3D Pose Estimation for Navigation Using HD Map Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems rely on 2D map matching, which can be inaccurate and require additional sensors like LiDAR, and face challenges in providing precise initial navigation parameters for sensor fusion, leading to delayed convergence on true values.
Innovation Solution
A navigation apparatus that uses GPS, steering, and velocity data to generate 3D pose, velocity, and position parameters through line fitting, RANSAC, and dead reckoning, eliminating the need for expensive sensors like LiDAR by leveraging high-definition map data and sensor fusion for accurate initial information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 2D map matching is used with GPS data, then the navigation system can provide basic position information, but the accuracy is insufficient and additional sensors like LiDAR are required
Solution Approach 1:
The patent transitions from traditional 2D map matching to 3D pose estimation by incorporating vertical dimension information from high-definition maps. The system uses 3D map elements (buildings, terrain, road surfaces) to constrain and refine position estimation, adding dimensional information that significantly improves accuracy without requiring additional physical sensors like LiDAR.
Solution Approach 2:
The system changes the parameter space from 2D coordinates (latitude, longitude) to 3D pose parameters (position x, y, z, roll, pitch, yaw). By estimating six-degree-of-freedom pose parameters instead of just 2D position, the system achieves higher precision while utilizing existing GPS and sensor data more effectively.
2Productivity
If traditional map matching methods are used, then the system can operate with basic sensors, but convergence on true values is delayed
Solution Approach 1:
The system performs preliminary 3D pose estimation using high-definition map constraints before final position determination. By pre-establishing 3D spatial relationships from HD maps and using them to guide the estimation process, the system accelerates convergence to accurate position values compared to traditional incremental 2D matching methods.
Solution Approach 2:
The system implements feedback mechanisms where 3D pose estimation results are continuously refined by comparing predicted sensor measurements with actual measurements. The high-definition map provides a reference framework that feeds back into the estimation process, correcting deviations and accelerating convergence to true position values.
3Measurement precision
If 3D pose parameters are estimated using line fitting and RANSAC, then precise navigation information is obtained, but computational complexity increases
Solution Approach 1:
The patent segments the 3D pose estimation problem into distinct computational stages: (1) identifying 3D map elements and their spatial relationships, (2) performing line fitting on subsets of map elements to estimate initial pose, (3) applying RANSAC to filter outliers and refine estimates, and (4) final pose parameter calculation. This segmentation reduces computational complexity by breaking down the complex optimization problem into manageable steps.
Solution Approach 2:
The system uses high-definition map data as a pre-computed reference model (copy) of the environment. Instead of processing raw sensor data from scratch, the system leverages the pre-structured 3D map representation, which already encodes spatial relationships, surface orientations, and geometric constraints, significantly reducing the computational burden of pose estimation.
Data Source
AI summary
An operation method of a navigation apparatus includes: obtaining valid global positioning system (GPS) data at a current time point corresponding to a current position of a target device; determining first neighboring map elements corresponding to a first region indicated by the valid GPS data at the current time point from among a plurality of map elements of map data; and determining a pose parameter of the target device at the current time point based on a first direction specified by at least a portion of the first neighboring map elements.