Will the Pedestrian Cross? A Study on Pedestrian Path Prediction

被引:227
作者
Keller, Christoph G. [1 ]
Gavrila, Dariu M. [2 ,3 ]
机构
[1] Daimler Res & Dev, Environm Percept Dept, D-71034 Boblingen Hulb, Germany
[2] Daimler Res & Dev, Environm Percept Dept, D-89081 Ulm, Germany
[3] Univ Amsterdam, Intelligent Autonomous Syst Grp, NL-1098 Amsterdam, Netherlands
关键词
Computer vision; pedestrian safety; prediction methods; HUMAN MOTION; MODELS;
D O I
10.1109/TITS.2013.2280766
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
Future vehicle systems for active pedestrian safety will not only require a high recognition performance but also an accurate analysis of the developing traffic situation. In this paper, we present a study on pedestrian path prediction and action classification at short subsecond time intervals. We consider four representative approaches: two novel approaches (based on Gaussian process dynamical models and probabilistic hierarchical trajectory matching) that use augmented features derived from dense optical flow and two approaches as baseline that use positional information only (a Kalman filter and its extension to interacting multiple models). In experiments using stereo vision data obtained from a vehicle, we investigate the accuracy of path prediction and action classification at various time horizons, the effect of various errors (image localization, vehicle egomotion estimation), and the benefit of the proposed approaches. The scenario of interest is that of a crossing pedestrian, who might stop or continue walking at the road curbside. Results indicate similar performance of the four approaches on walking motion, with near-linear dynamics. During stopping, however, the two newly proposed approaches, with nonlinear and/or higher order models and augmented motion features, achieve a more accurate position prediction of 10-50 cm at a time horizon of 0-0.77 s around the stopping event.
引用
收藏
页码:494 / 506
页数:13
相关论文
共 37 条
[1]  
Abramson Y, 2004, 2004 IEEE INTELLIGENT VEHICLES SYMPOSIUM, P590
[2]  
Andriluka M., 2008, PROC IEEE C CVPR, P1
[3]  
[Anonymous], 2007, P INT C ART INT STAT
[4]  
[Anonymous], 2003, STRATEGIES TERMS VUL
[5]  
[Anonymous], P IEEE INT TRANSP SY
[6]   Behavioral priors for detection and tracking of pedestrians in video sequences [J].
Antonini, Gianluca ;
Martinez, Santiago Venegas ;
Bierlaire, Michel ;
Thiran, Jean Phillppe .
INTERNATIONAL JOURNAL OF COMPUTER VISION, 2006, 69 (02) :159-180
[7]  
Bar-Shalom Y., 2004, Estimation with applications to tracking and navigation: Theory algorithms and software
[8]  
Black M.J., 1998, P EUROPEAN C COMPUTE, V1, P909
[9]  
Blackman S., 1999, Design and Analysis of Modern Tracking Systems
[10]   Interacting multiple model particle filter [J].
Boers, Y ;
Driessen, JN .
IEE PROCEEDINGS-RADAR SONAR AND NAVIGATION, 2003, 150 (05) :344-349