Distortion Correction Using Non-Polynomial Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing distortion correction methods for fisheye lens images are inefficient due to the need for numerous correction coefficients and are not suitable for real-time hardware implementation, leading to issues in feature extraction and recognition, especially in vehicle-mounted cameras with broad field of view.
Innovation Solution
A hardware-based apparatus and method using a non-polynomial estimation function and morphing form for distortion correction, which includes an image corrector that estimates image coordinates and brightness, and adjusts correction coefficients to correct distorted images in real-time, utilizing trigonometric functions and arctangent values for efficient hardware operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a polynomial model with multiple correction coefficients is used to correct fisheye lens distortion, then distortion correction accuracy is improved, but computational complexity and hardware implementation difficulty increase
Solution Approach 1:
The patent transforms the polynomial correction model into a parameterization-based model where distortion is corrected by changing parameters (u', v') rather than using multiple correction coefficients. This converts the complex polynomial computation into simpler parameter transformation operations, reducing computational complexity while maintaining correction accuracy.
Solution Approach 2:
The patent replaces the traditional polynomial mathematical model with a parameterization-based geometric transformation model. Instead of using complex polynomial equations with multiple coefficients, the system uses parameter transformations that map distorted coordinates to corrected coordinates through simpler geometric relationships, making hardware implementation more feasible.
2Device complexity
If non-polynomial models with logarithm and tangent functions are used for distortion correction, then computational simplicity is improved, but hardware operability deteriorates due to complex estimation functions
Solution Approach 1:
The patent creates a hardware-friendly version of the non-polynomial model by copying the essential geometric transformation logic while replacing complex estimation functions with simpler operations. The parameterization approach copies the functional relationship between distorted and corrected coordinates while using only basic arithmetic and trigonometric operations that are easily implementable in hardware.
Solution Approach 2:
The patent changes the mathematical parameters from complex estimation functions (logarithm, tangent) to simpler parameter transformations. By using parameterization where distortion correction is achieved through coordinate transformation rather than complex function evaluation, the system maintains computational simplicity while improving hardware operability.
3Measurement precision
If multiple correction coefficients are used in distortion correction algorithms, then correction accuracy is improved, but real-time processing capability deteriorates due to increased calculation amount
Solution Approach 1:
The patent changes the computational approach from evaluating multiple correction coefficients to transforming parameters (u', v'). This parameter transformation approach requires significantly fewer calculations per pixel while achieving the same correction accuracy, enabling real-time processing in hardware implementations.
Solution Approach 2:
The patent segments the distortion correction process into distinct stages: parameter transformation, coordinate mapping, and image rendering. This segmentation allows each stage to be optimized independently and implemented efficiently in hardware, improving real-time processing capability while maintaining correction accuracy.
4Device complexity
If fixed correction coefficients are used in distortion correction, then hardware implementation is simplified, but adaptability to different images deteriorates due to connection disconnection
Solution Approach 1:
The patent introduces dynamic parameter adjustment capability to the distortion correction system. Instead of using fixed correction coefficients, the system uses parameter transformations that can adapt to different image characteristics and distortion levels. This dynamic approach maintains hardware simplicity while significantly improving adaptability to various imaging conditions.
Solution Approach 2:
The patent enables parameter changeability in the distortion correction model. By using parameterization rather than fixed coefficients, the system can dynamically adjust correction parameters based on image content and distortion characteristics. This allows seamless transition between different correction modes (general correction and morphing type correction) while maintaining hardware implementation simplicity.
Data Source
AI summary
With regard to a distortion correction apparatus and method which are applicable to hardware for real-time distortion correction, provided are a distorted image correction apparatus and method for correcting a distortion of an image by using a non-polynomial estimation function suitable for a hardware operation, and for correcting the distortion of the image by utilizing a morphing form.