摘要
为了解决当前在复杂背景下视频目标存在跟踪提取不准确与不稳定等不足,论文设计了基于CamShift与混合高斯模型的目标跟踪与提取算法。首先,耦合MeanShift与CamShift算法,定义了目标跟踪算子,完成对目标的精确定位;然后计算图像像素灰度的均值与方差,设计混合高斯模型,实现定位目标与背景的分离;最后基于OpenCV开源函数,利用C++编程实现算法跟踪与提取功能。实验数据显示:与当前目标跟踪算法相比,在面对复杂背景条件下的运动目标跟踪提取时,论文技术拥有更高的追踪精度与稳定性。
In order to solve the current inaccurate and unstable deficiencies under the background of complex video target tracking extraction,the CamShift and target tracking of the gaussian mixture model and extract algorithm is designed.First of all,coupled with MeanShift and CamShift algorithm,the target tracking operator is defined,the accurate positioning of target is completed.Then based on pixel gray mean and variance of the image,the gaussian mixture model is designed and localization of target and background separation is realized.Finally,based on OpenCV open-source function C++ programming is used to realize algorithm tracking and extracting features.Experimental data shows that,compared with the current target tracking algorithm,in the face of the moving target tracking under complex background extraction,the technology in this paper has higher tracking accuracy and stability.
出处
《计算机与数字工程》
2016年第6期1053-1056,1063,共5页
Computer & Digital Engineering
