Radar Object Tracking Using Road Topology Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Vehicle perception systems, particularly those using radar, face challenges with noise from static objects and atmospheric conditions, leading to inaccurate object detection and tracking, as moving objects appear spread out and background noise causes positional shifts, making motion returns ineffective.
Innovation Solution
A processor-implemented method that organizes radar measurements into time-ordered clusters, removes noise by comparing them to a road topology map, and uses a constrained Kalman filter to track objects within aligned lanes, ensuring accurate detection and prediction of object movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If radar measurements are used for object detection and tracking, then object detection capability is provided, but noise from static objects and atmospheric conditions causes inaccurate detection and tracking
Solution Approach 1:
The patent introduces road topology maps and lane information as intermediary data structures that mediate between raw radar measurements and object detection results. These intermediaries provide contextual constraints (valid regions, allowed directions of travel) that filter out noise from static objects and atmospheric conditions while preserving valid moving object detections.
Solution Approach 2:
The patent changes the parameter space by transforming radar measurements from raw coordinate data into constrained tracks that satisfy road topology constraints. By enforcing constraints on position (within lanes), direction (allowed directions of travel), and motion patterns (consistent with lane geometry), the system filters noise while maintaining detection accuracy.
2Reliability
If motion returns are used for tracking, then moving objects can be detected, but objects appear spread out over larger regions and positional shifts occur due to background noise
Solution Approach 1:
The patent implements feedback by continuously comparing radar measurements against road topology constraints and adjusting tracks accordingly. The constrained Kalman filter uses feedback from lane boundaries and allowed directions to correct positional drift and maintain accurate object tracking even in the presence of background noise.
Solution Approach 2:
The patent applies parameter changes by transforming uncertain radar measurements into constrained tracks with reduced positional uncertainty. By enforcing geometric constraints from road topology, the system reduces the spread of object positions and eliminates positional shifts caused by background noise.
3Productivity
If conventional radar tracking without constraints is used, then detection speed is maintained, but tracking accuracy deteriorates when objects move outside expected patterns
Solution Approach 1:
The patent applies preliminary action by pre-computing road topology maps and lane constraints before radar detection occurs. These pre-computed constraints (valid regions, allowed directions, lane geometries) are ready to immediately constrain and correct radar measurements, enabling fast yet accurate tracking without sacrificing detection speed.
Data Source
AI summary
A processor-implemented method in a vehicle for detecting objects from radar data includes: retrieving radar measurements taken at different periodic time increments; organizing the radar measurements into appropriate time windows; building a sequence cluster of radar measurements wherein the sequence cluster comprises a sliding window the latest time windows of radar measurements; removing noise from the sequence cluster of radar measurements by removing a cluster of radar measurements from the sequence cluster of radar measurements that is contradictory to a road topology map for an area in which the first object is estimated to be situated; and outputting the sequence cluster of radar measurements after removal of contradictory radar measurements as a new cluster of radar measurements.