Fisheye Image Distortion Correction for Road Mirror Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fisheye camera images used for driving assistance and automated driving have significant distortion, limiting their effective use for advanced recognition processing, as only a central portion of the image can be utilized after distortion correction.
Innovation Solution
An image processing apparatus that performs distortion correction on partial regions of the fisheye camera image based on a rule, with the ability to adjust this rule when a road mirror is detected, prioritizing distortion correction for regions corresponding to specific directions of the vehicle.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If distortion correction processing is applied to the entire captured image with the center as the correction center, then the central region achieves good distortion correction, but the distortion increases in portions away from the center
Solution Approach 1:
The captured image is divided into multiple partial regions (first partial region, second partial region, third partial region, fourth partial region) with different correction centers. Each region is corrected independently using its own correction center, allowing the central regions to be corrected with high accuracy while peripheral regions maintain their own local correction quality without being degraded by global correction.
Solution Approach 2:
Different partial regions are assigned different correction centers tailored to their specific locations. The first and second partial regions use one correction center, while the third and fourth partial regions use another correction center. This local customization of correction parameters ensures that each region achieves optimal distortion correction without introducing artifacts or increased distortion in other areas.
2Manufacturing precision
If only the central portion of the fisheye camera image is used for recognition processing, then distortion correction accuracy is improved, but the effective utilization of the extensive captured image is reduced
Solution Approach 1:
The image processing apparatus processes multiple partial regions simultaneously with different correction centers, enabling the entire captured image to be effectively utilized for recognition processing. Each segmented region contributes to the overall recognition accuracy without being limited to only the central portion, thereby maximizing the productive use of the extensive fisheye camera coverage.
3Productivity
If distortion correction is applied to partial regions with different correction centers, then the effective use of extensive captured image is improved, but the system complexity increases
Solution Approach 1:
The captured image is divided into a first partial region and a second partial region (and optionally third and fourth partial regions), with each region assigned a specific correction center. This segmentation allows the system to process multiple regions in parallel with dedicated correction parameters, improving effective image utilization while managing complexity through structured regional division.
Solution Approach 2:
Different correction centers are assigned to different partial regions based on their spatial locations. The first and second partial regions share one correction center, while the third and fourth partial regions share another correction center. This localized approach optimizes correction quality for each region while maintaining systematic organization that prevents excessive complexity growth.
Data Source
AI summary
An image processing apparatus performs external recognition for driving assistance or automated driving of a vehicle based on a captured image obtained from a photographing apparatus capturing an image requiring distortion correction. The image processing apparatus: moves and sets a partial region to be a target of distortion correction processing according to a rule, and applies the distortion correction processing to the set partial region of the captured image to acquire a partial image corrected in distortion; detects a road mirror around the vehicle; and changes the rule so that a partial region corresponding to a predetermined direction of the vehicle is preferentially set as a target of the distortion correction processing when a road mirror is detected.