Subword Parallelism Color Space Conversion Overflow Prevention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The conventional subword parallelism technique for multimedia data processing faces issues with overflow and increased hardware cost due to unsuitable data arrangement and computation overhead, leading to performance degradation.
Innovation Solution
A method is introduced to convert image signals from RGB to YCbCr color space, reducing the number of data bits through coefficient manipulation and barrel shifting, allowing accurate computation without overflow and minimizing hardware requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional subword parallelism technique is used to process 8-bit pixels, then parallel processing capability is improved, but overflow occurs during computation and hardware cost increases
Solution Approach 1:
The patent segments the color conversion computation into two distinct stages: first computing intermediate values using 8-bit ALUs, then processing these intermediates using 16-bit ALUs. This segmentation allows each stage to operate within its appropriate bit-width range, preventing overflow while maintaining parallel processing efficiency.
Solution Approach 2:
The patent transitions from purely 8-bit computation to a mixed-dimensional approach by introducing 16-bit ALUs for specific intermediate computations. This dimensional change in data representation allows the system to handle values that exceed 8-bit range without overflow, while still leveraging 8-bit parallelism where applicable.
2Reliability
If unpacking process is used to prevent overflow in 8-bit computation, then computation accuracy is improved, but processing delay increases
Solution Approach 1:
The patent performs preliminary computation of intermediate values (such as Y+Cb and Y-Cr) using 8-bit ALUs before the final color conversion step. By preparing these intermediate results in advance with appropriate bit-width, the system avoids the need for unpacking operations during critical path computations, thereby reducing processing delay.
3Reliability
If 12-bit ALUs are used to prevent overflow, then computation accuracy is improved, but hardware cost and size increase
Solution Approach 1:
The patent applies different computational precision to different parts of the computation process: 8-bit ALUs are used for initial pixel value processing where full precision is not yet required, and 16-bit ALUs are used specifically for intermediate computations that require extended range. This local differentiation of computational quality avoids the need for uniformly high-precision (and costly) 12-bit ALUs throughout the entire system.
Data Source
AI summary
Provided is a method for converting a first image signal expressed in a first color space into a second image signal expressed in a second color space on a subword parallelism basis. The method includes determining a first image parameter of a second image signal matrix as a function calculated by multiplying conversion coefficients of a first row constituting an image conversion matrix by image parameters constituting a first image signal matrix; determining a second image parameter of the second image signal matrix as a function expressed by at least one of image parameters constituting the first image signal matrix and the determined first image parameter of the second image signal matrix; determining a third image parameter of the second image signal matrix as a function expressed by at least one image parameter unused for determining the second image parameter of the second image signal matrix among the image parameters constituting the first image signal matrix, and by the determined first image parameter of the second image signal matrix; and finally calculating the second image signal matrix by shifting the determined image parameters constituting the second image signal matrix by a predetermined number of bits.