三维头部区域的跟踪与深度估计方法研究通信与信息系统专业论文.docxVIP

山东大学硕士学位论文3．4云台控制系统 山东大学硕士学位论文 3．4云台控制系统 ．．36 3．5映射算法 。37 3．6实验结果与分析 39 3．7本章小结 43 第四章基于单幅正面人脸图片的深度估计 ．44 4．1基于2D特征点的深度估计系统设计 44 4．2 2D人脸数据库建立及使用 ．45 4．3 2D特征点提取及三角划分 ．46 4．3．1 2D特征点提取方法 46 4．3．2基于Delaunay的三角剖分方法 47 4．4基于三角相似的头部深度估计方法 。48 4．5实验结果与分析 。49 4．6本章小结 。50 第五章总结与展望 51 5．1总结 ．5 l 5．2未解决的问题和以后工作展望 ．．52 jjl；：jjl5}’：j2：：119C 53 致 攻读学位论文期间发表的学术论文目录 61 万方数据 山东大学硕士学位论文C0lNTENTS 山东大学硕士学位论文 C0lNTENTS ABSTRACT(In Chinese) ．．I 』jljljISTRACT ．H Symbol D憾cription IV Chapter 1 Overview ．．1 1．1 Subject Background and Significance ． ． 1 1．2 Research Smms ． ． ． ．3 1．2．1 Head Detection and Tracking Development Status ． ．3 1．2．2 Facial Feature Depth Estimation Development Status 9 1．2．3 Kinect and Depth Image ．．1() 1．3 Main Content ofthe Thesis ．14 1．4 Organization ofthe Thesis 15 Chapter 2 Head Detection Based on Depth Image ．16 2．1 Design of Head Detection System ．1 6 2．2 Preprocess ofDepth Image ．．16 2．3 Distance Transform and Application ． ．17 2．4 Head Detection Based on Template Matching 。1 9 2．4．1 Optimal Correlation Threshold ．21 2．4．2 Optimal Stride Threshold 24 2．5 Selecting Seed—Point Strategy 26 2．6 Region Grow Algorithm 28 2．7 Experiment and Data Analysis ．．28 2．8 Summary ofThis Chapter ．31 Chapter 3 Head Tracking System Based on Depth Image ．32 ：；．1 Design ofHead Tracking System ．．32 3．2 Head Detection Based on Frame Difference Algorithm． ． ． 33 3．3 Head Tracking Based on Modified Mean-shift ． ． 35 万方数据 山东大学硕士学位论文3．4 山东大学硕士学位论文 3．4 Stepper Motor Controll System ．36 3．5 Mapping Algorithm ．．37 3．6 Experiment and Data Analysis ．．39 3．7 Summary ofThis Chapter ．43 Chapter 4 Design of Feature Depth Estimation System ．．44 4．1 Design of Depth Estimation System Based on 2D Features ．．44 4．2 Establish and Application of 2D Face Database ． ．．45 4．3 Extract 2D Feature and Triangulation ．46 4．3．1 Extract 2D Facial Features ．46 4．3．2 Delaunay Triangulation 47 4．4 Depth Estimation Based on Triangle S