Multi-Camera Image Recognition With Selective Distortion Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems for vehicles face challenges in balancing image recognition accuracy and processing speed when using multiple cameras with different angles, particularly when combining high-resolution and wide-angle views, leading to increased complexity and processing load.
Innovation Solution
An image processing system with multiple imaging units capturing both low-distortion and high-distortion regions, utilizing separate image recognition units for partial and wider regions, and an integration processing unit to combine results, optimizing recognition accuracy and speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If distortion correction is performed on images captured by fisheye lenses, then image recognition accuracy is improved, but processing load increases and processing time lengthens
Solution Approach 1:
The patent divides the image processing into two distinct paths: a first image recognition unit that processes images without distortion correction for quick results, and a second image recognition unit that performs distortion correction for higher accuracy. This segmentation allows the system to balance speed and accuracy by processing different regions or images through different pipelines.
Solution Approach 2:
The patent applies distortion correction selectively rather than to all images. The second image recognition unit performs distortion correction on specific images or regions where high accuracy is critical, while the first image recognition unit handles other images without correction. This partial application of distortion correction reduces overall processing load while maintaining accuracy where needed.
2Area of stationary object
If multiple cameras are mounted to achieve both high resolution and wide angle coverage, then imaging coverage and resolution are improved, but device complexity increases
Solution Approach 1:
The patent makes a single imaging unit perform multiple functions by capturing both high-resolution images for electronic rearview mirror display and wide-angle images for surrounding situation detection. The imaging unit is designed to capture images with different characteristics (high-resolution and wide-angle) simultaneously or alternately, eliminating the need for separate camera systems for each function.
Solution Approach 2:
The patent merges the functions of multiple cameras into a single integrated system. Instead of using separate cameras for electronic rearview mirror and surrounding detection, the patent combines these functions into one imaging unit that can capture both types of images, thereby reducing device complexity while maintaining comprehensive imaging coverage.
3Measurement precision
If distortion correction is performed on multiple images from multiple cameras, then overall recognition accuracy is improved, but processing load becomes excessively large
Solution Approach 1:
The patent segments the image processing workload by creating two separate image recognition units with different processing requirements. The first unit processes images without distortion correction, while the second unit handles distortion correction. This segmentation distributes the processing load across different units and allows parallel processing, reducing the overall computational burden compared to correcting all images in a single pipeline.
Solution Approach 2:
The patent applies distortion correction selectively to only those images or image regions where it is most beneficial, rather than uniformly correcting all images from all cameras. The second image recognition unit performs distortion correction on specific images or regions where high accuracy is critical, while the first unit handles other images without correction, thereby reducing the total processing load while maintaining accuracy where it matters most.
Data Source
AI summary
An image processing system includes a plurality of imaging units configured to capture optical images including a low-distortion region and a high-distortion region; a first image recognition unit configured to perform image recognition on at least a partial region out of image data obtained from the imaging unit and output a first image recognition result; a second image recognition unit configured to perform image recognition on image data in a wider region than the partial region out of the image data obtained from at least one of the imaging units and output a second image recognition result; and an integration processing unit configured to output an image recognition result integrated on the basis of the first image recognition result and the second image recognition result.