作　　者： ; ; ;

机构地区： 华南理工大学电子与信息学院

出　　处： 《计算机仿真》 2007年第3期207-209,共3页

摘　　要： 针对人脸轮廓跟踪,提出一种约束的B-Snake模型的非线性粒子滤波新算法。该算法引入轮廓的主动模型,采用B -Snake表示人脸轮廓,通过建立人脸轮廓变形的形状空间,对人脸轮廓的可变形范围进行约束。同时,利用非线性滤波方法在人脸轮廓的变形区域内进行搜索,求解人脸轮廓参数,进行人脸轮廓跟踪。运用人脸的运动模型进行了跟踪仿真实验,仿真结果表明,该方法有效地解决了复杂背景下人脸自由运动、光照变化及部分遮挡的问题,具有较高的鲁棒性。 Aiming at face contour tracking, an algorithm of non - linear particle filter with restricted B - Snake model was presented . The method introduced the contour initiative model, and made B - Snake to express face contour, the deformable range of face contour was restricted by building the form space of face contour. In tracking, the area of face contour distortion were searched by the method of particle filter, and the face contour parameters were solved. Simulator experimental results showed the algorithm＇ s robustness to the agile motion of face, the change of illumination and partial occlusion in the presence of complex background.

关 键 词： 人脸跟踪 非线性粒子滤波 轮廓跟踪

领　　域： [自动化与计算机技术] [自动化与计算机技术]