摘要
针对户外环境光线和气候条件多变以及目标间相互遮挡对目标检测和跟踪的影响,提出了一种基于改进的高斯混合模型方法来检测运动目标,并消除噪声和阴影;同时采用基于Kalman滤波器的预测模型和最大后验概率目标匹配相结合的方法来实现目标的连续跟踪。实验表明,该方法能实现目标的稳定跟踪,且能够处理目标相互遮挡的情况,计算复杂度较低,基本满足实时应用的需求。
In order to overcome the adverse effect,which caused by the inconstant light and climate and occlusion between objects,on object detecting and tracking,a method based on the improved mixture Gaussian model is proposed to detect the moving object,then shadow and noise is removed.While using the method of predict model based on Kalman filter combines with maximum posterior probability for object matching it realizes the moving objects tracking.The experimental results indicate that the method can construct a robust real-time object tracking system which can easily handle the occlusion.
出处
《计算机工程与应用》
CSCD
北大核心
2011年第11期187-189,200,共4页
Computer Engineering and Applications