视频跟踪的目标是将连续视频帧中的目标物体相关联。当物体相对于帧速率快速移动时,这种关联可能会特别困难。另一种情况是当被跟踪物体的方向随时间改变时,问题的复杂性会增加。对于这种情况,视频跟踪系统通常采用运动模型,以描述目标图像随着物体的不同可能运动会如何变化。
简单运动模型有:
为了实现视频跟踪,算法分析连续的视频帧,并输出不同帧之间物体的移动。目前存在各种各样的算法,各有优缺点。选择使用哪种算法时,考虑预期用途很重要。视觉跟踪系统有两个主要组成部分:目标表示和定位以及过滤和数据关联。
目标表示和定位主要是一个自下而上的过程。其提供了多种识别运动物体的工具。能否成功定位和跟踪目标取决于算法。例如,使用斑点跟踪对于识别人类运动是适用的,因为人的轮廓是动态变化的。[6] 通常来说这些算法的计算复杂度很低。以下是一些常见的目标表示和定位算法:
过滤和数据关联主要是一个自上而下的过程,其整合场景或物体的先验信息、处理物体运动以及评估不同的假设。这些方法适用于跟踪具有复杂物体交互的复杂物体,例如跟踪于障碍物后面移动的物体。[8] 此外,如果视频跟踪器(也称为TV跟踪器或目标跟踪器)非固联安装,而是安装在移动的平台上,人们通常会使用惯性测量系统来预稳定视频跟踪器,以减少相机系统所需的动态和带宽,这种情况下算法复杂性会增加。[9]过滤和数据关联算法的计算复杂度通常要高得多。以下是一些常见的过滤算法:
^Peter Mountney, Danail Stoyanov & Guang-Zhong Yang (2010). "Three-Dimensional Tissue Deformation Recovery and Tracking: Introducing techniques based on laparoscopic or endoscopic images." IEEE Signal Processing Magazine. 2010 July. Volume: 27". IEEE Signal Processing Magazine. 27 (4): 14–24. doi:10.1109/MSP.2010.936728..
^Lyudmila Mihaylova, Paul Brasnett, Nishan Canagarajan and David Bull (2007). Object Tracking by Particle Filtering Techniques in Video Sequences; In: Advances and Challenges in Multisensor Data and Information. NATO Security Through Science Series, 8. Netherlands: IOS Press. pp. 260–268. CiteSeerX 10.1.1.60.8510. ISBN 978-1-58603-727-7.CS1 maint: Multiple names: authors list (link).
^Kato, Hirokazu & Mark Billinghurst (1999). "Marker Tracking and HMD Calibration for a Video-based Augmented Reality Conferencing System" (PDF). IWAR '99 Proceedings of the 2nd IEEE and ACM International Workshop on Augmented Reality. IEEE Computer Society, Washington, DC, USA..
^"High-speed Catching System (exhibited in National Museum of Emerging Science and Innovation since 2005)". Ishikawa Watanabe Laboratory, University of Tokyo. Retrieved 12 February 2015..
^"Basic Concept and Technical Terms". Ishikawa Watanabe Laboratory, University of Tokyo. Retrieved 12 February 2015..
^S. Kang; J. Paik; A. Koschan; B. Abidi & M. A. Abidi (2003). "Real-time video tracking using PTZ cameras". Proc. SPIE. 5132: 103–111. CiteSeerX 10.1.1.101.4242. doi:10.1117/12.514945..
^Comaniciu, D.; Ramesh, V.; Meer, P., "Real-time tracking of non-rigid objects using mean shift," Computer Vision and Pattern Recognition, 2000. Proceedings. IEEE Conference on , vol.2, no., pp.142,149 vol.2, 2000.
^Black, James, Tim Ellis, and Paul Rosin (2003). "A Novel Method for Video Tracking Performance Evaluation". Joint IEEE Int. Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance: 125–132. CiteSeerX 10.1.1.10.3365.CS1 maint: Multiple names: authors list (link).
^Gyro Stabilized Target Tracker for Off-shore Installation.
^M. Arulampalam; S. Maskell; N. Gordon & T. Clapp (2002). "A Tutorial on Particle Filters for Online Nonlinear/Non-Gaussian Bayesian Tracking". IEEE Transactions on Signal Processing. 50 (2): 174. CiteSeerX 10.1.1.117.1144. doi:10.1109/78.978374..
^Emilio Maggio; Andrea Cavallaro (2010). Video Tracking: Theory and Practice. 1. Video Tracking provides a comprehensive treatment of the fundamental aspects of algorithm and application development for the task of estimating, over time..
^Karthik Chandrasekaran (2010). Parametric & Non-parametric Background Subtraction Model with Object Tracking for VENUS. 1. Background subtraction is the process by which we segment moving regions in image sequences..
^J. Martinez-del-Rincon, D. Makris, C. Orrite-Urunuela and J.-C. Nebel (2010). "Tracking Human Position and Lower Body Parts Using Kalman and Particle Filters Constrained by Human Biomechanics". IEEE Transactions on Systems Man and Cybernetics – Part B', 40(4)..
暂无