RAW Data Encoding for High Dynamic Range Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image capture devices face challenges in encoding RAW data with varying exposure times, leading to increased data volume due to high frequency components generated by pixel level differences, which reduces coding efficiency.
Innovation Solution
The proposed solution involves generating and encoding RAW data for specific exposure times by separating data into short and long exposure components, using a processor and memory to perform operations such as channel transformation, frequency transform, and quantization, which reduces high frequency components and data volume.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If RAW data with different exposure times is encoded directly, then high dynamic range imaging is achieved, but large level differences generate high frequency components that reduce coding efficiency
Solution Approach 1:
The patent segments the RAW data encoding process by separating pixels into different exposure time groups (first exposure time and second exposure time). This segmentation allows each group to be encoded independently with appropriate quantization parameters, avoiding the mixing of large level differences that would generate high frequency components and reduce coding efficiency.
Solution Approach 2:
The patent applies different quantization parameters based on exposure time groups. By changing the quantization parameter according to the exposure time category of each pixel, the encoding process adapts to the specific characteristics of each group, reducing unnecessary high frequency components while preserving important image information and improving overall coding efficiency.
2Quantity of substance
If RAW data is compressed encoded, then recording data amount is reduced, but coding efficiency drops due to high frequency components from pixel level differences
Solution Approach 1:
The patent dynamically adjusts quantization parameters based on the exposure time groups of pixels. This parameter change strategy allows for effective compression by using coarser quantization for pixels where it is less critical, while maintaining finer quantization where needed, thus reducing overall data amount without severely compromising coding efficiency.
Solution Approach 2:
The patent applies different encoding quality levels to different pixel groups based on their exposure times. By making the encoding quality local rather than uniform, the system achieves better overall compression ratios while maintaining acceptable quality in critical regions, effectively balancing data reduction with coding efficiency.
Data Source
AI summary
An apparatus comprises a generating unit configured to generate a plurality of pieces of RAW data for respective exposure times from RAW data obtained from a sensor that can perform shooting at an exposure time that is different for each pixel, and an encoding unit configured to encode the generated plurality of pieces of RAW data.