Realtime vision based multi-target-tracking with particle filters in automotive applications

被引：11

作者：

Idler, Corvin ^{[1
]}

Schweiger, Roland ^{[3
]}

Paulus, Dietrich ^{[1
]}

Maehlisch, Mirko ^{[2
]}

Ritter, Werner ^{[2
]}

机构：

[1] Univ Koblenz Landau, Inst Computat Visualist, D-56070 Koblenz, Germany

[2] DaimlerChrysler AG, Res & Technol REI AI, D-89081 Ulm, Germany

[3] Univ Ulm, Dept Measurement, D-89069 Ulm, Germany

来源：

2006 IEEE INTELLIGENT VEHICLES SYMPOSIUM | 2006年

关键词：

D O I：

10.1109/IVS.2006.1689626

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In the following we present a multi-instance-based multi-target-tracking method using particle filters. We developed a robust and flexible system capable of tracking an unknown and changing number of objects (vehicles) using early-vision image features. The system works in real time.

引用

页码：188 / +

页数：2

共 25 条

[1] [Anonymous], BRIT MACH VIS C
[2] [Anonymous], 2003, HIGHLY STRUCTURED ST
[3] A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking
Arulampalam, MS
Maskell, S
Gordon, N
Clapp, T
[J]. IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2002, 50 (02) : 174 - 188
[4] BAEHRING D, 2003, WORKSH FAHR FAS2003, P9
[5] Broggi A, 2004, 2004 IEEE INTELLIGENT VEHICLES SYMPOSIUM, P310
[6] Collado JM, 2004, 2004 IEEE INTELLIGENT VEHICLES SYMPOSIUM, P572
[7] DELLAERT F, 1997, C INT TRANSP SYST IN, P72
[8] Doucet A., 2001, SEQUENTIAL MONTE CAR
[9] Doucet A., 1998, CUEDFINFENGTR310
[10] Egelhaaf JC, 2003, VDI BERICHT, V1768, P247

← 1 2 3 →