说明：  学位论文；运动物体跟踪方法主要包括卡尔曼滤波，Mean-shift，Camshifi算法，粒子滤波器，Snake模型等；应用卡尔曼滤波方法设计了一套煤矿矿工出入自动监测系统；提出了一种新的基于高斯混合模型的颜色特征提取方法，该方法克服了现有的Camshift算法Continuousl y Adaptive eanshift中跟踪目标特征提取精确度低和计算复杂度高的缺陷
(Dissertation moving object tracking methods include Kalman filtering, Mean-shift, Camshifi algorithm, particle filter, Snake model the application of Kalman filtering method designed a coal miners out of automatic monitoring system a new Gaussian mixture model based on the color feature extraction method to overcome the existing Camshift algorithm Continuousl y Adaptive eanshift track target feature extraction accuracy and low computational complexity and high defects)