Vehicle Pose Assessment Using Map Polyline Deviation Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle localization methods, particularly for autonomous vehicles, face challenges in accurately determining pose due to inaccuracies in satellite-based positioning systems and sensor noise, leading to positioning errors and unreliable vehicle pose estimation, especially in scenarios with poor satellite connections.
Innovation Solution
A vehicle pose assessment system that predicts vehicle pose using sensor data, transforms digital map road references into a selected coordinate system, identifies corresponding sensor-captured features, projects these features onto polyline segments, determines deviation parameters based on projection distances, and combines these parameters to calculate path deviations, thereby aligning sensor measurements with digital map elements for improved localization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If satellite-based positioning systems (GNSS) are used for vehicle positioning, then positioning coverage is provided, but positioning accuracy deteriorates with errors of several meters
Solution Approach 1:
The patent combines multiple positioning methods (GNSS, IMU, visual odometry, map matching) into a unified pose assessment system. The system integrates satellite-based positioning with inertial measurement unit data and visual odometry results to compensate for the limitations of each individual method, achieving both high accuracy and reliability in vehicle pose determination
Solution Approach 2:
The patent creates a composite positioning solution by fusing data from heterogeneous sources (satellite signals, inertial sensors, camera data, digital map information). This composite approach leverages the strengths of each data source while mitigating their individual weaknesses, producing a robust pose estimation that maintains accuracy across diverse operating conditions
2Measurement precision
If visual odometry is used for vehicle pose estimation, then positioning accuracy can be improved, but measurement noise from sensors causes unreliable pose prediction
Solution Approach 1:
The patent implements a feedback mechanism where the pose assessment system continuously evaluates the reliability of pose predictions and uses this information to adjust its operation. When measurement noise is detected or reliability drops below a threshold, the system requests updated digital map data or switches to alternative positioning methods, ensuring consistent and reliable pose estimation
Solution Approach 2:
The patent employs dynamic switching between different positioning methods based on current operating conditions and data quality. The system adaptively selects and combines GNSS, IMU, and visual odometry approaches according to their current reliability levels, maintaining robust pose estimation across varying environmental conditions
3Measurement precision
If landmark-based positioning is used to improve pose accuracy, then positioning precision increases, but system complexity increases due to multiple sensor integrations
Solution Approach 1:
The patent designs a multi-functional pose assessment system where a single integrated system performs multiple positioning functions using different methods. The system can operate with GNSS alone, visual odometry alone, or in combination, and can also perform map matching and reliability assessment, reducing the need for separate dedicated systems for each function
4Reliability
If GNSS and IMU are combined for positioning, then positioning coverage is maintained, but large scale and bias errors result in several meters positioning error
Solution Approach 1:
The patent introduces digital map data and visual odometry as intermediary references to correct the scale and bias errors inherent in GNSS-IMU combinations. By comparing sensor measurements against known map features and visual landmarks, the system can detect and correct cumulative drift and systematic errors, maintaining both coverage and accuracy
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure relates to a method performed by a vehicle pose assessment system (1) for supporting determining a pose of a vehicle (2) in view of a digital map (22). The vehicle pose assessment system predicts (1001) a pose of the vehicle based on sensor data acquired by a vehicle localization system (23); transforms (1002) to a selected coordinate system (3) a set of map road references of a portion of the digital map based on the predicted pose of the vehicle, wherein the transformed set of map road references form a set of polylines in the selected coordinate system, which set of polylines forms a set of polyline paths respectively comprising segments (SEG1-SEG10) of polylines; identifies (1003) a set of corresponding sensor-captured road reference features (S1-S11) acquired by a vehicle-mounted surrounding detecting device (24), each identified road reference feature defining a set of measurement coordinates in the selected coordinate system; projects (1004) each of the identified set of road reference features onto the polyline segments in order to obtain a set of projection points (P1-P11), wherein each projection point defines a set of projection coordinates; determines (1004) for each polyline segment, deviation parameters in view of each identified road reference feature, based on a projection distance (D1-D11) between respective road reference feature's measurement coordinates and its corresponding polyline segment projection coordinates, wherein for each polyline segment onto which one or more road reference features are having deviations fulfilling deviation criteria, the polyline segment is assigned predeterminable deviation parameters in view of those one or more road reference features; and determines (1006) by combining the deviation parameters of respective polyline path's polyline segments, a respective path deviation for each polyline path. The disclosure also relates to a vehicle pose assessment system in accordance with the foregoing, a vehicle comprising such a vehicle pose assessment system, and a respective corresponding computer program product and non-volatile computer readable storage medium.