Fast compressive tracking combined with Kalman filter
- 18 Downloads
Compressive tracking refers to a group of high-speed algorithms for real-time object tracking. Many tracking algorithms may not generate accurate tracking results because they used fixed learning rates, and sometime lose targets when objects are occluded or deformed. To address these problems, a fast tracking algorithm combined with Kalman filter was proposed in this research. Firstly, an object location was initialized by the predicted value of Kalman filter when it was occluded, and the Kalman update was implemented only when the object was detected. The object location obtained in the Kalman update stage was used later as the initial position in the next frame. Secondly, when the distribution of positive samples satisfied a threshold, an adaptive learning rate was then updated. Finally, the naive Bayes classifier was updated with samples which had more different features. In the experiment, the proposed algorithm was compared with other state-of-the-art algorithms on seven publicly tested sequences, demonstrating that it had higher tracking accuracy and robustness in conditions such as occlusion, deformation and rotation.
KeywordsCompressive tracking Kalman filter Secondary localization Object tracking
This work was supported by National Natural Science Foundation of China (61601358), the Natural Science Basic Research Plan in Shaanxi Province of China (2016JM6030), the Scientific Research Program funded by Shaanxi Provincial Education Department (18JK0349).
Compliance with ethical standards
Conflicts of interest
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
- 2.Bao C, Wu Y, Ling H et al. (2012) Real time robust L1 tracker using accelerated proximal gradient approach. IEEE Conference on Computer Vision and Pattern Recognition, Providence, RI, USA: 1830–1837. doi: https://doi.org/10.1109/CVPR.2012.6247881
- 3.Bolme SD, Beveridge RJ, Draper AB, Lui MY (2010) Visual object tracking using adaptive correlation filters. IEEE Conf Comput Vision Pattern Recognition, San Francisco, CA, USA: 2544–2550. doi: https://doi.org/10.1109/CVPR.2010.5539960
- 6.Dinh TB, Vo N, Medioni G (2011) Context tracker: exploring supporters and distracters in unconstrained environments. IEEE Conference on Computer Vision and Pattern Recognition, Providence, RI, USA: 1177–1184. doi: https://doi.org/10.1109/CVPR.2011.5995733
- 8.Han D, Lee J, Lee J et al. A low-power deep neural network online learning processor for real-time object tracking application. IEEE Transactions on Circuits and Systems I: Regular Papers Online in advance. doi: https://doi.org/10.1109/TCSI.2018.2880363
- 10.Huang S, Hong J (2011) Moving object tracking system based on Camshift and Kalman filter. International Conference on Consumer Electronics, Communications and Networks XianNing, China, pp 1423–1426. doi: https://doi.org/10.1109/CECNET.2011.5769081
- 11.Jia X, Lu H, Yang MH (2012) Visual tracking via adaptive structural local sparse appearance model. IEEE Conference of Computer Vision Pattern Recognition, Providence, RI, USA: 1822–1829. doi: https://doi.org/10.1109/CVPR.2012.6247880
- 18.Oron S, Bar-Hillel A, Levi D et al. (2012) Locally orderless tracking. IEEE Conference on Computer Vision and Pattern Recognition, Providence, RI, USA: 1940–1947. doi: https://doi.org/10.1109/CVPR.2012.6247895
- 21.Sevilla-Lara L, Learned-Miller E (2012) Distribution fields for tracking. IEEE Conference on Computer Vision and Pattern Recognition, Providence, RI, USA: 1910–1917. doi: https://doi.org/10.1109/CVPR.2012.6247891
- 23.Song Y, Ma C, Gong L, Zhang J, Lau HWR, Yang M (2017) CREST: convolutional residual learning for visual tracking. IEEE International Conference on Computer Vision (ICCV), Venice, Italy: 2574–2583. doi: https://doi.org/10.1109/ICCV.2017.279
- 25.Wang JT, Yang JY (2007) Object tracking based on Kalman-mean shift in occlusions. J Syst Simul 19(18):4216–4220Google Scholar
- 26.Wu Y, Lim J, Yang MH (2013) Online object tracking: a benchmark. IEEE Conference on Computer Vision and Pattern Recognition, Portland, OR, USA: 2411–2418. doi: https://doi.org/10.1109/CVPR.2013.312
- 28.Yan JH, Chen SH, Ai SF et al (2014) Target tracking with improved CAMShift based on Kalman predictor. J Chin Inert Technol 22(4):536–542Google Scholar
- 31.Zhang T, Ghanem B, Liu S, Ahuja N (2012) Robust visual tracking via multi-task sparse learning. Proc IEEE Conference of Computer Vision Pattern Recognition, Providence, RI, USA: 2042–2049, . doi: https://doi.org/10.1109/CVPR.2012.6247908