Vehicle Surround-View Imaging for Backlight Distortion Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Vehicle image distortion caused by backlight sources can hinder a driver's ability to perceive obstacles and distances, especially when camera angles correspond to backlight environments, leading to potential traffic accidents.
Innovation Solution
A vehicle image processing device that determines the presence of backlight by analyzing shadow recognition and optical property information from multiple cameras, calculating light source location, and generating a display image that excludes light bleed areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple cameras are used to expand driver's field of view, then the driver can perceive areas difficult to identify with eyes, but image distortion occurs when camera shooting direction corresponds to backlight
Solution Approach 1:
The system uses shadow recognition as an intermediary indicator to indirectly detect backlight presence. Instead of directly analyzing complex image distortion caused by backlight, the system detects the shadow cast by the vehicle on the ground, which serves as a reliable mediator to infer backlight conditions and trigger appropriate image processing corrections.
Solution Approach 2:
The system replaces direct optical analysis of distorted images with a computational approach based on shadow detection and geometric calculations. By using coordinate-based shadow analysis and light source location calculation, the system substitutes complex optical distortion analysis with mathematical computations to determine backlight presence and generate corrected display images.
2Reliability
If shadow recognition is used to detect backlight presence, then backlight can be immediately determined for each camera, but additional processing steps are required
Solution Approach 1:
The system uses the vehicle's own shadow, which is naturally present in the camera images, as the detection target. Instead of requiring external detection devices or additional sensors, the system leverages the shadow that already exists in the captured images, turning the image data itself into the detection resource and eliminating the need for separate detection hardware.
Solution Approach 2:
The shadow recognition system serves multiple functions: it detects backlight presence, determines light source location, and provides information for image correction. By using the shadow as a universal indicator, the system achieves multiple detection objectives through a single analysis process, reducing the need for separate detection mechanisms for each camera.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution effectively reduces image distortion, enhancing the driver's field of view and safety by accurately determining backlight presence and eliminating light bleed areas from the vehicle's periphery images.
Implementation Method 1
the processor may calculate location information of a light source for at least one of the first camera or the vehicle by using coordinates of a shadow object of the vehicle, which is recognized from the vehicle periphery image, and coordinates of the vehicle
Data Source
AI summary
The present disclosure relates to a vehicle image processing device and a method therefor. A vehicle image processing apparatus may include a storage that stores optical property information of a first camera among a plurality of cameras for obtaining a vehicle periphery image, a processor that determines whether backlight is present in the vehicle periphery image and generates a display image based on whether the backlight is present, and a communication device controlled by the processor and communicating with a device in the vehicle. The processor may calculate location information of a light source for at least one of the first camera or the vehicle by using coordinates of a shadow object of the vehicle, which is recognized from the vehicle periphery image, and coordinates of the vehicle, and may determine whether the backlight is present, by comparing location information of the light source with the optical property information.