Raw Image Data Storage Across Striped Parallel Memory Modules
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern digital cameras lose image quality when converting raw image data to full-color format due to the reconstruction process, resulting in a reduction from 4,096 to 65,535 shades of color, as seen in JPEG formats compared to raw formats.
Innovation Solution
An image capture system that stores digital image data in a raw format, stripes it into blocks, and distributes these blocks across multiple memory modules for parallel processing, maintaining high color depth and dynamic range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If digital image data is converted from raw format to full-color format for display, then the image becomes displayable, but image quality is lost due to reduction in color depth from 4,096-65,535 shades to 256 shades
Solution Approach 1:
The patent divides the image data processing into separate stages: raw data storage is separated from reconstruction operations. The raw image data is stored in an unprocessed state, while reconstruction to full-color format is performed only when needed for display, allowing both high-quality storage and display capability
Solution Approach 2:
The patent performs preliminary storage of raw image data in its unprocessed form with full color depth before any reconstruction operations. This preliminary action preserves the complete information while allowing later conversion to displayable formats only when required
2Productivity
If image data is stored and processed sequentially in a single memory module, then the system is simpler, but read and write performance is limited
Solution Approach 1:
The patent segments the memory system into multiple memory modules, allowing parallel read and write operations on different portions of the image data. This segmentation increases throughput while distributing the storage load across multiple independent modules
Solution Approach 2:
The patent transitions from a single-memory-module architecture to a multi-memory-module architecture, adding spatial dimensionality to the memory system. This allows simultaneous access to multiple memory locations across different modules, dramatically improving read and write performance
Data Source
AI summary
Systems, methods, and apparatuses can store digital image data that is related to an image, or a series of images, commonly visualized as video. These systems, methods, and apparatuses can store the digital image data as unprocessed digital image data in a raw image format that includes color information, for example, luminance and/or chrominance color components of YUV color model and/or red, green, and/or blue color components of a red, green, blue (RGB) color model to provide some examples, for each pixel of the image. These systems, methods, and apparatuses can stripe the digital image data into multiple blocks of image data and can, thereafter, partitions the multiple blocks of image data into multiple blocks of pixel data. These systems, methods, and apparatuses can distribute the multiple blocks of pixel data across multiple memory modules that can operate in parallel to improve read and/or write performance.