Virtual Boundary Projection for Vehicle Maneuvering Assistance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing maneuvering assistance systems for vehicles, particularly those using reversing cameras, fail to display boundary lines or structures delimiting a parking space effectively at the end of the maneuver, and surround view systems are costly due to multiple cameras.
Innovation Solution
A method that detects areas outside the current camera view, calculates the position of detected elements, and inserts them as virtual elements into subsequent images, allowing for precise alignment with parking structures using a single or dual camera setup, reducing costs by eliminating the need for multiple cameras.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a reversing camera is used to provide maneuvering assistance, then the system cost is reduced, but the ability to display boundary lines and structures outside the current camera view is lost
Solution Approach 1:
The system captures and stores images at multiple predetermined positions before the vehicle reaches the final parking position. By pre-capturing images at these positions, the system ensures that boundary lines and structures are recorded while still visible, allowing them to be displayed even when they move out of the current camera view during the final maneuvering stages.
Solution Approach 2:
The system creates virtual copies of boundary lines and structures by detecting them in stored images and rendering them as virtual elements in the current camera view. These virtual boundary lines are superimposed on the live camera feed, allowing the driver to see parking space boundaries even when they are outside the current camera field of view, effectively copying visual information from previous positions to the current display.
2Loss of information
If multiple cameras are used to provide a surround view, then the ability to display all-around views and boundary lines is improved, but the system cost increases
Solution Approach 1:
The system transitions from a single two-dimensional camera view to a multi-dimensional solution by capturing images at multiple positions along the vehicle's path. This creates a temporal and spatial dimension to the camera data, allowing the system to reconstruct and display boundary lines that would otherwise be invisible in any single static view, achieving surround-view functionality with a single camera.
Solution Approach 2:
The system introduces virtual boundary lines as an intermediary element between the physical camera and the driver's perception. These virtual lines act as a mediator that bridges the gap between what the single camera can physically capture and what the driver needs to see for safe maneuvering, effectively creating the perception of multiple views without requiring multiple physical cameras.
3Device complexity
If the camera captures only the current area, then the image processing is simplified, but the ability to track and display moving boundary lines is lost
Solution Approach 1:
The system maintains continuous tracking of boundary lines by processing a sequence of images captured at predetermined positions. Instead of treating each image independently, the system continuously detects boundary lines across multiple images and maintains their positions, creating an unbroken visual track of boundary lines throughout the vehicle's movement, which simplifies the processing compared to attempting to track lines across complex multi-camera feeds.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method for manoeuvring a vehicle and to a manoeuvring assistance system, which facilitates precise orientation of the vehicle on existing structures which delimit a parking space and can be implemented cost-effectively. According to the invention, at a first time a first region is sensed by means of a first image and at least one first element within the first region in the first image is detected, and at a second time a second region is sensed by means of a second image and the position of the first element detected in the first image at the first time is calculated with regard to the second time, the first element being inserted and displayed at this calculated position into the second image as a virtual first element.