Integrated Camera and Radar Tracking using Multi-Model Cubature Kalman Filter

Venkata Pathuri Bhuvana, Mario Huemer

Research output: Conference proceeding/Chapter in Book/Report/Conference Paperpeer-review

Abstract

In this paper, a multiple camera and radar based tracking method is proposed using an interactive multi-model (IMM) based cubature Kalman filter (CKF). In real world scenarios, object movement cannot be characterized with a single motion model. Furthermore, measurements from a single camera/radar may not contain relevant information to achieve desired tracking accuracy for critical applications. The approach proposed in this paper utilizes multiple motion models and the measurements from multiple camera and radar sensors to perform tracking. The proposed method operates a CKF with information weighted sensor fusion for each motion model. Furthermore, the likelihood of each motion model at a given time is calculated using the measurement residuals from all the camera and radar sensors. In the considered simulation set up, the proposed method achieves better tracking accuracy than an IMM based on the extended Kalman filter method and better switching capability among different motion models compared to the single sensor scenario.
Original languageEnglish
Title of host publication2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP)
PublisherIEEE Computer Society
Pages1-5
Number of pages5
ISBN (Electronic)978-1-7281-2723-1
ISBN (Print)978-1-7281-2724-8
DOIs
Publication statusPublished - 14 Nov 2019
Event2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP) - Ottawa, ON, Canada
Duration: 11 Nov 201914 Nov 2019

Publication series

NameIEEE Global Conference on Signal and Information Processing
PublisherIEEE

Conference

Conference2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP)
Period11/11/1914/11/19

Fingerprint

Dive into the research topics of 'Integrated Camera and Radar Tracking using Multi-Model Cubature Kalman Filter'. Together they form a unique fingerprint.

Cite this