• To obtain structure and camera motion from an image sequence • Utilize inter-picture dynamics of a sequence – Such as constant speed, acceleration of the camera etc. Part 0 Basic concept of Kalman filter Part 1 Introduction to Kalman filter (KF) and Extended Kalman filters (EKF) Part II Two-pass Kalman filter for structure and pose estimation Description Major reference [Yu, June 2005] Part B Interacting Multiple Model Method (IMM) structure and pose estimation Major reference [Yu Aug.2005] Part C USE Trifocal tensor (TRI-KSFM) structure and pose estimation Major reference [Yu Feb.2005]

提出一种基于灰度信息和支持向量机的人眼检测方法.首先,利用人眼区域灰度变化比人脸其他部位灰度变化明显的特征,采用图像灰度二阶矩(方差)建立人眼方差滤波器,在固定人眼搜索区域内,应用人眼方差滤波器搜索候选人眼图像;然后,使用训练的支持向量机分类器精确检测人眼区域位置;最后,采用图像灰度信息率定位人眼中心(虹膜中心).该方法在BioID、FERET和IMM人脸数据库中的测试结果显示:没有佩戴眼镜人脸图像正确率分别为98.2%、97.8%和98.9%,406幅佩戴眼镜人脸图像正确率为94.9%;人眼中心定位正确率分别为90.5%、88.3%和96.1%.通过与目前方法比较,该方法获得较好的检测效果

BIOMATERIALS-TISSUE INTERACTIONS BIOMATERIAL TISSUE Strength ize scale IOnm 100nm lum 10 um 100 um Imm Modulus of Elasticity Fracture mechanics