Bayer Sensor Image Compression via Quarter-Resolution Plane Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing technologies face challenges in efficiently compressing and decoding video data from single-sensor Bayer imagers, as direct compression methods result in little data reduction or image distortion due to the interleaved and less correlated color primaries, and traditional demosaicing processes are computationally intensive and introduce visual artifacts.
Innovation Solution
The method involves separating a Bayer frame into four quarter-resolution planes, applying color differencing and encoding these planes using common compression techniques like DCT or wavelet compression, and modifying the decompression algorithm to reconstruct images at quarter or full resolution as needed, allowing for efficient data reduction without visual distortion and reducing computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If direct compression is applied to Bayer sensor data, then compression speed is improved, but data reduction is minimal and image distortion occurs
Solution Approach 1:
The Bayer frame is segmented into four quarter-resolution planes (R1, G1, G2, B1) by separating and downsampling the color channels. This segmentation allows each plane to be compressed independently, achieving significant data reduction while preserving the ability to reconstruct the full-resolution image through selective decoding of specific planes based on the needed color information.
2Measurement precision
If traditional demosaicing is applied, then full-resolution color image is obtained, but computational load increases and visual artifacts are introduced
Solution Approach 1:
Instead of performing full demosaicing on all pixels, the method applies partial action by selectively decoding only the necessary quarter-resolution planes (R1, G1, G2, B1) based on the specific color information needed. This partial decoding approach eliminates the need for computationally intensive demosaicing algorithms while avoiding visual artifacts, as the Bayer pattern's inherent structure is preserved and directly utilized.
3Measurement precision
If all four quarter-resolution planes are decoded, then full image quality is achieved, but data transmission and processing volume increases
Solution Approach 1:
The method applies local quality by selectively decoding only the specific quarter-resolution planes needed for the current processing task. For example, if only luminance information is needed, only the G1 and G2 planes are decoded; if only red channel information is needed, only the R1 plane is decoded. This selective approach maintains high image quality where needed while minimizing data transmission and processing volume by avoiding unnecessary decoding of other planes.
Data Source
AI summary
A method is described to greatly improve the efficiency of and reduce the complexity of image compression when using single-sensor color imagers for video acquisition. The method in addition allows for this new image compression type to be compatible with existing video processing tools, improving the workflow for film and television production.