Overlapping Radar Sensors for Close-User Hand Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing radar systems struggle with accurately tracking multiple individuals and their hand gestures in complex environments, particularly when users are close together, leading to incorrect cluster assignments and loss of track history.
Innovation Solution
A method and device using two radar sensors with overlapping fields of view, combined with enhanced preprocessing and estimation algorithms, including a density-based cluster analysis, to separate body and hand tracks and maintain accurate position profiles, even in complex scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a traditional radar sensor with tracking algorithm is used for personal tracking, then the system can track persons in indoor and outdoor areas, but it fails to accurately distinguish multiple users when they are close together, leading to incorrect cluster assignments and loss of track history
Solution Approach 1:
The patent divides the tracking task into separate processing pipelines: one for person tracking and another for hand/gesture tracking. The point cloud data is segmented into body points and hand points using clustering algorithms, allowing independent tracking of each with dedicated algorithms optimized for their specific characteristics
Solution Approach 2:
The patent implements a history buffer that stores past track assignments and uses this historical information in the tracking algorithm. This preliminary action of preserving track history allows the system to maintain reliable tracking even when users are close together, preventing track loss by referencing previous frame assignments
2Adaptability or versatility
If the radar system tracks hand gestures for gesture recognition, then it can enable gesture-based control functions, but it becomes vulnerable to incorrect assignment of hand points to wrong user tracks when multiple users are present
Solution Approach 1:
The patent separates hand tracking from body tracking by implementing distinct processing pipelines. Hand points are identified and clustered separately from body points, and the hand tracking algorithm operates independently on the hand point cloud data, using hand-specific parameters and update rules
Solution Approach 2:
The patent implements a feedback mechanism where the system continuously monitors the spatial relationship between hand points and body points. The hand tracking algorithm uses feedback from body track positions to correctly associate hand gestures with the appropriate user, even when multiple users are present and close together
3Measurement precision
If the system uses density-based cluster analysis to separate users, then it can distinguish close users, but it may still incorrectly assign hand points to wrong clusters in complex scenarios
Solution Approach 1:
The patent divides the complex tracking problem into simpler sub-problems by separating body tracking and hand tracking into different pipelines. This segmentation allows each pipeline to use simpler, more specialized algorithms rather than one complex algorithm trying to handle all aspects simultaneously
Solution Approach 2:
The patent implements dynamic track management where the system adapts to changing scenarios frame by frame. The tracking algorithms dynamically adjust to the current situation, maintaining separate tracks for different users and correctly assigning hand points based on real-time spatial relationships and historical track information
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
Enables robust and accurate tracking of multiple individuals and their hand gestures, preventing incorrect cluster assignments and maintaining track history, even when users are close together.
Implementation Method 1
reading in sensor data from an interface to a preprocessing device for preprocessing raw data of two spaced-apart radar sensors
Data Source
AI summary
A method for providing personal tracking data for controlling at least one function of a technical system. The method includes reading in, executing, and generating. During the reading in, sensor data from two spaced-apart radar sensors with partially overlapping fields of view are read in, the sensor data representing a point cloud made up of point targets that are detected with the aid of the radar sensors. During the execution, an estimation algorithm is executed, using the sensor data, in order to generate at least one corrected position profile track, the estimation algorithm applying a density-based cluster analysis algorithm. During the generating, the tracking data are generated using the at least one corrected position profile track.