Lens Image Distortion Correction via Segmented Unit Shifting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for correcting image distortion in lenses with wide angles, such as those used in automotive cameras, require complex calculations and large memory capacity, leading to high implementation complexity and limited distortion correction capability.
Innovation Solution
An apparatus that segments image information into units, shifts them by a variation ratio relative to a set center, and inserts these units into interpolation spaces to correct distortion, minimizing operation and memory logic while maximizing correction capability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If mathematical model is used for distortion correction, then correction can be applied to all pixels, but implementation complexity is high and large memory capacity is required
Solution Approach 1:
The patent divides the image into multiple image subdivision units (e.g., quadrants or blocks) and processes each unit independently. This segmentation approach reduces the complexity of processing the entire image at once, as each unit can be handled with simpler operations while collectively achieving comprehensive distortion correction across the full image.
2Manufacturing precision
If Cartesian coordinate pattern is used for distortion correction, then distortion can be corrected by shifting pixels, but capability is limited when using only some pixels for relationship information
Solution Approach 1:
The patent applies different correction methods to different regions of the image based on their specific distortion characteristics. Each image subdivision unit is processed according to its local distortion properties, allowing for optimized correction in each region rather than applying a uniform approach. This enables accurate distortion correction while reducing the need to store and process all pixel values.
3Manufacturing precision
If all pixel values before and after correction are stored, then correction can be performed, but large memory capacity is required
Solution Approach 1:
By segmenting the image into smaller units, the patent reduces the amount of data that needs to be stored and processed simultaneously. Instead of storing all pixel values for the entire image, only the necessary pixel values for each subdivision unit are retained, significantly reducing memory requirements while maintaining correction accuracy.
Solution Approach 2:
The patent extracts only the essential pixel information needed for distortion correction in each image subdivision unit, rather than retaining all pixel values. This selective extraction approach minimizes memory usage by discarding redundant information while preserving the critical data required for accurate distortion correction.
Data Source
AI summary
Disclosed herein is an apparatus for correcting image distortion of a lens, including an input unit for receiving image information, a segmentation unit for segmenting the image information into one or more image subdivision units, a setting unit for setting a center of the image information, a controller for shifting each of the one or more image subdivision units by a variation ratio with respect to the center of the image information, an interpolation unit for inserting one or more image subdivision units corresponding to the shifted image subdivision units into an interpolation space generated between the shifted image subdivision units, and an output unit for outputting corrected image information. According to an embodiment of the present invention, by shifting one or more image subdivision units by a variation ratio, operation logic for coordinate calculation and internal memory logic may be minimized, and implementation complexity may be lowered.