Adaptive Image Bracket Determination for HDR Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image fusion techniques struggle to generate low noise and high dynamic range (HDR) images in various capturing conditions, especially in low-light situations, due to predetermined exposure settings that fail to adapt to changing lighting conditions and scene compositions, leading to ghosting artifacts and memory inefficiencies in mobile devices.
Innovation Solution
An adaptive approach to image bracket determination and memory-efficient image fusion, where an incoming preview image stream is analyzed to determine target exposure times and capture parameters, allowing for dynamic adjustment of exposure settings and image fusion operations, breaking down fusion into smaller sub-sets to manage memory usage effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple images are captured with predetermined exposure settings for fusion, then noise reduction and HDR image quality are improved, but ghosting artifacts occur due to inability to adapt to changing lighting conditions and scene motion
Solution Approach 1:
The patent implements dynamic exposure time adjustment by analyzing preview image streams to determine scene characteristics (lighting conditions, motion levels) and adaptively selecting exposure times for bracketed captures. This dynamic approach replaces fixed predetermined exposure settings, allowing the system to optimize exposure parameters in real-time based on actual scene conditions, thereby reducing ghosting artifacts while maintaining image quality.
Solution Approach 2:
The system uses feedback from preview image analysis to guide the capture process. By continuously monitoring the preview stream and analyzing scene characteristics, the system adjusts exposure parameters for subsequent captures based on actual observed conditions rather than predetermined settings. This feedback loop enables adaptive optimization of capture parameters to minimize ghosting while maximizing image quality.
2Reliability
If multiple images are captured with longer aggregate exposure times to reduce noise in low-light conditions, then signal-to-noise ratio is improved, but device stability becomes difficult to maintain and memory usage increases
Solution Approach 1:
The patent segments the total exposure requirement into multiple individual image captures with optimized exposure times rather than requiring a single long exposure. By capturing multiple images with shorter individual exposure times and fusing them, the system achieves the necessary signal-to-noise ratio while minimizing the impact of device motion, as each short capture freezes motion better than a single long exposure would.
Solution Approach 2:
The system employs periodic capture of multiple images at optimized exposure times rather than continuous long exposure. This periodic action allows the device to capture sufficient light information across multiple brief intervals, achieving low-noise results while maintaining stability between captures. The fusion process combines these periodic captures to produce the final high-quality image.
3Reliability
If multiple images are captured with different exposures for HDR imaging, then dynamic range is improved, but capture time increases and device stability becomes more difficult to maintain
Solution Approach 1:
The patent implements dynamic exposure time selection based on real-time preview analysis. Instead of using fixed predetermined exposure brackets, the system analyzes the preview stream to determine optimal exposure times tailored to the specific scene conditions. This dynamic approach captures the necessary dynamic range information more efficiently, reducing the total capture time required while maintaining HDR quality.
4Reliability
If image fusion is performed on complete sets of captured images, then comprehensive noise reduction is achieved, but memory footprint becomes prohibitively large for mobile devices
Solution Approach 1:
The patent segments the image fusion process into batches of smaller subsets rather than loading all images into memory simultaneously. The system processes images in manageable groups, performing fusion operations on each batch sequentially. This segmentation approach maintains comprehensive noise reduction quality while keeping memory footprint within acceptable limits for mobile devices by processing and discarding batches as they are completed.
Data Source
AI summary
An adaptive approach to image bracket determination and a more memory-efficient approach to image fusion, which are designed to generate low noise and high dynamic range (HDR) images in a wide variety of capturing conditions, are described. An incoming preview image stream may be obtained from an image capture device. When a capture request is received, an analysis may be performed on an image from the preview image stream that has a predetermined temporal relationship to the image capture request. Based on the analysis, a set of images (and their respective capture parameters, e.g., exposure time) may be determined for the image capture device to capture. As the determined set of images are captured, they may be registered and fused in a memory-efficient manner that, e.g., places an upper limit on the overall memory footprint of the registration and fusion operations—regardless of how many images are captured in total.