Driver Assistance Corridor Training for Obstacle-Aware Vehicle Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current driver assistance systems for vehicles lack effective support for navigating specific locations and avoiding obstacles, particularly in variable environments, and do not allow for user-defined trajectories to prevent undesirable paths during automated driving.
Innovation Solution
A training method that captures a sequence of camera images during a user-guided drive to determine a driving corridor, which is then adjusted and stored, allowing for semi-automatic or automatic navigation within this corridor, avoiding obstacles and user-defined paths, even in unfavorable conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a fixed trajectory is used for automated driving, then the driving path is precisely controlled, but the system cannot adapt to variable obstacles or unfavorable environmental conditions
Solution Approach 1:
The patent implements a dynamic driving corridor that can adapt its boundaries and shape based on detected obstacles and environmental conditions. The corridor is not a fixed geometric constraint but a flexible guide that adjusts in real-time, allowing the vehicle to navigate around obstacles while maintaining automated control. This resolves the contradiction by making the trajectory both precise (through controlled boundaries) and adaptable (through real-time adjustment).
Solution Approach 2:
The system changes the parameters of the driving corridor dynamically based on sensor input. When obstacles are detected, the corridor's boundaries, width, and shape parameters are modified to accommodate the new conditions. This allows the automated system to adapt to variable obstacles while maintaining precise control within the updated corridor parameters.
2Reliability
If the driving corridor is strictly enforced, then undesirable paths are prevented, but the system cannot handle unexpected obstacles within the corridor
Solution Approach 1:
The driving corridor is implemented as a dynamic constraint rather than a static boundary. The system continuously monitors for obstacles and adjusts the corridor boundaries in real-time, allowing the vehicle to deviate from the original path when necessary while still preventing truly undesirable paths through the maintained corridor structure. This dynamic approach provides both reliability and adaptability.
Solution Approach 2:
The system performs preliminary detection and planning to establish the driving corridor before execution. By anticipating potential obstacles and pre-defining the corridor boundaries based on available information, the system can prevent undesirable paths while maintaining flexibility to handle unexpected obstacles that arise during execution through real-time adjustments.
3Measurement precision
If user training is required to define driving corridors, then location-specific navigation is improved, but the setup time and complexity increase
Solution Approach 1:
The system performs self-calibration and automatic corridor definition by capturing images during a brief user-driven training drive and processing them to determine the driving corridor. This eliminates the need for extensive manual configuration while still achieving location-specific navigation accuracy. The system serves itself by automatically extracting corridor information from the training data without requiring detailed user input.
Solution Approach 2:
The patent replaces manual, mechanical corridor definition processes with automated image processing and computational analysis. Instead of requiring users to manually map and define corridors, the system captures images during a short training drive and uses algorithms to automatically determine the corridor boundaries and characteristics, significantly reducing setup time while maintaining precision.
4Measurement precision
If extensive image processing is performed during training, then the driving corridor is precisely determined, but the processing time and computational load increase
Solution Approach 1:
The system performs image capture and preliminary processing during a brief user-driven training drive, extracting key features and determining the driving corridor in advance. By completing the computationally intensive image processing before actual automated operation, the system achieves precise corridor determination without imposing continuous computational load during vehicle operation. The preprocessing results are then used for efficient real-time navigation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Training method for a driver assistance system of a vehicle (100), wherein the vehicle (100) comprises at least one camera (105, 106, 107), comprising the following method steps: acquisition (310) of a sequence of camera images during a user-guided training drive of the vehicle (100) along a desired trajectory (400); determination (350) of a driving corridor (500) for the driver assistance system along the desired trajectory (400) depending on image processing of the acquired sequence of camera images; display (370) of the acquired sequence of camera images and/or an environment model determined depending on the acquired sequence of camera images on a display (103, 122), wherein at least the determined driving corridor (500) is displayed as an overlay superimposed on the display;and storage (395) of the driving corridor (500, 600, 700) in an electronic memory of the vehicle, wherein a detection (380) of user input for adjusting a limit of the determined driving corridor (500, 600) during the display, and an adjustment (390) of the determined driving corridor (500, 600) depending on the detected input are carried out.;