RADAR Object Tracking with RLS Velocity Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Extended Object Tracking (EOT) solutions face challenges in grouping return points from the same object together, associating new return points with existing tracks, updating tracks, and detecting noise in RADAR sensor data from automotive applications.
Innovation Solution
The proposed solution involves processing RADAR return points using a pipeline that includes pre-processing, clustering, cluster velocity estimation, and track management. This involves filtering out noise, accumulating points over multiple time-steps, using density-based clustering algorithms like DBSCAN, and employing Recursive Least Squares (RLS) filters and Kalman Filters to estimate velocities and update tracks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If density-based clustering algorithms like DBSCAN are used to group return points, then object detection accuracy is improved, but computational complexity and processing time increase
Solution Approach 1:
The patent applies preliminary filtering and preprocessing to RADAR return points before clustering, removing obvious noise and organizing data structures in advance. This reduces the computational burden on the clustering algorithm while maintaining detection accuracy, as the algorithm processes pre-filtered data rather than raw return points.
Solution Approach 2:
The patent segments the clustering process into multiple stages: initial grouping of return points, formation of candidate clusters, and refinement of final clusters. This segmentation allows complex clustering to be broken down into manageable steps, reducing overall computational complexity while maintaining accuracy through iterative refinement.
2Measurement precision
If multiple RADAR time-steps are accumulated to increase point density, then tracking accuracy is improved, but processing time and data volume increase
Solution Approach 1:
The patent accumulates RADAR return points from multiple time-steps in advance before clustering operations, creating denser point clouds that improve tracking accuracy. By performing accumulation as a preliminary step, the system prepares optimized input data for subsequent processing, reducing the need for repeated processing of sparse data.
Solution Approach 2:
The patent implements periodic accumulation of RADAR frames at optimized intervals, balancing point density requirements with processing time constraints. By accumulating data periodically rather than continuously, the system maintains adequate point density for accurate tracking while avoiding excessive data volumes that would overwhelm processing resources.
3Measurement precision
If Recursive Least Squares filters are used to estimate cluster velocity, then velocity estimation accuracy is improved, but computational load increases
Solution Approach 1:
The patent performs preliminary calculations and data organization before applying Recursive Least Squares (RLS) filters, pre-computing quantities that will be needed for velocity estimation. This preliminary preparation reduces the computational burden during the actual RLS filtering operation, maintaining high velocity estimation accuracy while lowering real-time processing requirements.
4Reliability
If return points are filtered based on distance from map features and range rate thresholds, then noise rejection is improved, but risk of filtering valid return points increases
Solution Approach 1:
The patent applies different filtering criteria and threshold values for different spatial regions and object types. Instead of using uniform filtering parameters across all return points, the system adapts filtering strictness to local conditions, such as applying more lenient criteria in regions where valid objects are expected and stricter criteria where noise is predominant. This preserves valid return points while effectively rejecting noise.
Data Source
AI summary
Techniques are provided for extended object tracking using RADAR return points only. The techniques include receiving the return points from at least one RADAR sensor of the vehicle. One or more clusters based on or from the return points are generated. An estimated position and velocity of each of the one or more clusters is computed. A Recursive Least Squares (RLS) based algorithm is proposed to estimate the instantaneous velocity of a cluster in real-time that allows for accurate track-cluster association, removes the need to perform computationally expensive non-linear state updates, and allows for the estimation of the true velocity even in frames with a large amount of clutter. If it is determined that the one or more clusters are associated with an existing object track, the existing object track is updated using at least the respective positions of the one or more clusters associated with the existing object track.