the approximated 4x4 DCT matrix. The scaling quantization unit may quantize the integer transform coefficient to generate the data in the second format (e.g., H.264 data).

Claims

1. A method of transcoding MPEG (moving picture experts group)-2 data into H.264 data, the method comprising:

dequantizing the MPEG-2 data to generate a DCT (discrete cosine transform) coefficient;

approximating a 4×

4 H.264 integer transform matrix as a 4×

4 DCT matrix, and transforming the DCT coefficient into an integer transform coefficient using the 4×

4 DCT matrix; and

quantizing the integer transform coefficient to generate the H.264 data;

wherein the transforming of the DCT coefficient includes,

changing an 8×

8 DCT matrix into a product of a first diagonal matrix and a plurality of matrices and changing the 4×

4 H.264 integer transform matrix into a product of a second diagonal matrix and the 4×

4 DCT matrix,

changing a combination matrix expressed as a combination of some of the plurality of matrices and the 4×

4 DCT matrix into a product of a third diagonal matrix and a coefficient transform matrix having a symmetric structure,

performing primary scaling on the DCT coefficient using the first diagonal matrix,

two-dimensionally transforming the primarily scaled data using a product of the coefficient transform matrix and remaining ones of the plurality of matrices except the some of the plurality of matrices, and

performing secondary scaling on the two-dimensionally transformed data using the second diagonal matrix and the third diagonal matrix to generate the integer transform coefficient.

2. The method of claim 1, wherein changing the combination matrix comprises:

generating the combination matrix through multiplication of the some of the plurality of matrices and four 4×

4 DCT matrices; and

changing the combination matrix into a product of the third diagonal matrix and the coefficient transform matrix such that the value of one pair among symmetric element pairs of the coefficient transform matrix is 1.

3. The method of claim 1, wherein when transcoding the MPEG-2 data into the H.264 data with a ½

reduction of resolution, the transforming of the DCT coefficient into the integer transform coefficient comprises;

changing the 4×

4 H.264 integer transform matrix into a product of the second diagonal matrix and the 4×

4 DCT matrix;

generating a 4×

4 DCT coefficient matrix corresponding to low frequency components from the DCT coefficient; and

performing scaling on the generated 4×

4 DCT coefficient matrix using the second diagonal matrix to generate the integer transform coefficient.

4. An apparatus for transcoding MPEG (moving picture experts group)-2 data into H.264 data, the apparatus comprising:

a scaling dequantization unit configured to dequantize the MPEG-2 data to generate a DCT (discrete cosine transform) coefficient;

a coefficient transform unit configured to approximate a 4×

4 H.264 integer transform matrix as a 4×

4 DCT matrix, and transform the DCT coefficient into an integer transform coefficient using the 4×

4 DCT matrix; and

a scaling quantization unit configured to quantize the integer transform coefficient to generate the H.264 data;

wherein

when transcoding the MPEG-2 data into the H.264 data without a resolution change, the coefficient transform unit changes an 8×

8 DCT matrix into a product of a first diagonal matrix and a plurality of matrices, changes the 4×

4 H.264 integer transform matrix into a product of a second diagonal matrix and the 4×

4 DCT matrix, and changes a combination matrix expressed as a combination of some of the plurality of matrices and the 4×

4 DCT matrix into a product of a third diagonal matrix and a coefficient transform matrix having a symmetric structure.

5. The apparatus of claim 4, wherein the scaling dequantization unit is configured to perform primary scaling on the DCT coefficient using the first diagonal matrix, the coefficient transform unit is configured to two-dimensionally transform the primarily scaled data using a product of the coefficient transform matrix and remaining ones of the plurality of matrices except the some of the plurality of matrices, and the scaling quantization unit is configured to perform secondary scaling on the two-dimensionally transformed data using the second diagonal matrix and the third diagonal matrix to generate the integer transform coefficient.

6. The apparatus of claim 4, wherein the coefficient transform unit is configured to generate the combination matrix through multiplication of some of the plurality of matrices and four 4×

4 DCT matrices, and configured to change the combination matrix into a product of the third diagonal matrix and the coefficient transform matrix such that the value of one pair among symmetric element pairs of the coefficient transform matrix is 1.

7. The apparatus of claim 4, wherein when transcoding the MPEG-2 data into the H.264 data with a ½

reduction of resolution, the coefficient transform unit changes the 4×

4 H.264 integer transform matrix into a product of the second diagonal matrix and the 4×

4 DCT matrix and generates a 4×

4 DCT coefficient matrix corresponding to low frequency components from the DCT coefficient, and the scaling quantization unit performs scaling on the 4×

4 DCT coefficient matrix using the second diagonal matrix to generate the integer transform coefficient.