Traffic accident detection through a hydrodynamic lens

In this paper we present a novel method for automatic traffic accident detection, based on Smoothed Particles Hydrodynamics (SPH). In our method, a motion flow field is obtained from the video through dense optical flow extraction. Then a thermal diffusion process (TDP) is exploited to turn the motion flow field into a coherent motion field. Approximating the moving particles to individuals, their interaction forces, represented as endothermic reactions, are computed using the enthalpy measure, thus obtaining the potential particles of interest. Furthermore, we exploit SPH that accumulates the contribution of each particle in a weighted form, based on a kernel function. The experimental evaluation is conducted on a set of video sequences collected from Youtube, and the obtained results are compared against a state of the art technique.

[1]  Jianxin Wu,et al.  Finding Coherent Motions and Semantic Regions in Crowd Scenes: A Diffusion and Clustering Approach , 2014, ECCV.

[2]  Mubarak Shah,et al.  A Streakline Representation of Flow in Crowded Scenes , 2010, ECCV.

[3]  Qingshan Liu,et al.  Abnormal detection using interaction energy potentials , 2011, CVPR 2011.

[4]  Christian Bauckhage,et al.  Loveparade 2010: Automatic video analysis of a crowd disaster , 2012, Comput. Vis. Image Underst..

[5]  M. Aoki,et al.  Traffic flow measurement using stereo slit camera , 2004, Proceedings. The 7th International IEEE Conference on Intelligent Transportation Systems (IEEE Cat. No.04TH8749).

[6]  Nicola Conci,et al.  Dominant Motion Analysis in Regular and Irregular Crowd Scenes , 2014, HBU.

[7]  Nicola Conci,et al.  Structured learning for crowd motion segmentation , 2013, 2013 IEEE International Conference on Image Processing.

[8]  S. Chapman M.A. D.Sc.,et al.  XXII. A note on thermal diffusion , 1917 .

[9]  Sergio A. Velastin,et al.  Crowd analysis: a survey , 2008, Machine Vision and Applications.

[10]  Gunnar Farnebäck,et al.  Two-Frame Motion Estimation Based on Polynomial Expansion , 2003, SCIA.

[11]  Helbing,et al.  Social force model for pedestrian dynamics. , 1995, Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics.

[12]  Mubarak Shah,et al.  Learning object motion patterns for anomaly detection and improved object detection , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[13]  Simone Calderara,et al.  Visual Tracking: An Experimental Survey , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[14]  Nicola Conci,et al.  Real-time anomaly detection in dense crowded scenes , 2014, Electronic Imaging.

[15]  Mubarak Shah,et al.  Abnormal crowd behavior detection using social force model , 2009, 2009 IEEE Conference on Computer Vision and Pattern Recognition.

[16]  Ayoub Al-Hamadi,et al.  Real-Time Automatic Traffic Accident Recognition Using HFG , 2010, 2010 20th International Conference on Pattern Recognition.

[17]  Guangyi Shi,et al.  Image Sequences Based Traffic Incident Detection for Signaled Intersections Using HMM , 2009, 2009 Ninth International Conference on Hybrid Intelligent Systems.

[18]  J. Monaghan.  Smoothed Particle Hydrodynamics and Its Diverse Applications , 2012 .

[19]  Nicu Sebe,et al.  Particles cross-influence for entity grouping , 2013, 21st European Signal Processing Conference (EUSIPCO 2013).

[20]  Alessio Del Bue,et al.  Optimizing interaction force for global anomaly detection in crowded scenes , 2011, 2011 IEEE International Conference on Computer Vision Workshops (ICCV Workshops).