Vehicle Pose Correction Using Radar Landmarks and Barriers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle localization technologies using radar sensors face challenges in cluttered and sparse environments, leading to inaccurate or unreliable pose corrections due to noisy or sparse data.
Innovation Solution
The system employs a method that receives a radar occupancy grid (ROG), a map, and the vehicle's pose, identifies landmark and barrier locations, calculates association probabilities, and determines pose corrections based on ripple point locations and a cost function to generate an updated vehicle pose.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If radar sensors are used for vehicle localization in cluttered environments, then the system can operate without additional sensors, but the data becomes noisy leading to inaccurate pose corrections
Solution Approach 1:
The patent extracts and isolates specific reliable features (landmarks and barriers) from the cluttered environment data. By identifying and focusing only on these distinct, recognizable features rather than processing all radar data points, the system filters out noise while maintaining sensor independence. The landmarks and barriers serve as extracted key elements that provide reliable localization references despite the cluttered surroundings.
Solution Approach 2:
The patent applies local quality by treating different regions of the environment differently - landmarks and barriers receive special processing and weighting compared to general cluttered areas. The system identifies specific locations with high confidence (landmarks/barriers) and uses these localized reliable features to anchor the pose correction, rather than uniformly processing all environmental data. This allows accurate localization by focusing computational resources on high-value regions.
2Adaptability or versatility
If radar sensors are used for vehicle localization in sparse environments, then the system can operate without additional sensors, but the data becomes sparse leading to unreliable pose corrections
Solution Approach 1:
The patent merges multiple data sources - the radar occupancy grid data with pre-stored map data containing known landmark and barrier locations. By combining the real-time sparse radar measurements with the comprehensive prior knowledge from maps, the system compensates for data sparsity. The map provides expected locations of reliable features that guide the interpretation of limited radar data, making pose corrections reliable even when radar returns are sparse.
Solution Approach 2:
The patent performs preliminary action by pre-storing detailed map information about landmarks and barriers before the vehicle reaches the environment. This advance preparation allows the system to have expectations about where reliable features should be located, enabling it to make sense of sparse radar data by comparing against predetermined feature locations. The preliminary map data acts as a guide for interpreting limited real-time sensor information.
3Device complexity
If traditional pose correction methods are used, then the processing is simpler, but the accuracy deteriorates in challenging environments
Solution Approach 1:
The patent segments the pose correction process into distinct stages: identifying candidate landmarks and barriers, calculating association probabilities between radar detections and map features, selecting the best matches, and computing pose corrections. This segmentation allows the system to apply sophisticated algorithms only where needed (in the association and selection stages) while keeping other parts relatively simple. The breakdown into discrete steps makes the complex process manageable and implementable.
Solution Approach 2:
The patent introduces an intermediary element - the association probability calculation - that bridges the gap between simple radar detections and accurate pose corrections. Rather than directly mapping radar points to pose parameters, the system uses probability-based association as an intermediate step to evaluate and select the most likely landmark-barrier correspondences. This intermediary layer enables accurate localization by systematically evaluating multiple hypotheses before committing to a pose correction.
Data Source
AI summary
The techniques and systems herein enable pose correction based on landmarks and barriers. A vehicle pose, one or more radar-occupancy grid (ROG) landmark locations relative to the vehicle, and one or more map landmark locations are received. Based on determined association probabilities of candidate pairs (e.g., treating the map landmark locations as observations), one of the ROG landmark locations and one of the map landmark locations are selected as corresponding to each other. An ROG barrier location and a map barrier location are identified, and ripple point locations are identified that are along the barriers at a radial distance from the landmarks. Based on the ripple point locations and a cost function, a pose correction for the pose is determined. By using the described techniques, reliable vehicle localization can be performed using radar data and a map in a wide array of environments without necessitating other sensors.