Image Stabilization Weighting for Wide-Angle Perspective Distortion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image stabilization techniques are ineffective in correcting image blur caused by ultra wide-angle lenses, which result in decreased accuracy of motion vector detection and unstable image stabilization due to the significant perspective distortion they introduce.
Innovation Solution
An image stabilization apparatus that adjusts weights applied to motion vectors and camera shake data based on lens perspective information to calculate the image blur correction amount, improving the accuracy and stability of image stabilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If ultra wide-angle lenses are used to capture images, then the angle of view is widened, but perspective distortion increases causing decreased motion vector detection accuracy
Solution Approach 1:
The patent applies different weights to motion vectors from different regions of the image. Specifically, motion vectors from the peripheral region (which suffers from perspective distortion) are given lower weights, while motion vectors from the central region (which has less distortion) are given higher weights. This local differentiation of quality allows the system to utilize motion vector information while mitigating the impact of perspective distortion.
Solution Approach 2:
The patent changes the parameter of motion vector weighting based on the detected perspective amount. When perspective distortion is detected to be large, the system adjusts the weights applied to motion vectors accordingly, reducing the influence of distorted peripheral motion vectors. This dynamic parameter adjustment allows the system to adapt to varying degrees of perspective distortion.
2Reliability
If conventional image stabilization techniques are used, then image blur correction is achieved, but stability deteriorates when perspective distortion is present
Solution Approach 1:
The patent implements a feedback mechanism where the system detects the perspective amount from the captured image and uses this information to adjust the weighting of motion vectors in real-time. The detected perspective amount feeds back into the image stabilization calculation, allowing the system to dynamically adapt its correction strategy based on the actual distortion present in the image.
Solution Approach 2:
The patent makes the image stabilization system dynamic by adjusting the motion vector weights based on the detected perspective amount. Rather than using fixed weights, the system dynamically modifies the weighting scheme according to the actual imaging conditions, allowing it to adapt to varying degrees of perspective distortion and maintain stability.
3Quantity of substance
If motion vectors from peripheral regions are used for image stabilization, then more data is available for calculation, but accuracy decreases due to perspective distortion
Solution Approach 1:
The patent divides the image into different regions (central and peripheral) and applies different quality assessments to motion vectors from each region. Peripheral motion vectors, which are more numerous but more distorted, are given lower weights, while central motion vectors, which are fewer but more accurate, are given higher weights. This approach allows the system to utilize the quantity of peripheral vectors while compensating for their lower quality.
Data Source
AI summary
An image stabilization apparatus includes an acquisition unit configured to acquire information about a lens, a motion vector detection unit configured to detect a motion vector from a plurality of images captured by an image capture unit for capturing an object image, and a calculation unit configured to calculate an image blur correction amount for correcting an image blur, on the basis of the motion vector detected by the motion vector detection unit, and a shake of an image capturing apparatus detected by a shake detection unit for detecting the shake of the image capturing apparatus. The calculation unit changes weights applied to the motion vector and the shake of the image capturing apparatus for calculation of the image blur correction amount, on the basis of information related to perspective of the lens acquired by the acquisition unit.