Descriptor-based tracking algorithm using a depth camera
- 52 Downloads
The appearance of inexpensive high-quality video cameras and depth cameras resulted in the development of a large number of object-tracking algorithms. In this paper, a new descriptor-based algorithm for real-time object tracking using the information from a Microsoft Kinect depth camera is proposed. As a descriptor for the object tracked, histograms of oriented gradients calculated from the circular sliding regions of the scene image are used. The information on the depth of the scene is used when the image of the object of interest is partially occluded by other objects in the scene. To speed up the tracking process, a model for predicting the object motion is used. To ensure real-time tracking with the proposed algorithm, a multicore graphics processor is used.
Unable to display preview. Download preview PDF.
- 2.H. Schweitzer, J. Bell, and F. Wu, “Very fast template matching,” in Proc. 7th Eur. Conf. on Comput. Vis. (ECCV 2002), Copenhagen, Denmark, May 28–31, 2002 (Springer-Verlag, 2002), pp. 358–372.Google Scholar
- 3.S. Nejhum, J. Ho, and M. Yang, “Online visual tracking with histograms and articulating blocks,” in Comput. Vis. Image Underst., 2010, pp. 901–914.Google Scholar
- 6.F. Talu, I. Turkoglu, and M. Cebeci, “A hybrid tracking method for scaled and oriented objects in crowded scenes,” Expert Syst. Appl. 38, 13682–13687 (2011).Google Scholar
- 7.A. Buchanan and A. Fitzgibbon, “Document image dewarping using robust estimation of curled text lines,” Combining local and global motion models for feature point tracking, in Comput. Vision Pat. Recogn., 2007, pp. 1–8.Google Scholar
- 10.B. Babenko, Y. C. Ming-Hsuan, and S. Belongie, “Visual tracking with online multiple instance learning,” in Proc. IEEE Conf. on Comput. Vision and Patt. Rec. (CVPR 2009), Miami, Florida, June 20–25, 2009, (IEEE, New York, 2009), pp. 983–990.Google Scholar
- 11.S. Song and J. Xiao, “Cluster based weighted SVM for the recognition of Farsi handwritten digits,” in Tracking Revisited Using Rgbd Camera: Unified Benchmark and Base-Lines, (2013), pp. 233–240.Google Scholar
- 12.K. Meshgi, S. Maeda, S. Oba, and S. Ishii, “Fusion of multiple cues from color and depth domains using occlusion aware bayesian tracker,” IEICE Tech. Rep. Neurocomp. 114 (500), 127–132 (2014).Google Scholar
- 13.N. Dalal and B. Triggs, “Histograms of oriented gradients for human detection,” Comput. Vis. Patt. Rec. 1, 886–893 (2005).Google Scholar
- 14.D. Miramontes-Jaramillo, V. Kober, and V. Díaz-Ramírez, “CWMA: Circular window matching algorithm,” in Proc. 18th Iberoam. Cong. in Patt. Rec., 2013, LNCS 8258, pp. 439–446 (2013).Google Scholar
- 16.V. Gupta, Nonlinear Filters with Spatially Connected Neighborhoods (Laxmi Publications, 2005).Google Scholar
- 17.E. M. Ramos and V. Kober, “Design of correlation filters for recognition of linearly distorted objects in linearly degraded scenes,” J. OSA A 24, 3403–3417 (2007).Google Scholar
- 21.G. Takacs, V. Chandrasekhar, S. Tsai, R. Grzeszczuk, and B. Girod, “Distortion invariant pattern recognition with local correlations,” Fast Computation of Rotation- Invariant Image Features by Approximate Radial Gradient Transform, 22, No. 8, pp. 2970–2982 (2013).Google Scholar