Single-Image HDR Processing via Pixel Block Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current high dynamic range (HDR) imaging techniques require multiple original images of different exposure levels to alleviate underexposure and overexposure issues, limiting real-time brightness adjustment capabilities.
Innovation Solution
A method that calculates long and short exposure values for pixel blocks within a single image, using weight values derived from adjacent pixel blocks to adjust brightness, allowing for real-time HDR image generation from a single image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple original images of different exposure levels are used for HDR synthesis, then the recognition rate is improved and underexposure/overexposure issues are alleviated, but the real-time processing capability is reduced and the complexity increases
Solution Approach 1:
The image is divided into multiple pixel blocks, and each pixel block is processed independently to calculate long and short exposure values. This segmentation allows parallel processing of different regions, improving real-time capability while maintaining HDR quality through localized exposure adjustment.
Solution Approach 2:
Instead of requiring multiple actual captured images, the patent creates synthetic long and short exposure values by copying and processing pixel data from a single input image through different calculation paths. This eliminates the need for multiple physical images while achieving HDR effects.
2Manufacturing precision
If multiple original images are required for HDR synthesis, then brightness adjustment accuracy is improved, but the device complexity and operational requirements increase
Solution Approach 1:
The patent makes the single input image serve multiple functions by extracting both long and short exposure information from it through different calculation methods. This multi-functional use of the same input data eliminates the need for multiple input images while maintaining brightness adjustment accuracy.
Solution Approach 2:
The patent introduces weight values as an intermediary element that mediates between the calculated long and short exposure values to produce the final HDR pixel values. This intermediary mechanism enables precise brightness control without requiring complex multi-image alignment and fusion systems.
3Loss of information
If multiple images are used for HDR synthesis, then detail preservation in bright and dark areas is improved, but the processing time and computational load increase
Solution Approach 1:
The patent performs preliminary calculations of brightness sum values and brightness average values for each pixel block before the final HDR synthesis. This preliminary processing organizes the data in advance, enabling faster final computation and reducing overall processing time while preserving detail information.
Data Source
AI summary
A high dynamic range image sensing method for an image sensing device is provided. The high dynamic range image sensing method comprises: calculating a brightness sum value of a pixel block to obtain a long exposure value; calculating a brightness average value of the pixel block to obtain a short exposure value; calculating a block average value of a plurality of adjacent pixel blocks; obtaining a weight value according to the block average value; and outputting a high dynamic range pixel value of one pixel according to the long exposure value, the short exposure value and the weight value.