MPEG-2 to H.264 Transcoding via Coefficient Matrix Approximation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current transcoding methods for converting MPEG-2 video data into H.264 format are inefficient in terms of time and computation, particularly due to the differences in discrete cosine transform (DCT) and integer transform matrices between the two formats, leading to increased computational costs and errors.
Innovation Solution
The proposed solution involves an apparatus and method that uses a combination of inverse discrete cosine transform (IDCT) in MPEG-2 encoding and integer transform in H.264 encoding by approximating the integer transform matrix in H.264 to a DCT matrix in MPEG-2, utilizing scaling dequantization, coefficient transformation, and scaling quantization units to reduce computational complexity and errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pixel region transforming method is used to convert MPEG-2 data to H.264 data, then format conversion is achieved, but time and computational efficiency deteriorate due to full encoding and decoding in two formats
Solution Approach 1:
The patent extracts only the essential transform operations (DCT to integer transform) from the complete encoding/decoding process. By isolating and optimizing the specific transformation step needed for format conversion, it avoids the computational overhead of full pixel-based re-encoding while maintaining format compatibility.
Solution Approach 2:
Instead of following the conventional forward path of decoding MPEG-2 to pixels then re-encoding to H.264, the patent inverts the approach by directly transforming the DCT coefficients through an inverse transform path. This reverse engineering of the transform chain enables efficient coefficient-level conversion without pixel intermediate processing.
2Productivity
If simple transform of 8×8 DCT coefficient matrix to 4×4 integer transform coefficient matrix is performed, then format conversion is achieved, but computational complexity increases due to matrix approximation requirements
Solution Approach 1:
The patent segments the 8×8 DCT coefficient matrix into multiple 4×4 blocks, allowing independent transformation of each block. This segmentation enables the use of simpler 4×4 integer transform matrices while maintaining overall image quality, reducing the computational burden compared to transforming the entire 8×8 matrix at once.
Solution Approach 2:
The patent changes the transformation parameters by selecting specific approximation matrices (such as using rounded cosine values) that simplify the mathematical operations. By adjusting the precision parameters of the transform matrix elements, it achieves a balance between computational simplicity and transformation accuracy.
3Ease of operation
If DCT coefficient matrix element arrangement is changed to match integer transform matrix, then transformation simplicity is improved, but manufacturing precision deteriorates due to element repositioning errors
Solution Approach 1:
The patent introduces an intermediary transformation step that maps the 8×8 DCT coefficient elements to the 4×4 integer transform positions through a defined sampling pattern. This intermediary mapping process preserves the relative positional relationships and weighting of coefficients while adapting them to the target matrix structure, minimizing accuracy loss.
Data Source
AI summary
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).