Anomaly detection with a moving camera using multiscale video analysis
- 140 Downloads
This paper addresses the problem of abandoned object detection in a cluttered environment using a camera moving along a straight track. The developed system compares captured images to a previously recorded reference video, thus requiring proper temporal alignment and geometric registration between the two signals. A real-time constraint is imposed onto the system to allow an effective surveillance capability in practical situations. In this paper, we propose to deal with the simultaneous detection of objects of different sizes using a multiresolution approach together with normalized cross-correlation and a voting step. In order to develop and properly assess the proposed method we designed a database recorded in a real surveillance scenario, consisting of an industrial plant containing a large number of pipes and rotating machines. Also, we have devised a systematic parameter tuning routine that allows the system to be adapted to different scenarios. We have validated it using the designed database. The obtained results are quite effective, achieving real-time, robust abandoned object detection in an industrial plant scenario.
KeywordsVideo surveillance Moving camera Abandoned object detection Cluttered environment
This work was developed with the partial support of Statoil Brazil, Petrobras, ANP, and CAPES and CNPq funding agencies.
Funding was provided by Conselho Nacional de Desenvolvimento Científico e Tecnológico (Grant No. 203876/2017-2), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (Grant No. 88881.135449/2016-01), Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro.
- Chang, L., Zhao, H., Zhai, S., Ma, Y., & Liu, H. (2013). Robust abandoned object detection and analysis based on online learning. In International conference on robotics and biomimetics, Shenzhen, China.Google Scholar
- da Silva, A. F., Thomaz, L. A., Carvalho, G., Nakahata, M. T., Jardim, E., de Oliveira, J. F. L., et al. (2014). An annotated video database for abandoned-object detection in a cluttered environment. In International telecommunications symposium, Sao Paulo, Brazil.Google Scholar
- da Silva, A. F., Thomaz, L. A., Netto, S. L., & da Silva, E. A. B. (2017). Online video-based sequence synchronization for moving camera object detection. In IEEE international workshop on multimedia signal processing (pp. 1–6). Luton, UK.Google Scholar
- FLIR. (2016). Flir GF320. http://www.flir.com/ogi/display/?id=55671. Accessed January 14, 2016.
- iRobot. (2016). iRobot Roomba vacuum cleaning robot. http://www.irobot.com/For-the-Home/Vacuuming/Roomba.aspx. Accessed February 26, 2018.
- Kim, J., Ye, G., & Kim, D. (2010). Moving object detection under free-moving camera. In IEEE international conference on image processing, Hong Kong.Google Scholar
- Kucharczak, F., da Silva, A. F., Thomaz, L. A., Carvalho, G., da Silva, E. A. B., & Netto, S. L. (2014). Comparison and optimization of image descriptors for real-time detection of abandoned objects. In Simpósio de Processamento de Sinais da UNICAMP, Campinas, Brazil. http://www.sps.fee.unicamp.br/anais/vol01/VSPS_a24_LThomaz.pdf.
- Kundu, A., Jawahar, C. V., & Krishna, K. M. (2010). Realtime moving object detection from a freely moving monocular camera. In IEEE international conference on robotics and biomimetics (pp. 1635–1640). Tianjin, China.Google Scholar
- Mukojima, H., Deguchi, D., Kawanishi, Y., & Ide, I. (2016). Moving camera background subtraction for obstacle detection on railway tracks. In International conference on image processing, Arizona, USA.Google Scholar
- Nakahata, M. T., Thomaz, L. A., da Silva, A. F., da Silva, E. A. B., & Netto, S. L. (2017). Anomaly detection with a moving camera using spatio-temporal codebooks. Multidimensional Systems and Signal Processing. https://doi.org/10.1007/s11045-017-0486-8.
- VDAO. (2014). VDAO—Video database of abandoned objects in a cluttered industrial environment. http://www.smt.ufrj.br/~tvdigital/database/objects. Accessed February 26, 2018.
- VDAO-200. (2017). 200-frame excerpts form VDAO database. http://www02.smt.ufrj.br/~tvdigital/database/research/. Accessed February 26, 2018.
- Weinzaepfel, P., Revaud, J., Harchaoui, Z., & Schmid, C. (2013). Deppflow: Large displacement optical flow with deep matching. In International conference on computer vision, Sydney, Australia.Google Scholar
- Xie, Y., Lin, L., & Jia, Y. (2010). Tracking objects with adaptive feature patches for PTZ camera visual surveillance. In International conference on pattern recognition (pp. 1739–1742). Istanbul, Turkey.Google Scholar
- Xue, K., Ogunmakin, G., Liu, Y., Vela, P. A., & Wang, Y. (2011). PTZ camera-based adaptive panoramic and multi-layered background model. In IEEE international conference on image processing, Brussels, Belgium.Google Scholar