Radar Localization via Sensor Segmentation and Feature Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in localization, particularly in adverse weather conditions, where GPS and sensor data effectiveness is reduced, leading to inaccuracies in determining geographical position.
Innovation Solution
A localization method that constructs self-consistent sensor data groups based on attributes like Doppler shift, return power, and neighborhood similarity, using a feature correlation module to estimate vehicle position through a processor, with outlier removal procedures like RANSAC, focusing on static-nonmoveable objects for accurate positioning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If GPS and sensor data are used for localization, then position estimation can be obtained, but localization accuracy deteriorates in adverse weather conditions
Solution Approach 1:
The patent segments sensor returns into multiple groups based on attributes like Doppler shift, return power, and neighborhood similarity. This segmentation allows the system to process different types of objects (static, dynamic, moving) separately, improving localization reliability by focusing on stable features from static-nonmoveable objects even in adverse weather conditions.
Solution Approach 2:
The patent changes parameters by constructing multiple sensor data groups with different attribute combinations and selecting the most reliable features for correlation. By adjusting which attributes are used for feature construction based on environmental conditions, the system maintains measurement precision when GPS and traditional sensor data become unreliable in adverse weather.
2Measurement precision
If multiple sensor attributes are processed, then localization accuracy improves, but computational complexity increases
Solution Approach 1:
The patent divides sensor returns into multiple groups based on attribute similarity, processing each group separately to identify stable features. This segmentation reduces computational complexity by focusing processing on relevant subsets of data rather than processing all sensor attributes uniformly, while still achieving high localization accuracy through correlated feature matching.
Solution Approach 2:
The patent extracts and removes outliers from sensor data using procedures like RANSAC, isolating only the most reliable features for localization. By taking out and discarding unreliable data points and attributes, the system reduces processing complexity while maintaining or improving localization accuracy through focused processing of high-quality features.
3Measurement precision
If outlier removal procedures are applied, then position estimation accuracy improves, but processing time increases
Solution Approach 1:
The patent performs preliminary segmentation of sensor returns into groups based on attributes before conducting outlier removal. This preliminary organization reduces the data volume that requires intensive outlier removal processing, thereby reducing processing time while still achieving high position estimation accuracy through subsequent correlated feature matching.
Solution Approach 2:
By segmenting sensor data into multiple groups based on attribute similarity before outlier removal, the patent reduces the computational burden of outlier detection. Each group can be processed more efficiently, and the segmentation itself pre-identifies potential outliers, reducing the overall processing time while maintaining high accuracy in position estimation.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances localization accuracy in autonomous vehicles by effectively processing sensor data to improve position estimation, even in challenging weather conditions, by constructing meaningful features and correlating them over time, thereby improving navigation reliability.
Implementation Method 1
The plurality of corresponding attributes includes at least one of Doppler shift, return power, and neighborhood similarity; and the sensor data is at least one of radar data and lidar data
Data Source
AI summary
Systems and method are provided for controlling a vehicle. In one embodiment, a localization method includes receiving sensor data relating to an environment of a vehicle, the sensor data including a plurality of sensor returns associated with objects in the environment, each of the sensor returns having a plurality of corresponding attributes, and constructing a first plurality of sensor data groups, each including a self-consistent subset of the plurality of sensor returns based on their corresponding attributes. The method further includes defining, for each of the first plurality of sensor data groups, a first set of features, wherein each feature is based on at least one of the corresponding attributes and each has an associated feature location, and determining, with a processor, a feature correlation between the first set of features and a second, previously determined set of features.