Camera Shake Correction Using Adaptive Exposure and Frame Addition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional camera shake correction methods for low-luminance/low-illuminance image sensing in video cameras are ineffective due to unnatural motion and insufficient camera shake correction, especially when using digital image extraction methods, which cannot correct camera shake within a single frame and result in increased noise with multiple image superposition.
Innovation Solution
An image sensing apparatus that detects camera shake and adjusts exposure time based on shake amount, using a first exposure time for small shake and a second, shorter exposure time for larger shake, and adds image signals from multiple frames to produce a single corrected image, thereby improving camera shake correction and reducing noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If a slow shutter mode is used to prolong exposure time for low-luminance image sensing, then image brightness is improved, but camera shake correction effectiveness deteriorates and motion becomes unnatural
Solution Approach 1:
The patent segments the exposure process into multiple frames, each captured at a normal shutter speed. Instead of using a single prolonged exposure, the image sensing unit captures multiple individual frames sequentially, and the image processing unit combines them through addition. This segmentation allows each frame to be properly exposed without motion blur while accumulating sufficient light signal through multiple captures.
Solution Approach 2:
The patent performs preliminary camera shake detection and correction on each individual frame before combining them. The camera shake detection unit detects shake amounts for each frame, and the image processing unit corrects shake for each frame individually before adding them together. This preliminary correction ensures that each frame contributing to the final image has already had its shake corrected, maintaining effectiveness even though the total exposure is distributed across multiple frames.
2Illumination intensity
If multiple images are superimposed to correct camera shake in low-light conditions, then image brightness is improved, but noise increases
Solution Approach 1:
The patent replaces the mechanical slow shutter approach with a digital processing system. Instead of physically extending exposure time with a mechanical shutter, the system uses digital image addition and processing to achieve the same light-gathering effect. The image processing unit adds multiple frames together and applies noise reduction processing, which is more effective at suppressing noise than simply prolonging a single exposure.
Solution Approach 2:
The patent changes the processing parameters by performing camera shake correction and noise reduction on each frame individually before combination, rather than correcting the combined image after summation. This frame-by-frame processing approach, combined with the addition operation, allows for better noise characteristics in the final image compared to traditional slow shutter methods.
3Ease of manufacture
If digital image extraction method is used for camera shake correction, then cost is reduced by eliminating mechanical components, but correction effectiveness deteriorates for single-frame shake
Solution Approach 1:
The patent merges multiple frames together through addition to create a final image with improved camera shake correction. By combining multiple individually corrected frames, the system achieves better overall correction effectiveness than could be obtained from a single frame, while still using only digital processing without mechanical components.
Solution Approach 2:
The patent performs preliminary camera shake correction on each individual frame before combining them. The camera shake detection unit detects shake for each frame, and the image processing unit corrects each frame individually before addition. This preliminary correction of each frame ensures that even single-frame shake is addressed, and the combined effect of multiple corrected frames provides enhanced overall correction effectiveness.
Data Source
AI summary
In an image sensing apparatus which has a camera shake correction function of performing a camera shake correction on a plurality of images sensed by an image sensing unit by extracting a partial image from each of the sensed images on the basis of correlation between the sensed images, a camera shake amount of said image sensing apparatus is detected, and said image sensing unit is exposed for a first exposure time period when the camera shake amount is equal to or less than a predetermined amount. When the camera shake amount is greater than the predetermined amount, said image sensing unit is exposed for a second exposure time period which is shorter than the first exposure time period, image signals of a plurality of images which undergo the camera shake correction are added, and the added image signals is output as one image.