Surround-View Underside Imaging Without Shadow Distortion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Panoramic imaging systems in vehicles often reconstruct vehicle underside images with shadows and distorted shadows due to affine transformations, affecting the driver's observation and judgment of the vehicle's underside environment.
Innovation Solution
A method and system for generating a surround-view image that intercepts shadow-free images from specific regions, using filling ranges and affine transformations based on vehicle operating parameters to reconstruct shadow-free underside images, and optionally includes a transparent vehicle model for enhanced visibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If affine transformation is performed on the acquired image containing shadows to reconstruct the underside image, then the complete surround-view image covering the vehicle underside can be generated, but the reconstructed underside image will contain shadows and distorted shadows affecting driver observation
Solution Approach 1:
The patent segments the image processing into distinct regions: the first filling range covers the vehicle underside region, the second filling range covers the shadow region, and the image interception region is positioned between them. By segmenting the processing areas, the patent enables selective handling of shadow-free regions versus shadow-containing regions, allowing the underside image to be reconstructed without shadow interference while maintaining complete coverage.
Solution Approach 2:
The patent extracts the shadow-free image data from the image interception region, which is positioned between the first and second filling ranges. This extracted shadow-free image is then used to reconstruct the underside image, effectively taking out the useful shadow-free information from the broader environment image while leaving the shadow regions separate and excluded from the underside reconstruction.
2Object-affected harmful factors
If the image interception region is positioned outside the first filling range to avoid shadows, then shadow-free underside images can be reconstructed, but the time delay between first moment and second moment may affect image accuracy
Solution Approach 1:
The patent performs preliminary action by capturing the environment image at a first moment and identifying the image interception region that is free from shadows before the vehicle moves significantly. The image interception region is predetermined based on the first filling range and second filling range positions. This preliminary identification allows the system to use the shadow-free image data for reconstruction even at a second moment after vehicle movement, compensating for the time delay.
Solution Approach 2:
The patent applies dynamics by adjusting the image interception region based on vehicle operating parameters and movement. The system dynamically determines the appropriate image interception region at different moments, allowing the region to adapt to vehicle movement while maintaining its shadow-free property. This dynamic adjustment enables accurate reconstruction despite time delays between capture and display moments.
Data Source
Figure 1A~1B
Figure 1C~1D
Figure 2
AI summary
Provided in the present disclosure are a method, device and system for generating a surround-view image, and a computer program product and a motor vehicle. An image can be captured from an image capture region where there is no shadow, and the image can be used to reconstruct an image which covers a second filling range corresponding to a first filling range, that is, also covering a region corresponding to the vehicle bottom of a motor vehicle. Since there is no shadow in an image which is captured from an image capture region, after affine transformation is performed on the image, a reconstructed image, which covers the second filling region and comprises an under-vehicle image, does not comprise a shadow. Therefore, the observation and determination of under-vehicle-related information by a driver of a motor vehicle being affected due to a reconstructed under-vehicle image of the vehicle containing a lot of shadows and distorted shadows is effectively prevented.