Image Noise Reduction via Alignment Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image noise reduction methods struggle with geometric misalignment between images, leading to blur and reduced sharpness in combined images, especially when capturing scenes with motion.
Innovation Solution
A method involving the application of transformations to images using multiple kernel tracking and Lucas Kanade Inverse algorithms to align them with a reference image, followed by weighting and combining these aligned images to form a reduced noise image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple images are combined to reduce noise, then noise reduction is improved, but geometric misalignment causes blur and reduces image sharpness
Solution Approach 1:
The patent applies alignment transformations to images before combining them. The system determines geometric transformations (translation, rotation, scaling) for each image relative to a reference image, and applies these transformations in advance to correct misalignment. This preliminary alignment action ensures that when images are subsequently combined for noise reduction, they are properly registered, preventing blur and maintaining image sharpness while still achieving noise reduction benefits
Solution Approach 2:
The patent uses alignment metrics to measure the degree of misalignment between images and employs optimization algorithms that iteratively adjust transformation parameters to minimize misalignment. The system calculates alignment quality metrics, compares them against targets, and uses this feedback to refine the geometric transformations applied to each image, ensuring optimal alignment before combination while maintaining both noise reduction and image sharpness
2Measurement precision
If images are captured in quick succession for noise reduction, then noise reduction effectiveness is improved, but motion between images causes misalignment and blur
Solution Approach 1:
The patent performs alignment transformations on images captured in quick succession before combining them. The system detects motion between rapidly captured images and applies corrective geometric transformations (translation, rotation, scaling) in advance to register them with a reference image. This preliminary alignment action enables effective noise reduction from quickly captured images while compensating for motion-induced misalignment, maintaining image sharpness
Solution Approach 2:
The patent dynamically adjusts alignment transformations based on the specific motion characteristics observed between each pair of rapidly captured images. The system calculates individual transformation parameters for each image based on its relative motion, allowing flexible adaptation to varying motion conditions while maintaining proper geometric alignment across all images for effective noise reduction
Data Source
AI summary
A reduced noise image can be formed from a set of images. One of the images of the set can be selected to be a reference image and other images of the set are transformed such that they are better aligned with the reference image. A measure of the alignment of each image with the reference image is determined. At least some of the transformed images can then be combined using weights which depend on the alignment of the transformed image with the reference image to thereby form the reduced noise image. By weighting the images according to their alignment with the reference image the effects of misalignment between the images in the combined image are reduced. Furthermore, motion correction may be applied to the reduced noise image.