Image Correction Circuit Using Non-Linear Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image correction methods for barrel-shaped and pillow-shaped distortions caused by wide-angle and telephoto lenses compromise the viewing angle, are complex, and unsuitable for super wide-angle lenses, with high computational requirements and unnatural image stretching, especially when dealing with super wide-angle lenses.
Innovation Solution
An approximately non-linear regression approach is used to calculate horizontal and vertical ratio parameters for each pixel based on its distance from a reference point, allowing for fast and accurate image correction without triangular functions, preserving the horizontal field of view and simplifying hardware implementation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If de-warping transformation mathematical models are used to perform image correction operations, then barrel-shaped distortion is corrected, but horizontal field of view is lost and viewing angle is compromised
Solution Approach 1:
The patent changes the mathematical model parameters from traditional de-warping transformations to a new correction model that uses different calculation formulas. This allows achieving image correction while preserving the horizontal field of view by altering the fundamental parameters of the correction approach rather than using conventional mathematical transformations.
Solution Approach 2:
The patent replaces the traditional mechanical/mathematical de-warping transformation system with a new computational approach that calculates correction values differently. This substitution enables correction of barrel-shaped distortion without the field of view loss inherent in traditional transformation-based methods.
2Manufacturing precision
If traditional image correction methods are used, then distortion is corrected, but computational complexity increases due to triangular functions and their inverse functions
Solution Approach 1:
The patent extracts and removes the complex triangular functions and their inverse functions from the correction calculation process. By taking out these computationally intensive mathematical operations, the patent achieves image correction with significantly reduced calculation complexity, making the system more suitable for hardware implementation.
Solution Approach 2:
The patent employs simpler, more computationally efficient mathematical operations that can be executed quickly and with fewer resources. This approach uses 'cheaper' computational methods that sacrifice some mathematical complexity to gain efficiency, making the correction process more practical for real-time applications.
3Manufacturing precision
If de-warping transformation is applied to super wide-angle lenses, then barrel-shaped distortion is corrected, but image stretching becomes significant and unnatural
Solution Approach 1:
The patent applies local quality by using different correction calculations for different regions of the image. The correction approach adapts to local characteristics, particularly for super wide-angle lenses, to avoid uniform stretching artifacts. This allows distortion correction while maintaining natural appearance in different parts of the image.
4Ease of operation
If polynomial coefficients are calculated using known input and output images, then image correction can be performed, but the process becomes extremely complicated and requires manual coordinate acquisition
Solution Approach 1:
The patent enables the system to perform self-service by automatically determining correction parameters without requiring manual coordinate acquisition. The correction model can be configured to work automatically with the lens characteristics, eliminating the need for users to manually map coordinates between input and output images.
Solution Approach 2:
The patent performs preliminary action by pre-configuring the correction model with lens-specific parameters before actual image correction is needed. This preliminary setup eliminates the need for complex real-time calculations and manual coordinate work during operation, simplifying the user experience.
Data Source
AI summary
An image correction method arranged for processing an original image to obtain a corrected image includes steps: receiving the original image from an image sensor; regarding each pixel of the original image, calculating a horizontal distance and a vertical distance between the pixel and a reference point in the original image; determining a horizontal ratio parameter and a vertical ratio parameter according to the horizontal distance and the vertical distance between the pixel and the reference point in the original image; and performing an approximately non-linear regression calculation on the horizontal ratio parameter, the vertical ratio parameter and a coordinate of the pixel to obtain a position of the pixel in the corrected image.