Thermal Tracking of Sports Players

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

13 Citationer (Scopus)
333 Downloads (Pure)

Resumé

We present here a real-time tracking algorithm for thermal video from a sports game. Robust detection of people includes routines for handling occlusions and noise before tracking each detected person with a Kalman filter. This online tracking algorithm is compared with a state-of-the-art offline multi-target tracking algorithm. Experiments are performed on a manually annotated 2-minutes video sequence of a real soccer game. The Kalman filter shows a very promising result on this rather challenging sequence with a tracking accuracy above 70 % and is superior compared with the offline tracking approach. Furthermore, the combined detection and tracking algorithm runs in real time at 33 fps, even with large image sizes of 1920 × 480 pixels.
OriginalsprogEngelsk
TidsskriftSensors
Vol/bind14
Udgave nummer8
Sider (fra-til)13679-13691
Antal sider13
ISSN1424-8220
DOI
StatusUdgivet - 29 jul. 2014

Fingerprint

Sports
Hot Temperature
Kalman filters
games
Soccer
Target tracking
Noise
Pixels
occlusion
pixels
Experiments

Citer dette

@article{fe6f0555d0124f89873e32d5757b3f5e,
title = "Thermal Tracking of Sports Players",
abstract = "We present here a real-time tracking algorithm for thermal video from a sports game. Robust detection of people includes routines for handling occlusions and noise before tracking each detected person with a Kalman filter. This online tracking algorithm is compared with a state-of-the-art offline multi-target tracking algorithm. Experiments are performed on a manually annotated 2-minutes video sequence of a real soccer game. The Kalman filter shows a very promising result on this rather challenging sequence with a tracking accuracy above 70 {\%} and is superior compared with the offline tracking approach. Furthermore, the combined detection and tracking algorithm runs in real time at 33 fps, even with large image sizes of 1920 × 480 pixels.",
author = "Rikke Gade and Moeslund, {Thomas B.}",
year = "2014",
month = "7",
day = "29",
doi = "10.3390/s140813679",
language = "English",
volume = "14",
pages = "13679--13691",
journal = "Sensors",
issn = "1424-8220",
publisher = "M D P I AG",
number = "8",

}

Thermal Tracking of Sports Players. / Gade, Rikke; Moeslund, Thomas B.

I: Sensors, Bind 14, Nr. 8, 29.07.2014, s. 13679-13691.

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

TY - JOUR

T1 - Thermal Tracking of Sports Players

AU - Gade, Rikke

AU - Moeslund, Thomas B.

PY - 2014/7/29

Y1 - 2014/7/29

N2 - We present here a real-time tracking algorithm for thermal video from a sports game. Robust detection of people includes routines for handling occlusions and noise before tracking each detected person with a Kalman filter. This online tracking algorithm is compared with a state-of-the-art offline multi-target tracking algorithm. Experiments are performed on a manually annotated 2-minutes video sequence of a real soccer game. The Kalman filter shows a very promising result on this rather challenging sequence with a tracking accuracy above 70 % and is superior compared with the offline tracking approach. Furthermore, the combined detection and tracking algorithm runs in real time at 33 fps, even with large image sizes of 1920 × 480 pixels.

AB - We present here a real-time tracking algorithm for thermal video from a sports game. Robust detection of people includes routines for handling occlusions and noise before tracking each detected person with a Kalman filter. This online tracking algorithm is compared with a state-of-the-art offline multi-target tracking algorithm. Experiments are performed on a manually annotated 2-minutes video sequence of a real soccer game. The Kalman filter shows a very promising result on this rather challenging sequence with a tracking accuracy above 70 % and is superior compared with the offline tracking approach. Furthermore, the combined detection and tracking algorithm runs in real time at 33 fps, even with large image sizes of 1920 × 480 pixels.

U2 - 10.3390/s140813679

DO - 10.3390/s140813679

M3 - Journal article

VL - 14

SP - 13679

EP - 13691

JO - Sensors

JF - Sensors

SN - 1424-8220

IS - 8

ER -