Radar Reference Mapping Using HD Map Objects and Occupancy Grids
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current radar localization systems face challenges in generating accurate and complete radar reference maps, especially in environments with insufficient or low-quality radar localization objects, leading to increased driver takeovers and decreased safety and satisfaction in autonomous vehicle operations.
Innovation Solution
The method involves receiving a high-definition map, determining HD map objects, and indicating occupancy cells in a radar occupancy grid based on object attributes to generate a robust and spatially efficient radar reference map, which can be updated through multiple iterations using radar detections and HD map data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If radar reference maps are generated using insufficient or low-quality radar localization objects, then map generation is simpler and faster, but localization accuracy deteriorates leading to increased driver takeovers
Solution Approach 1:
The system performs preliminary actions by pre-processing radar detections to identify and classify localization objects before map generation. It pre-computes object attributes such as positions, sizes, and types, and prepares a structured representation of the environment. This preliminary organization of data enables faster subsequent map generation while maintaining high localization accuracy through quality filtered input data.
Solution Approach 2:
The system changes parameters by dynamically adjusting the selection criteria for radar localization objects based on environmental conditions and map quality requirements. It modifies parameters such as object detection thresholds, classification criteria, and attribute precision levels to optimize the balance between map generation speed and localization accuracy for different operating scenarios.
2Measurement precision
If complete and accurate radar reference maps are generated, then localization accuracy improves, but map generation complexity and processing time increase
Solution Approach 1:
The system segments the map generation process into distinct modular stages: radar detection and filtering, localization object identification, attribute extraction and validation, spatial relationship computation, and map structure construction. Each segment processes specific aspects of the data independently, enabling parallel processing and reducing overall complexity while maintaining comprehensive accuracy through systematic coverage of all necessary processing steps.
Solution Approach 2:
The system introduces intermediary data structures and processing layers between raw radar detections and the final reference map. It uses intermediate representations such as filtered object lists, validated attribute sets, and pre-computed spatial relationships to bridge the gap between complex raw data and the structured final map, simplifying the overall generation process while preserving accuracy.
3Reliability
If comprehensive HD map objects and attributes are processed, then reference map completeness improves, but processing time and computational resources increase
Solution Approach 1:
The system applies local quality by processing HD map objects and attributes with varying levels of detail based on their importance and impact on localization accuracy. Critical objects such as road boundaries and intersections receive comprehensive processing with full attribute sets, while less critical objects use simplified representations. This selective processing approach ensures reference map completeness for essential features while reducing overall processing time through differential detail levels.
Data Source
Figure 1
Figure 2-1
Figure 2-2
AI summary
Methods and systems are described that enable radar reference map generation. A high-definition (HD) map is received and one or more HD map objects within the HD map are determined. Attributes of the respective HD map objects are determined, and, for each HD map object, one or more occupancy cells of a radar occupancy grid are indicated as occupied space based on the attributes of the respective HD map object. By doing so, a radar reference map may be generated without a vehicle traversing through an area corresponding to the radar reference map.