Cramér-Rao Bound for Line Constrained Trajectory Tracking
ICASSP(2018)
摘要
In this paper, target tracking constrained to short-term linear trajectories is explored. The problem is viewed as an extension of the matrix decomposition problem into low-rank and sparse components by incorporating an additional line constraint. The Cramér-Rao Bound (CRB) for the trajectory estimation is derived; numerical results show that an alternating algorithm which estimates the various components of the trajectory image is near optimal due to proximity to the computed CRB. In addition to the theoretical contribution of incorporating an additional constraint in the estimation problem, the alternating algorithm is applied to real video data and shown to be effective in estimating the trajectory despite it not being exactly linear.
更多查看译文
关键词
Cramer-Rao bound,sparse methods,low-rank,augmented Lagrange Multiplier method,object tracking
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要