Multi-Lane Road Geometry Estimation With Bayesian Hypothesis Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current road estimation systems for autonomous vehicles face challenges in accurately modeling complex road geometries, especially in scenarios with multiple lanes, on-ramps, and off-ramps, due to limitations in data association and clutter handling, which affects the reliability and redundancy of road feature detection.
Innovation Solution
A model-based algorithm using a Bayesian approach with a multiple hypothesis tracking filter that predicts road model hypotheses based on vehicle motion data and perception data from sensors, associating road features to estimate lane center curves and existence probabilities, and selecting the most probable road model for accurate road geometry estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a traditional road estimation system is used, then the system complexity is low, but the measurement precision of road geometry in complex environments deteriorates
Solution Approach 1:
The road model is segmented into multiple hypotheses, each representing a different possible road configuration. The tracking filter maintains multiple road model hypotheses simultaneously, allowing the system to evaluate different road geometry possibilities independently and select the most probable one, thereby improving measurement precision without overwhelming system complexity.
Solution Approach 2:
The road estimation system dynamically adapts by continuously updating the probability of each road model hypothesis based on new sensor measurements and vehicle motion data. The system transitions between different road model hypotheses as the vehicle encounters complex road geometries, enabling accurate real-time estimation while maintaining manageable computational complexity through probabilistic filtering.
2Reliability
If multiple road model hypotheses are tracked, then the reliability of road estimation improves, but the device complexity increases
Solution Approach 1:
The system tracks multiple road model hypotheses beyond what a single deterministic model would provide, maintaining a set of probable road configurations with associated probabilities. This partial multiplicity approach improves reliability by considering alternative road geometries without fully enumerating all possible configurations, balancing reliability improvement with acceptable filter complexity.
Solution Approach 2:
The tracking filter incorporates feedback mechanisms where each new sensor measurement updates the probability distribution across multiple road model hypotheses. The system uses feedback from vehicle motion data and sensor associations to reinforce or eliminate hypotheses, improving reliability through continuous validation while managing complexity through probabilistic convergence.
3Adaptability or versatility
If real-time road model updating is performed, then the adaptability to changing conditions improves, but the loss of time for computation increases
Solution Approach 1:
The system performs preliminary actions by pre-defining the structure of road model hypotheses and their associated probability distributions before actual road estimation is needed. The tracking filter is pre-configured with multiple possible road configurations, allowing real-time adaptation to changing conditions without the computational overhead of generating hypotheses on-the-fly, thus reducing computation time while maintaining high adaptability.
Solution Approach 2:
The system achieves real-time adaptability by changing parameters within the existing road model hypotheses rather than reconstructing entire models. The tracking filter adjusts probability distributions and geometric parameters of predefined hypotheses based on incoming sensor data, enabling rapid adaptation to changing road conditions while minimizing computation time through parameter optimization rather than full model regeneration.
Data Source
Figure 1~2(d)
Figure 3a~3b
Figure 4
AI summary
A computer-implemented method for estimating a road geometry of a road upon which a vehicle is traveling and related aspects are disclosed. The disclosed embodiments provide a model-based technique for estimating a road model having an arbitrary number of lanes on a road as well as any entrance lanes (i.e. on-ramp or slip roads) and exit lanes (i.e. off ramps or off-slip roads). In more detail, the herein disclosed embodiments utilize a Bayesian approach to multi-lane tracking in various traffic scenarios, such as e.g. highway scenarios. The employed model-based algorithm estimates the lane center curves of multi-lane roads based on vehicle motion data (e.g. speed, angular velocity) and perception data (e.g. camera output, lidar output, radar output, etc.) that comprises detected road features (e.g. lane markers, road edges, road barriers, guard rails, road markers, etc.) or other objects (e.g. other road users).