一种尺度估计的Mean Shift目标跟踪研究
|
摘要
目标跟踪技术被广泛应用于各行各业,特别是在摄像行业中,稳定的目标跟踪有利于提升拍摄效果。而在实际应用中,由于摄像机与目标可能均处于动态中,需要考虑因相对距离而产生的目标影响尺度变化问题。针对Mean Shift算法中尺度不变的缺点,本文采用一种尺度自适应的方式对目标进行尺寸估计,利用椭圆特征来调整核函数的参数,并以当前帧的参数作为下一帧的初始数据,达到自适应调节目标尺度的效果。通过实验验证了算法的有效性和稳定性。
Target tracking technology is widely used in all walks of life,especially in the camera industry.Stable target tracking is conducive to improving the shooting effect.In the practical application,since the camera and the target may both be in dynamic state,the scale change of the target due to the relative distance needs to be considered.Aiming at the defect of scale invariability in Mean Shift algorithm,a scale adaptive method is used to estimate the size of the target,the elliptical feature is used to adjust the parameters of kernel function,and the parameters of the current frame is taken as the initial data of the next frame to achieve the effect of adaptive adjustment of target scale.The validity and stability of the algorithm are verified by the experiment.
Target tracking technology is widely used in all walks of life,especially in the camera industry.Stable target tracking is conducive to improving the shooting effect.In the practical application,since the camera and the target may both be in dynamic state,the scale change of the target due to the relative distance needs to be considered.Aiming at the defect of scale invariability in Mean Shift algorithm,a scale adaptive method is used to estimate the size of the target,the elliptical feature is used to adjust the parameters of kernel function,and the parameters of the current frame is taken as the initial data of the next frame to achieve the effect of adaptive adjustment of target scale.The validity and stability of the algorithm are verified by the experiment.
引文
[1]管春苗.基于机器视觉的运动目标轨迹跟踪技术研究[D].沈阳:沈阳理工大学,2014.
[2]于金霞,许景民.多特征融合的自适应性粒子滤波跟踪算法[J].计算机工程与应用,2014,50(18):178-181.
[3]尹仕斌,任永杰,刘涛,等.机器视觉技术在现代汽车制造中的应用综述[J].光学学报,2018,38(8):1-11.
[4]沈豪,庄建军,郑茜颖,等.一种基于Mean Shift算法的目标跟踪系统的设计与实现[J].电子测量技术,2018,41(14):11-15.
[5]张颖,李孟歆,英宇.智能监控系统中自适应抗遮挡目标跟踪[J].现代电子技术,2018,41(17):59-61.
[6]朱闻亚.结合稳健估计和Mean Shift的视频目标跟踪算法[J].沈阳工业大学学报,2017,39(2):177-182.
[7]彭宝.基于机器视觉辅助驾驶系统中行人实时检测跟踪研究[D].上海:东华大学,2016.
[8]周传利.司机疲劳监测系统中眼睛检测与跟踪[D].西安:西安电子科技大学,2008.
[9]王佳.基于图像匹配的单目移动机器人自适应粒子滤波定位算法研究[D].秦皇岛:燕山大学,2016.
[10]张铁明.基于MeanShift的视频目标跟踪算法研究[D].秦皇岛:燕山大学,2016.
[11]TomasVojir,Jana Noskova,Jiri Matas.Robust scale-adaptive mean-shift for tracking[J].Pattern Recognition Letters,2014(49):250-258.
[2]于金霞,许景民.多特征融合的自适应性粒子滤波跟踪算法[J].计算机工程与应用,2014,50(18):178-181.
[3]尹仕斌,任永杰,刘涛,等.机器视觉技术在现代汽车制造中的应用综述[J].光学学报,2018,38(8):1-11.
[4]沈豪,庄建军,郑茜颖,等.一种基于Mean Shift算法的目标跟踪系统的设计与实现[J].电子测量技术,2018,41(14):11-15.
[5]张颖,李孟歆,英宇.智能监控系统中自适应抗遮挡目标跟踪[J].现代电子技术,2018,41(17):59-61.
[6]朱闻亚.结合稳健估计和Mean Shift的视频目标跟踪算法[J].沈阳工业大学学报,2017,39(2):177-182.
[7]彭宝.基于机器视觉辅助驾驶系统中行人实时检测跟踪研究[D].上海:东华大学,2016.
[8]周传利.司机疲劳监测系统中眼睛检测与跟踪[D].西安:西安电子科技大学,2008.
[9]王佳.基于图像匹配的单目移动机器人自适应粒子滤波定位算法研究[D].秦皇岛:燕山大学,2016.
[10]张铁明.基于MeanShift的视频目标跟踪算法研究[D].秦皇岛:燕山大学,2016.
[11]TomasVojir,Jana Noskova,Jiri Matas.Robust scale-adaptive mean-shift for tracking[J].Pattern Recognition Letters,2014(49):250-258.