Video Processing Algorithms for Detection of Pedestrians
Piniarski Karol, Pawłowski Paweł, Dąbrowski Adam
Poznan University of Technology
Department of Computing, Division of Signal Processing and Electronic Systems
Piotrowo 3, 60-965 Poznań, Poland
E-mails: {pawel.pawlowski, adam.dabrowski}@put.poznan.pl,
karol.piniarski@doctorate.put.poznan.pl
Received:
Received: 12 September 2014; revised: 26 May 2015; accepted: 08 June 2015; published online: 07 September 2015
DOI: 10.12921/cmst.2015.21.03.005
Abstract:
In this paper a video processing procedure for automatic detection of pedestrians is presented. It is planned to use it as a part of the automotive night vision system. Generally, such systems are either passive (i.e. those based on thermal vision) or active (i.e. equipped with illuminators and near infrared cameras). Passive systems provide a large range of detection, while their active counterparts, operating in a somehow smaller range, offer more readable images for car drivers. However, all images produced with both kinds of these systems are quite specific and special image processing procedures are needed for them. For this purpose the authors used modified and adapted algorithms, such as dual-threshold locally adaptive classification, connected component labeling, histogram of oriented gradients, and the support vector machine with a radial basic function kernel or with a linear kernel. Tests performed on the real night vision recordings show very high efficiency of the proposed solution with accuracy equal to 99.2 % for the linear kernel and even to 99.36 % for the radial basic function kernel.
Key words:
automotive systems, classification of pedestrians, night vision, object detection, video processing
References:
[1] European Commission, Towards a European road safety area: policy orientations on road safety 2011-2020, Brussels, COM(2010), 389, (2010).
[2] J. Broughton, C. Brandstaetter, G. Yannis, P. Evgenikos, et al., Basic Fact Sheet “Seasonality”, Deliverable D3.9 of the EC FP7 project DaCoTA, 4, (2012).
[3] J. F. Pace, J. Sanmartín, P. Thomas, A. Kirk, et al., Basic Fact Sheet “Pedestrians”, Deliverable D3.9 of the EC FP7 project DaCoTA, 1-12, (2012).
[4] European Commission, CARE, Road fatalities in the EU since 2001, EU road accidents database, (2013).
[5] P. Pawłowski, D. Prószyński, A. Dąbrowski, Real-time procedures for automatic recognition of road signs, Elektronika – konstrukcje, technologie, zastosowania, Sigma NOT, 3, 57-61 (2009).
[6] K. Piniarski, P. Pawłowski, A. Dąbrowski, Pedestrian Detection by Video Processing in Automotive Night Vision System, Proc. of IEEE Signal Processing Algorithms, Architectures, Arrangements and Applications, SPA 2014, Poznań, Poland, 104-109, (2014).
[7] Y. Luo, J. Remillard, D. Hoetzer, Pedestrian Detection in Near-Infrared Night Vision System, IEEE Intelligent Vehicles Symposium, 51-58, (2010).
[8] F. Jahard, D. A. Fish, A. A. Rio, C. P. Thompson, Far/Near Infrared Adapted Pyramid-Based Fusion for Automotive Night Vision, International Conference on Image Processing and its Applications, Vol. 8, 886-890, (1997).
[9] Q. Liu, J. Zhuang, S. Kong, Detection of Pedestrians at Night Time Using Learning-based Method and Head Validation, Proc. of IEEE Conf. on Imaging Systems and Techniques (IST 2012), 398-402, (2012).
[10] J. Ge, Y. Luo, G. Tei, Real Time Pedestrian Detection and Tracking at Night time for Driver-Assistance Systems, IEEE Transactions on Intelligent Transportation Systems, Vol. 10, No. 2, 283-298, (2009).
[11] Y. Zhang, Y. Zhao, G. Li, R. Cheng, Grey Self-similarity Feature for Night-time Pedestrian Detection, Journal of Computational Information System 10: 7, 2967-2974, (2014)
[12] V. E. Neagoe, C. T. Tudoran, M. Neghina, A neural network approach to pedestrian detection, Proc. of ICCOMP’09, 374-379, (2009).
[13] N. Dalal, B. Triggs, Histograms of Oriented Gradients for HumanDetection, Proc. of IEEE Conference on Computer Vision and Pattern Recognition, Vol. 1, 886-893, (2005).
[14] R. Walczyk, A. Armitage, D. Binnie, Comparative Study on Connected Component Labeling Algorithms for Embedded Video Processing Systems, Proc. of Int. Conf Image Processing, Computer Vision, and Pattern Recognition, IPCV 2010, CSREA Press, Vol. 2, 853-859, (2010).
[15] F. Chang, C. J. Chen, C. J. Lu, A linear-time component-labeling algorithm using contour tracing technique, Computer Vision and Image Understanding, Vol. 93, No. 2, doi:10.1016/j.cviu.2003.09.002, 206-220 (2004).
[16] G. Szwoch, P. Dalka, A. Ciarkowski, P. Szczuko, A. Czy˙zewski, Visual Object Tracking System Employing Fixed and PTZ Cameras, Journal of Intelligent Decision Technologies, Vol. 5, No 2, 177-188, (2011).
[17] C. Cortes, V. Vapnik, Support-Vector Networks, Journal of Machine Learning archive 20(3), 273-297 (1995).
[18] T. Fletcher, Support Vector Machines Explained, University College London, (2009).
[19] M. Bertozzi, A. Broggi, M. Felisa, G. Vezzoni, Low-level Pedestrian Detection by means of Visible and Far Infra-red Tetra-vision, Proc. of IEEE Intelligent Vehicles Symposium, 231-236 (2006).
[20] A. Dąbrowski, P. Pawłowski, M. Stankiewicz, F. Misiorek, Fast and accurate digital signal processing realized with GPGPU technology, Electrical Review, R. 88, No 6, 47-50 (2012).
[21] T. Marciniak, D. Jackowski, P. Pawłowski, A. Dąbrowski, Real-time people tracking using DM6437 EVM, Proc. of IEEE Signal Processing Conference SPA 2009, 116-120, (2009).
In this paper a video processing procedure for automatic detection of pedestrians is presented. It is planned to use it as a part of the automotive night vision system. Generally, such systems are either passive (i.e. those based on thermal vision) or active (i.e. equipped with illuminators and near infrared cameras). Passive systems provide a large range of detection, while their active counterparts, operating in a somehow smaller range, offer more readable images for car drivers. However, all images produced with both kinds of these systems are quite specific and special image processing procedures are needed for them. For this purpose the authors used modified and adapted algorithms, such as dual-threshold locally adaptive classification, connected component labeling, histogram of oriented gradients, and the support vector machine with a radial basic function kernel or with a linear kernel. Tests performed on the real night vision recordings show very high efficiency of the proposed solution with accuracy equal to 99.2 % for the linear kernel and even to 99.36 % for the radial basic function kernel.
Key words:
automotive systems, classification of pedestrians, night vision, object detection, video processing
References:
[1] European Commission, Towards a European road safety area: policy orientations on road safety 2011-2020, Brussels, COM(2010), 389, (2010).
[2] J. Broughton, C. Brandstaetter, G. Yannis, P. Evgenikos, et al., Basic Fact Sheet “Seasonality”, Deliverable D3.9 of the EC FP7 project DaCoTA, 4, (2012).
[3] J. F. Pace, J. Sanmartín, P. Thomas, A. Kirk, et al., Basic Fact Sheet “Pedestrians”, Deliverable D3.9 of the EC FP7 project DaCoTA, 1-12, (2012).
[4] European Commission, CARE, Road fatalities in the EU since 2001, EU road accidents database, (2013).
[5] P. Pawłowski, D. Prószyński, A. Dąbrowski, Real-time procedures for automatic recognition of road signs, Elektronika – konstrukcje, technologie, zastosowania, Sigma NOT, 3, 57-61 (2009).
[6] K. Piniarski, P. Pawłowski, A. Dąbrowski, Pedestrian Detection by Video Processing in Automotive Night Vision System, Proc. of IEEE Signal Processing Algorithms, Architectures, Arrangements and Applications, SPA 2014, Poznań, Poland, 104-109, (2014).
[7] Y. Luo, J. Remillard, D. Hoetzer, Pedestrian Detection in Near-Infrared Night Vision System, IEEE Intelligent Vehicles Symposium, 51-58, (2010).
[8] F. Jahard, D. A. Fish, A. A. Rio, C. P. Thompson, Far/Near Infrared Adapted Pyramid-Based Fusion for Automotive Night Vision, International Conference on Image Processing and its Applications, Vol. 8, 886-890, (1997).
[9] Q. Liu, J. Zhuang, S. Kong, Detection of Pedestrians at Night Time Using Learning-based Method and Head Validation, Proc. of IEEE Conf. on Imaging Systems and Techniques (IST 2012), 398-402, (2012).
[10] J. Ge, Y. Luo, G. Tei, Real Time Pedestrian Detection and Tracking at Night time for Driver-Assistance Systems, IEEE Transactions on Intelligent Transportation Systems, Vol. 10, No. 2, 283-298, (2009).
[11] Y. Zhang, Y. Zhao, G. Li, R. Cheng, Grey Self-similarity Feature for Night-time Pedestrian Detection, Journal of Computational Information System 10: 7, 2967-2974, (2014)
[12] V. E. Neagoe, C. T. Tudoran, M. Neghina, A neural network approach to pedestrian detection, Proc. of ICCOMP’09, 374-379, (2009).
[13] N. Dalal, B. Triggs, Histograms of Oriented Gradients for HumanDetection, Proc. of IEEE Conference on Computer Vision and Pattern Recognition, Vol. 1, 886-893, (2005).
[14] R. Walczyk, A. Armitage, D. Binnie, Comparative Study on Connected Component Labeling Algorithms for Embedded Video Processing Systems, Proc. of Int. Conf Image Processing, Computer Vision, and Pattern Recognition, IPCV 2010, CSREA Press, Vol. 2, 853-859, (2010).
[15] F. Chang, C. J. Chen, C. J. Lu, A linear-time component-labeling algorithm using contour tracing technique, Computer Vision and Image Understanding, Vol. 93, No. 2, doi:10.1016/j.cviu.2003.09.002, 206-220 (2004).
[16] G. Szwoch, P. Dalka, A. Ciarkowski, P. Szczuko, A. Czy˙zewski, Visual Object Tracking System Employing Fixed and PTZ Cameras, Journal of Intelligent Decision Technologies, Vol. 5, No 2, 177-188, (2011).
[17] C. Cortes, V. Vapnik, Support-Vector Networks, Journal of Machine Learning archive 20(3), 273-297 (1995).
[18] T. Fletcher, Support Vector Machines Explained, University College London, (2009).
[19] M. Bertozzi, A. Broggi, M. Felisa, G. Vezzoni, Low-level Pedestrian Detection by means of Visible and Far Infra-red Tetra-vision, Proc. of IEEE Intelligent Vehicles Symposium, 231-236 (2006).
[20] A. Dąbrowski, P. Pawłowski, M. Stankiewicz, F. Misiorek, Fast and accurate digital signal processing realized with GPGPU technology, Electrical Review, R. 88, No 6, 47-50 (2012).
[21] T. Marciniak, D. Jackowski, P. Pawłowski, A. Dąbrowski, Real-time people tracking using DM6437 EVM, Proc. of IEEE Signal Processing Conference SPA 2009, 116-120, (2009).
