Lens Distortion Correction Using Curve-Based Outlier Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing technologies face challenges in effectively correcting lens distortion, especially in wide-angle lenses, as they often require additional information or specific camera settings, and struggle with accurately identifying the distortion center, leading to incomplete or new distortions during correction.
Innovation Solution
An image processing apparatus and method that detects line segments, generates curves, applies weights to a cost function based on distance from the image center, removes outliers, and estimates a distortion level parameter using the RANSAC algorithm and Levenberg-Marquardt method to correct pixel positions, enabling automatic lens distortion correction without additional information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional distortion correction methods are used, then correction can be performed with simple algorithms, but accuracy deteriorates because they cannot accurately identify the distortion center and require additional information
Solution Approach 1:
The system uses the image content itself to automatically identify the distortion center and correction parameters without requiring external calibration patterns or additional information. The line segment detection and curve generation processes utilize only the input image data to self-determine the correction methodology.
Solution Approach 2:
The patent replaces traditional mechanical/optical correction methods (such as using distortion correction lenses or calibration targets) with computational image processing methods. The distortion correction is achieved through algorithmic analysis of image line segments and curves rather than physical intervention.
2Area of stationary object
If wide-angle lenses are used to capture large areas, then field of view increases, but distortion becomes severe especially at edges
Solution Approach 1:
The system dynamically adjusts correction parameters based on the detected distortion characteristics in different regions of the image. The correction amount varies according to the distance from the distortion center and the local curvature of detected line segments, allowing optimized correction for wide-angle distortions while preserving field of view.
3Manufacturing precision
If multiple lenses are attached to reduce distortion, then distortion correction improves, but device complexity and cost increase
Solution Approach 1:
The patent replaces complex mechanical lens assemblies designed for distortion correction with a computational approach. Instead of using multiple lenses or specialized optical elements, the system achieves distortion correction through image processing algorithms that analyze and rectify geometric distortions in the captured image.
4Ease of operation
If automatic correction without additional information is implemented, then ease of operation improves, but reliability deteriorates due to difficulty in accurately identifying distortion characteristics
Solution Approach 1:
The system employs iterative feedback mechanisms where the detected line segments and generated curves are used to refine the distortion center identification and correction parameters. The cost function minimization process continuously adjusts parameters based on the geometric relationships observed in the image, improving reliability through self-validation.
Data Source
AI summary
Provided are an apparatus and method for processing an image. The apparatus includes a detector configured to form one group with pieces of pixel information on an edge of a received image, detect line segments, and generate a plurality of curves using the detected line segments, an optimizer configured to apply a weight to a cost function based on the plurality of generated curves according to distance from the center of the image, define a curve having a shorter length than a threshold value as an outlier and a curve having a greater length than the threshold value as an inlier, remove the outlier, and estimate a distortion level parameter, and a corrector configured to correct a position of a pixel with a correction value using the distortion level parameter. Accordingly, it is possible to correct lens distortion of even an image that has not been directly photographed, and to correct images generated by devices equipped with any lens rather than a specific camera or photographing apparatus.