Scribd
Upload
40 views

Abstract 1.0

This document discusses recent approaches to track-to-track association and track estimation in multi-sensor tracking. It addresses the challenges of associating target tracks from multiple sensors to the appropriate target and estimating target state using track data from multiple sensors. The document proposes using convex optimization for the track-to-track association problem and a robust Kalman filter approach for the track estimation problem in centralized multi-sensor tracking.

Uploaded by

PawanKumar Barnwal
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
40 views

Abstract 1.0

This document discusses recent approaches to track-to-track association and track estimation in multi-sensor tracking. It addresses the challenges of associating target tracks from multiple sensors to the appropriate target and estimating target state using track data from multiple sensors. The document proposes using convex optimization for the track-to-track association problem and a robust Kalman filter approach for the track estimation problem in centralized multi-sensor tracking.

Uploaded by

PawanKumar Barnwal
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

Recent Approaches to Track to Track Association

and Track Estimation in Multi sensor Tracking


K P Jayaprakash

Harshavardhan Vajjaramatti

M.Tech
Electrical Engineering
IIT Kanpur
Kanpur, India 208016
Email: [email protected]

M.Tech
Electrical Engineering
IIT Kanpur
Kanpur, India 208016
Email: [email protected]

AbstractThe current day systems are generally system of


systems comprising of multiple sensors and a centralized computing engine. The recent interest in system design has been in
multi sensor data fusion and the need to associate and Estimate
simultaneously multiple tracks (estimated data) from multiple
sensors has arisen. The idea of using multiple tracks has been
efficient, due to the fact that the bandwidth constraints exist in
real systems and it is not feasible to transmit the observation
data (measurement data) directly but instead the local tracks
and covariances are transmitted. The problem of Multi target
tracking in a centralized fusion architecture comprises of two
distinct but tightly coupled challenges: (i) the naturally discrete
problem of Track to Track association, i.e assignment of target
tracks from multiple sensors to appropriate target; (ii) the
naturally continuous problem of target trajectory estimation; ie
estimation of target state using the track data from multiple
sensors. The recent approaches for Track to Track association
and Track Estimation are investigated and compared. The
convex optimization approach suits well for the Track to Track
association problem hence we have proposed the same for above
problem, similarly a Robust Kalman filtering approach well
suits for Track Estimation problem in Centralized Multi sensor
Tracking and hence is proposed.
Keywords : Convex Optimization, Robust Kalman Filter,
Track to Track Association, Track Estimation, Multi Sensor
Tracking, Multi sensor Data Fusion (MSDF).

R EFERENCES
[1] Siddharth. Joshi and Stephen. Boyd, Sensor Selection via Convex
Optimization, IEEE Trans. on Signal Process., vol.57, No.2, Feb. 2009.

You might also like