On-Chip Raw Image Compression for Low-Power Video Capture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video capture devices face challenges in efficiently capturing and storing high-quality video data while preserving creative flexibility and reducing power consumption, particularly in small-form factor devices like smartphones, due to the need for off-chip DRAM for image compression and limited processing capabilities.
Innovation Solution
Implementing on-chip memory and discrete cosine transform (DCT)-based compression algorithms without off-chip DRAM, along with sub-pixel shift-based resolution enhancement, to compress raw mosaiced image data efficiently, thereby reducing power consumption and enabling high-quality video capture.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If off-chip DRAM is used for image compression, then image quality can be maintained, but power consumption increases and device complexity increases
Solution Approach 1:
The patent merges the DRAM and compression algorithms onto the image sensor chip itself, eliminating the need for separate off-chip DRAM components. This integration reduces power consumption by removing the need for data transfer between chip components and eliminates the power overhead of external memory operations, while maintaining image quality through on-chip processing capabilities.
Solution Approach 2:
The patent extracts the compression functionality from external components and embeds it directly within the image sensor chip. By taking out the need for off-chip DRAM and implementing compression algorithms on-chip, the system reduces power consumption and device complexity while preserving the ability to maintain high image quality through efficient on-chip processing.
2Extent of automation
If off-chip DRAM is used for image compression, then image processing capability is enhanced, but device complexity increases
Solution Approach 1:
The patent combines multiple functions (image sensing, DRAM storage, and compression algorithms) into a single integrated chip. This merging eliminates the need for separate off-chip DRAM components and reduces the number of inter-component connections, thereby reducing device complexity while maintaining enhanced image processing capability through on-chip integration.
Solution Approach 2:
The image sensor chip is designed to perform multiple functions: capturing images, storing data in on-chip DRAM, and executing compression algorithms. This multi-functionality eliminates the need for separate dedicated components for each function, reducing overall device complexity while maintaining comprehensive image processing capability.
3Manufacturing precision
If resolution enhancement algorithm is applied to multiple image frames, then output image resolution is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary actions by capturing multiple frames and preparing them for resolution enhancement in advance. By having the frames ready and pre-processed before the actual resolution enhancement algorithm is applied, the system can optimize the processing workflow and reduce the overall processing time while achieving high output image resolution.
Solution Approach 2:
The patent uses multiple copies of the same scene captured in different frames to enhance resolution. By copying information from multiple frames and combining them through the resolution enhancement algorithm, the system achieves higher output resolution more efficiently than processing a single frame, as the redundant information from multiple copies accelerates the reconstruction process.
4Manufacturing precision
If multiple image frames are processed to enhance dynamic range, then dynamic range is improved, but computational load increases
Solution Approach 1:
The patent merges the processing of multiple image frames for dynamic range enhancement directly on the image sensor chip. By combining the computational tasks of capturing multiple frames with different exposures and processing them to enhance dynamic range on-chip, the system reduces the computational load that would otherwise be transferred to external processors, while achieving improved dynamic range performance.
Data Source
AI summary
A still or motion imaging device generates a plurality of image frames. The imaging device can before compression, demosaic the plurality of image frames. The imaging device can also compress the plurality of image frames before tonally processing the plurality of image frames.