Vehicle Positioning Using Road Marking Distance Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle positioning methods are not robust due to light and dynamic obstacles, and are costly due to hardware and system configuration requirements.
Innovation Solution
A method and apparatus for vehicle positioning that acquire a distance transformation image of road marking objects from a road image and a vector sub-graph from a global vector map, then determine a target positioning posture based on these images and sub-graphs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If positioning layer is used for vehicle positioning, then positioning can be achieved, but robustness deteriorates due to influences of light and dynamic obstacles
Solution Approach 1:
The patent creates a distance transformation image that copies and represents the road marking objects from the original road image. This transformed image serves as a simplified representation that preserves the essential geometric information of road markings while removing the harmful effects of light variations and dynamic obstacles, enabling robust positioning through image-subgraph matching
Solution Approach 2:
The patent replaces the traditional positioning layer approach with an image processing and matching system. Instead of relying on specialized positioning hardware layers that are sensitive to environmental factors, the system uses computer vision techniques (distance transformation, subgraph matching) to achieve positioning, thereby substituting a mechanical/optical system with a computational one that is more robust to environmental variations
2Measurement precision
If high-precision vector map is used for vehicle positioning, then positioning accuracy is improved, but cost increases due to hardware and system configuration requirements
Solution Approach 1:
The patent extracts only the necessary vector subgraph information related to road marking objects from the complete high-precision vector map. Instead of using the entire complex map data structure and associated hardware systems, the method extracts and matches only the relevant road marking geometric information, thereby achieving accurate positioning with reduced system complexity and lower costs
Solution Approach 2:
The patent segments the positioning problem into two independent parts: (1) extracting road marking objects and creating distance transformation images from camera images, and (2) matching these transformed images with corresponding vector subgraphs from the map. This segmentation allows the system to achieve high-precision positioning using only the necessary map data portions rather than requiring complete high-precision mapping systems
Data Source
AI summary
A method and an apparatus for vehicle positioning. The method includes: acquiring a distance transformation image of a road marking object in a road image based on the road image of an environment where a vehicle is located at a current time point; acquiring a vector subgraph of the road marking object from a global vector map; and determining a target positioning posture of the vehicle based on the distance transformation image and the vector subgraph.