Millimeter-Wave Radar Target Tracking Using Feature Extraction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current radar systems face challenges in efficiently tracking human targets at low frame-rates without relying on motion models, particularly in millimeter-wave frequency regimes, which limits their power efficiency and compliance with regulatory requirements while maintaining effective tracking performance.

Innovation Solution

A millimeter-wave radar system that uses feature extraction techniques such as Empirical Mode Decomposition (EMD) and Scale Invariant Feature Transform (SIFT) to associate targets with tracks, allowing for tracking without knowledge of target motion or localization, and incorporating motion models based on range and Doppler information to enhance tracking accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If radar systems operate at low frame-rates to reduce power consumption, then power efficiency improves, but tracking performance deteriorates

Engineering Contradiction:
Improvepower consumptionVSAvoidtracking performance
Core Design Contradiction:
Use of energy by moving objectVSReliability

Solution Approach 1:

The system employs feedback mechanisms where detection results from previous frames are fed into the tracking algorithm to predict and associate targets in current frames. This allows the radar to maintain effective tracking at lower frame rates by using historical data to compensate for reduced measurement frequency.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The tracking algorithm performs preliminary associations based on predicted target positions before new detections are fully processed. By pre-establishing track hypotheses using motion models and previous detections, the system maintains tracking continuity even when detections are sparse due to low frame rates.

Inventive Principle:
Principle #10Preliminary action

2Adaptability or versatility

If radar systems do not rely on motion models for target association, then adaptability to unknown target behaviors improves, but tracking accuracy deteriorates

Engineering Contradiction:
Improveadaptability to target motionVSAvoidtracking accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system dynamically adjusts its approach by combining model-based predictions with data-driven associations. When targets exhibit predictable motion, the motion model provides accurate predictions. When targets display unpredictable behavior, the system relies more on detection-based associations, creating a flexible hybrid tracking framework.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The tracking algorithm adjusts its parameters dynamically based on detection quality and target behavior. When detections are reliable and targets move predictably, the system increases reliance on motion model parameters. When detections are sparse or targets exhibit erratic motion, the system shifts weight toward detection-based association parameters.

Inventive Principle:
Principle #35Parameter changes

3Reliability

If feature extraction techniques are used for target association, then tracking performance without motion models improves, but computational complexity increases

Engineering Contradiction:
Improvetracking performanceVSAvoidcomputational complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system extracts only the most relevant features from radar detections for association purposes, such as range, velocity, and angular position. By selecting and extracting only the critical features needed for tracking rather than processing all available data, the system achieves effective target association with reduced computational burden.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system applies feature extraction selectively to only those detections that are relevant for tracking associations, rather than processing all detections uniformly. This partial application of feature extraction reduces computational complexity while maintaining tracking performance for the targets of interest.

Inventive Principle:
Principle #16Partial or excessive action

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 efficient tracking of human targets at low frame-rates, reducing power consumption and adhering to regulatory limits while maintaining high tracking performance, and is suitable for distributed radar systems where targets may move between fields-of-view.

Implementation Method 1

In some radar systems, the distance between the radar and a target is determined by transmitting a frequency modulated signal, receiving a reflection of the frequency modulated signal (also referred to as the echo), and determining a distance based on a time delay and/or frequency difference between the transmission and reception of the frequency modulated signal.

Methodology Applied
Scientific EffectRadar: Radar

Implementation Method 2

In some radar systems, the distance between the radar and a target is determined by transmitting a frequency modulated signal, receiving a reflection of the frequency modulated signal (also referred to as the echo), and determining a distance based on a time delay and/or frequency difference between the transmission and reception of the frequency modulated signal.

Methodology Applied
Scientific EffectDoppler Effect: Doppler Effect

Data Source

PatentEP4001960B1Radar-based tracker for target sensing
Publication Date: 2025.01.15 INFINEON TECHNOLOGIES AG
  • EP4001960B1 patent drawingFigure 1
  • EP4001960B1 patent drawingFigure 2~4
  • EP4001960B1 patent drawingFigure 5~6

AI summary

In an embodiment, a method for tracking targets includes: receiving data from a radar sensor of a radar; processing the received data to detect targets; identifying a first geometric feature of a first detected target at a first time step, the first detected target being associated to a first track; identifying a second geometric feature of a second detected target at a second time step; determining an error value based on the first and second geometric features; and associating the second detected target to the first track based on the error value.