Vehicle Under-View Image Reconstruction via Perimeter Camera Stitching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing driver assist systems for motor vehicles do not effectively provide information about the terrain and ground conditions directly under the vehicle, especially in off-road conditions, which can lead to potential hazards such as large rocks, depressions, and slippery surfaces that may cause vehicle damage or injury.
Innovation Solution
A driver assist system that uses perimeter view cameras to capture and store image sequences, processes them to construct an under-view image of the terrain below the vehicle, and displays this information on an interactive display monitor, allowing the driver to see the vehicle's surroundings, including wheel positions and potential hazards, without the need for cameras mounted underneath the vehicle.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If cameras are mounted underneath the vehicle to capture terrain information, then the driver can see ground conditions directly under the vehicle, but the device complexity and installation difficulty increase significantly
Solution Approach 1:
The patent uses perimeter view cameras mounted on the vehicle's perimeter (not underneath) as intermediary devices to capture terrain information. The processing circuitry acts as a mediator to stitch these perimeter images together to reconstruct the under-vehicle terrain view, avoiding the need for direct underneath camera mounting while still providing the required information.
Solution Approach 2:
The system creates a virtual copy of the under-vehicle terrain view by processing and stitching images from perimeter cameras. This virtual terrain image is displayed on the monitor to provide the driver with terrain information without requiring physical cameras mounted in the difficult-to-reach under-vehicle area.
2Loss of information
If multiple perimeter view cameras are used to capture terrain information, then the driver can see surrounding terrain, but the device complexity and cost increase
Solution Approach 1:
The perimeter view cameras serve multiple functions: they capture both the surrounding environment for standard perimeter view display and simultaneously provide the data needed to construct the under-vehicle terrain view. This multi-functionality reduces the need for additional dedicated cameras while maximizing the utility of each camera installed.
Solution Approach 2:
The patent merges the function of perimeter view capture and under-vehicle terrain capture into a single camera system. The same perimeter cameras that provide 360-degree surrounding views are also used to reconstruct terrain information, combining multiple functions into one camera system to reduce overall device complexity.
3Speed
If the processing circuitry processes image sequences in real-time to construct under view images, then the driver receives immediate terrain information, but the processing time and computational load increase
Solution Approach 1:
The system performs preliminary processing of the image sequences from perimeter cameras, pre-calculating and storing processed image data that can be quickly assembled into under-vehicle terrain views when needed. This preliminary processing reduces the real-time computational burden during actual terrain display generation.
Solution Approach 2:
The processing circuitry processes only the necessary portions of the image sequences required to construct the under-vehicle terrain view, rather than processing entire 360-degree panoramas. This selective partial processing reduces computational load while still providing complete terrain information under the vehicle.
Data Source
AI summary
A driver assist system and method for a motor vehicle that includes processing circuitry configured to receive and store a sequence of images from perimeter view cameras, select at least a portion of a previously stored image based on an estimated position displacement of the motor vehicle, construct an under view image of at least a portion of the motor vehicle, and display the under view image on a display monitor.