Lane-Level Navigation Path Planning for Complex Roads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems face inaccuracies in identifying vehicle location on complex roads, leading to errors in path planning due to low GNSS/GPS precision, especially when vehicles deviate from designated lanes or paths.
Innovation Solution
Utilizes high-definition map modules to provide lane-level road information, enabling accurate determination of vehicle location and updating navigation paths based on real-time lane-level data to prevent deviations and improve path planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GNSS/GPS positioning information is used for navigation path planning, then the navigation system can provide path planning information, but the positioning precision is relatively low causing errors in identifying vehicle location and lane
Solution Approach 1:
The patent combines GNSS/GPS positioning information with image recognition technology and high-definition map data to create a multi-source positioning system. The image recognition module captures road surface features and lane markings, which are then fused with satellite positioning and map data to achieve accurate lane-level positioning, resolving the contradiction between using simple GNSS and achieving high navigation accuracy.
Solution Approach 2:
The patent introduces image recognition technology as an intermediary between GNSS positioning and navigation path planning. The image recognition module processes visual data of road features and lane markings to determine precise vehicle location and lane information, acting as a mediator that translates imprecise GNSS coordinates into accurate lane-level positioning for reliable navigation.
2Productivity
If complex road designs are implemented to accommodate increasing traffic flow, then road capacity increases, but it becomes difficult to identify the current lane and detect path deviations
Solution Approach 1:
The patent uses image recognition technology to continuously capture and analyze road surface features, creating a dynamic visual feedback system. The system processes real-time images of lane markings and road features, enabling precise detection of vehicle position and lane status even on complex road designs with multiple lanes and intersections.
Solution Approach 2:
The patent implements a feedback mechanism where the image recognition module continuously monitors road features and lane markings, compares the detected position with the planned navigation path, and provides real-time feedback on lane identification and path deviation detection. This closed-loop feedback system enables accurate tracking on complex road networks.
Data Source
Figure 1A~1B
Figure 2~3
Figure 4
AI summary
This application discloses a navigation method and apparatus, to provide a navigation path for a target vehicle. The method includes: receiving first location information of the target vehicle sent by a high-definition map module, where the first location information includes lane information of a first location at which the target vehicle is currently located; and determining a navigation path based on the first location information and a destination of the target vehicle. The lane information of the current location of the target vehicle can be provided for a navigation map module by using the high-definition map module, so as to more accurately determine a navigation path suitable for a lane in which the target vehicle is currently located, thereby improving accuracy of navigation planning.