HD Route Building From SD Maps for Lane-Level Vehicle Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current mobile and in-vehicle navigation systems rely on standard definition (SD) maps that lack precise lane information necessary for assisted-driving and autonomous-driving systems, which require high definition (HD) map data for safe navigation.
Innovation Solution
A route builder system that matches navigation routes from SD maps to HD maps by interpolating and weighting lane geometry data, generating an HD navigation route that includes additional waypoints and lane details for precise vehicle path planning and control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If SD map databases are used for navigation routes, then the system is simpler and easier to operate, but the route precision and lane information are insufficient for assisted-driving and autonomous-driving systems
Solution Approach 1:
The patent segments the HD map database into multiple tiles that can be selectively loaded and processed. This allows the system to work with divided portions of high-precision map data rather than the entire database, reducing memory requirements and processing complexity while maintaining route precision.
Solution Approach 2:
The patent introduces an intermediary process that matches SD navigation routes with HD map tiles. This intermediary layer translates between the simplified SD route format and the detailed HD map format, enabling assisted-driving systems to utilize precise lane information without requiring a complete system redesign.
2Reliability
If HD map databases are used to provide precise lane information, then the route accuracy improves for assisted-driving and autonomous-driving systems, but the computing power and processing complexity increase significantly
Solution Approach 1:
The HD map database is divided into spatial tiles that are loaded only when needed for specific navigation segments. This segmentation reduces the amount of HD map data that must be held in memory and processed simultaneously, lowering computational energy requirements while maintaining reliable lane-level navigation accuracy.
Solution Approach 2:
The system loads and processes only the portion of HD map data that is currently needed for the active navigation route, rather than loading the entire database. This partial action approach provides sufficient lane information for safe navigation while significantly reducing the computing power and energy consumption required.
3Measurement precision
If grid-based spatial index methods are used to generate precise paths, then the lane-level detail accuracy improves, but the algorithm complexity and computing requirements increase
Solution Approach 1:
The patent divides the HD map into a grid of spatial tiles with predefined dimensions. This segmentation creates a structured framework that simplifies the matching process between SD routes and HD lane data, reducing algorithmic complexity compared to processing the entire map as a single grid structure.
Solution Approach 2:
The patent changes the parameter of spatial organization from a monolithic grid structure to a hierarchical tile-based structure. This parameter change allows for more efficient spatial indexing and matching operations, reducing the computational complexity of generating precise lane-level paths while maintaining measurement precision.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
This document describes techniques, apparatuses, and systems for standard-definition (SD) to high-definition (HD) navigation route determination. An example route builder receives a navigation route generated for a host vehicle. The navigation route includes a list of waypoints generated by an SD map database. The route builder matches the list of waypoints to lane geometry data maintained in an HD map database and outputs HD navigation route to a vehicle controller. The HD navigation route includes the list of waypoints, additional waypoints, and lane geometry data. The vehicle controller can then operate the host vehicle in a roadway environment along the HD navigation route. In this way, the described techniques and systems can provide HD navigation routes required by some assisted-driving and autonomous-driving systems based on SD navigation routes.