Spatially multiplexed exposure image processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image sensors struggle to capture high dynamic range images effectively, as images captured with different exposure times or sensitivities often result in loss of detail in either bright or dark areas due to saturation or underexposure, and combining these images can lead to blurring or loss of detail.
Innovation Solution
A method and apparatus that receive input image data from two sets of pixel locations with different exposure levels, generate interpolated image data, apply a noise model to detect motion, and combine the data based on motion detection to produce improved image data with higher dynamic range, avoiding blurring and preserving detail.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If images are captured with different exposure times to capture detail in both bright and dark areas, then the dynamic range is improved, but motion artifacts and blurring occur when combining the images
Solution Approach 1:
The patent applies preliminary action by performing motion detection and compensation operations before combining the short and long exposure images. The system detects motion between the two exposures and applies corrective transformations to align the images, preventing motion artifacts and blurring in the final HDR image while preserving the improved dynamic range.
2Loss of information
If interpolation is used to infer details at missing pixel positions when combining images from different exposure times, then the image completeness is improved, but noise and detail loss increase
Solution Approach 1:
The patent applies local quality by using motion detection data to guide the interpolation process locally. Instead of uniform interpolation across the entire image, the system identifies regions with motion and applies different processing strategies - using motion-compensated data in moving regions and interpolation only in static regions, thereby preserving detail accuracy while maintaining image completeness.
3Productivity
If some sensor pixels capture short exposure images and other pixels capture long exposure images, then the processing speed is improved, but the image quality deteriorates due to interpolation requirements
Solution Approach 1:
The patent applies feedback by using motion detection data derived from comparing the short and long exposure images to guide the combination process. The motion detection results provide feedback about scene changes, allowing the system to selectively combine pixel data from the two exposures based on actual motion patterns, thereby improving image quality while maintaining the parallel processing advantage.
Data Source
AI summary
Methods and apparatus for generating improved image data from received input image data comprising first input image data associated with a first exposure level and second input image data associated with a second, different, exposure level. Motion detection data is generated from the received input image data by applying a noise model and improved image data is generated by combining data from the first and second input data in dependence on the motion detection data.