基于混合高斯模型的运动车辆检测算法研究

Study of Moving Vehicle Detection Algorithm Based onGaussian Mixture Model

在军事领域,对轮式、履带式及坦克车辆等运动车辆进行高效的检测及跟踪,使军队可以大范围、全天候捕捉目标,对感知战场态势、侦察、决策指挥、评估威胁以及精确打击具有重要意义。主要研究检测运动车辆的方法,对常用几种方法的原理、优缺点进行了分析和对比,最终得到优化算法——融入了三帧差分法的基于混合高斯模型的背景差分法。从处理速度和检测前景精度等方面评估了3种算法,通过专项试验,验证并取得了一定的检测效果。实验结果表明,提出的优化算法在处理速度和检测效果两方面均有很大的提升。

In the military field,efficient detection and tracking of moving vehicles such as wheeled,tracked,and tank and other vehicles allows the military to capture targets on a large scale and all-weather,which is important for sensing the battlefield situation,reconnaissance,decision-making and commanding,assessing threats,and precision strikes.The methods of detecting moving vehicles are mainly studied,the principles,advantages and disadvantages of several commonly used methods are analyzed and compared,and finally the optimization algorithm——The background difference method based on the Gaussian mixture model incorporating three-frame difference is obtained.Three algorithms are evaluated in terms of processing speed and detection prospect accuracy,etc.the detection effect is verified through special experiments.The experimental results show that the improved algorithm proposed has greatly improved on the two aspects of processing speed and detection effect.