Texture Processing Apparatus Using Classification-Based Data Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 3D graphics rendering technologies consume significant power and resources due to the high number of operations and memory accesses required for texture processing, especially when rendering each frame of 3D graphics.
Innovation Solution
A method and apparatus that reduce the amount of texture data processed by using texture position information and texture classification information (TCI) to determine the minimum amount of texture data needed, allowing for efficient reading and filtering of texture data from memory, thereby minimizing power consumption and operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all texture data is read from memory for each pixel, then texture processing accuracy is maintained, but memory access frequency and power consumption increase significantly
Solution Approach 1:
The patent extracts only the necessary texture data based on TCI classification. By analyzing the texture image and classifying regions with similar texture factors, the system extracts only the minimum required texture data for pixels in homogeneous regions, rather than reading all texture data for every pixel, thus reducing memory access frequency and power consumption while maintaining processing accuracy.
Solution Approach 2:
The patent applies different texture data reading strategies to different regions of the texture image based on their TCI classification. Regions with high texture factor similarity (homogeneous regions) use minimum texture data reading, while regions with low similarity (heterogeneous regions) use complete texture data reading, allowing optimized power consumption and accuracy for each local area.
2Loss of information
If texture data is read for all pixels, then complete texture information is available, but the number of memory operations increases
Solution Approach 1:
The system extracts only the minimum necessary texture data based on TCI analysis. By identifying regions where texture factors are similar and determining that minimum texture data suffices for those regions, the system reduces the total number of memory operations while preserving complete texture information where needed and using minimal data where texture uniformity allows.
Solution Approach 2:
The patent applies partial action by reading only the necessary portion of texture data rather than all texture data. For pixels in homogeneous regions, the system uses partial texture data (minimum required), while for pixels in heterogeneous regions, it uses complete texture data, optimizing the balance between information completeness and processing efficiency.
3Loss of energy
If minimum texture data is read based on TCI, then power consumption and operations are reduced, but texture processing speed may be affected
Solution Approach 1:
The system performs preliminary TCI analysis and texture factor classification before reading texture data. By pre-classifying the texture image into homogeneous and heterogeneous regions based on texture factor similarity, the system can then efficiently read only the minimum necessary data for each region, reducing power consumption while maintaining fast processing through optimized data access patterns.
Data Source
AI summary
Provided is a method of processing a texture. The method includes acquiring texture position information in a texture image corresponding to pixel position information of pixels constituting a frame, acquiring texture classification information (TCI) representing a similarity between respective texture factors of two or more classified regions in the texture image based on the texture position information, determining an amount of texture data requested from a memory according to the TCI, and reading texture data corresponding to the determined amount of texture data based on the texture position information.