基于改进无序运动数学形态特征的弱目标跟踪  

Based on the Improved Disordered Motion Mathematical Morphological Characteristics of the Weak Target Tracking

作　　者：刘锡凤 

机构地区：[1]大连商业学校,辽宁大连116033

出　　处：《科技通报》2013年第2期221-223,共3页Bulletin of Science and Technology

摘　　要：在视频无序运动目标实时跟踪中,无序性的出现会使所跟踪目标的尺寸和色彩等外观线索失去强度可靠性,跟踪目标的弱化会导致对目标地错误跟踪。为了克服这一问题,本文提出了一种基于改进无序运动数学形态特征的弱目标跟踪算法。该算法根据数学形态模型,分割出图像中的可跟踪特征,进行弱化特征增强,利用卡尔曼滤波预测各目标是否遇到是真实跟踪模型,在目标较为弱化的情况下,可以计算最佳定位信息更新目标信息。计算机仿真结果表明,所提出算法能够在保证实时性的前提下,在目标较为弱化的情况下,均能实现准确跟踪,并且跟踪结果令人满意。In video disordered motion target real-time tracking,disordered sexual appear can make the tracking target size and color appearance clues to lose strength reliability,the tracking target can lead to the weakening of the target error tracking.In order to overcome this problem,this paper proposes a based on improved disordered motion mathematical morphological characteristics of the weak target tracking algorithm.The algorithm based on mathematical morphology model,the segmentation of the image can be tracking characteristic,carries on the weakening characteristic enhancement,using kalman filter forecast each target is true whether meet tracking model,the target is weakened circumstances,can calculate the best positioning information update target information.Computer simulation results show that the proposed algorithm can ensure real-time in the premise of the target is weakened cases,all can realize accurate tracking,and tracking the result is satisfactory.

关 键 词：目标跟踪 弱目标 卡尔曼滤波 数学形态特征 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]