Lens shading correction using threshold-based gain functions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lens shading correction methods are costly in terms of hardware implementation and often fail to maintain adequate brightness uniformity, especially when the relative illumination of an image is low, as they require significant memory or involve complex calculations.
Innovation Solution
A method involving a threshold-based approach with a first and second gain function, blended to calculate brightness gains for pixels within and outside a threshold range, using second-order polynomial functions to reduce hardware costs and improve brightness uniformity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a high-order polynomial is used for computing gain values, then measurement precision of brightness correction is improved, but device complexity and hardware cost increase
Solution Approach 1:
The image is divided into multiple regions (e.g., center region and peripheral regions), and different gain calculation methods are applied to different regions. The center region uses a simpler gain calculation while peripheral regions use more complex calculations, thereby reducing overall computational complexity while maintaining correction accuracy where needed.
Solution Approach 2:
Different gain functions are applied to different spatial locations in the image. The gain calculation is adapted locally based on the position of pixels, using appropriate polynomial orders for different regions to optimize both accuracy and computational efficiency.
2Ease of operation
If a Mesh Grid method is used for lens shading correction, then ease of operation is improved, but device complexity increases due to significant memory requirements
Solution Approach 1:
The patent extracts only the essential gain values needed for correction rather than storing complete gain maps for every pixel. By calculating and storing only critical gain parameters and computing others on-demand, memory requirements are significantly reduced while maintaining correction effectiveness.
Solution Approach 2:
The approach changes from storing fixed gain values for all pixels to using parametric gain functions that can generate gain values dynamically. This parameter-based approach reduces memory storage needs from storing thousands of gain values to storing just a few polynomial coefficients.
3Device complexity
If a low-order polynomial is used for gain computation, then device complexity is reduced, but measurement precision deteriorates when relative illumination is lower than 30%
Solution Approach 1:
The image is divided into multiple regions (e.g., center region and peripheral regions), and different gain calculation methods are applied to different regions. The center region uses a simpler gain calculation while peripheral regions use more complex calculations, thereby reducing overall computational complexity while maintaining correction accuracy where needed.
Solution Approach 2:
The gain calculation method is made dynamic rather than static. The polynomial order and complexity are adjusted dynamically based on the spatial position and illumination conditions, allowing the system to use simpler calculations in well-lit areas and more complex calculations in low-illumination areas where precision is more critical.
Data Source
AI summary
A lens shading correction method for pixels of an image is provided. The method includes the steps of: inputting coordinates of the pixels and setting a threshold range on the image; providing a first gain function and a second gain function, each relating the coordinates of the pixels to brightness gains; performing the first gain function on the pixels located at the interior of the threshold range for calculating the brightness gains; and performing a blended gain function on the pixels located at the exterior of the threshold range for calculating the brightness gains, wherein the blended gain function is the combination of the first gain function and the second gain function.