摘要
针对篮球比赛中的运动目标跟踪问题,提出了一种基于DSP和FPGA的嵌入式篮球运动视频目标跟踪算法,将Mean-Shift算法用于跟踪篮球运动视频目标,并给出了DSP-FPGA的硬件设计.该方法先建立一个包含运动目标的搜索窗,并初始化其位置和大小,进一步计算搜索窗的色彩直方图.然后,输入待搜索图像,运行Mean-Shift,获得新的搜索窗位置和大小;最后,基于Mean-Shift算法迭代,获得运动目标的新位置.FPGA方便电路设计,DSP则运行目标跟踪算法.实验证明,提出的方法在激烈的比赛中依然能够准确跟踪运动目标.
Aiming at the problem of moving target tracking in basketball game,an embedded basketball video target tracking method based on DSP and FPGA is proposed.The Mean-Shift algorithm is used to track basketball video targets,and the DSP-FPGA hardware is designed.The method first establishes a search window containing moving objects,and initializes its position and size to further calculate the color histogram of the search window.Then enter the image to be searched,run Mean-Shift to obtain the new search window position and size.Finally,iterate based on the Mean-Shift algorithm to obtain the new position of the moving target.FPGA facilitates circuit design,and DSP runs target tracking algorithms.Experiments show that the proposed method can still accurately track moving targets in a fierce competition.
作者
杨彬
王同喜
YANG Bin;WANG Tong-xi(School of Physical Education,Yangtze University,Jinzhou 434020;School of Computer Science,Yangtze University,Jinzhou 434020 China)
出处
《湘潭大学学报（自然科学版）》
CAS
2018年第6期104-108,共5页
Journal of Xiangtan University(Natural Science Edition)
基金
国家监管中心科研项目(2011235)
作者简介
通信作者:王同喜(1971—),男,湖北黄冈人,副教授.E-mail:qqaz891@163.com.
