Vehicle Surround Image Display Device with Deformed Side Views
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for generating a vehicle surrounding image display fail to create a realistic image of the circumstances behind a vehicle from multiple camera perspectives, leading to misalignment and distortion of road features and other vehicles.
Innovation Solution
A vehicle surrounding image display device that uses a central camera and offset cameras, with an image deformation unit adjusting the images to align vanishing points and reduce distortion, generating a synthetic image that simulates a single virtual viewpoint by inclining and resizing images to match scales and orientations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If images from multiple cameras are combined without deformation to match vanishing points, then the combination process is simple, but white lines on the road surface are misaligned at boundaries due to parallax
Solution Approach 1:
The patent applies image deformation to change the geometric parameters of images captured by offset cameras. Specifically, it inclines the inner side portions of images lower than the vanishing point to make the lower portion located more outward than the upper portion, with gradually increased inclination toward the lower end. This parameter transformation corrects parallax-induced misalignment of road features while maintaining a relatively simple processing framework.
2Manufacturing precision
If the center camera image is compressed horizontally at the lower portion to align boundaries, then road line continuity is improved, but images of vehicles directly behind are significantly distorted
Solution Approach 1:
The patent applies different deformation strategies to different regions of the images. For offset cameras, only the inner side portions below the vanishing point are inclined, while other regions remain unchanged. This localized deformation approach ensures road line continuity at boundaries while preserving the natural appearance of vehicles directly behind, avoiding excessive distortion in critical areas.
3Manufacturing precision
If the side camera images are deformed with oblique inclination to align boundaries, then road line continuity is improved, but images of vehicles diagonally behind are significantly distorted
Solution Approach 1:
The patent deforms only specific local regions of side camera images - specifically the inner side portions below the vanishing point - while leaving other regions including areas capturing diagonally positioned vehicles unchanged. This selective local deformation achieves road line continuity without distorting vehicles in diagonal positions.
4Reliability
If images are deformed to achieve smooth road line continuity, then visual realism is improved, but processing complexity and distortion risk increase
Solution Approach 1:
The patent employs controlled parameter changes through specific deformation operations - inclining inner side portions below vanishing points with gradually increased inclination toward the lower end. This targeted parameter transformation achieves smooth road line continuity and enhanced visual realism while maintaining manageable processing complexity through a well-defined deformation rule.
Data Source
AI summary
A vehicle surrounding image display device according to the present invention combines an image obtained by reducing in size an image taken by a camera provided at the center of the rear side of a vehicle, and images obtained by deforming images taken by cameras provided at the left and right sides of the vehicle such that the outward inclination of the inner side in the vehicle width direction of each of the left and right side images is gradually increased toward the lower end of the inner side. The vehicle surrounding image display device then generates a single image simulating the circumstances behind the user's own vehicle as observed from a single virtual viewpoint in a realistic form, and displays the generated image on a display device.