基于光谱匹配滤波的线结构光扫描人脸轮廓三维成像系统  

Three-dimensional Imaging System Based on Spectral Matched Filtering for Line Structured Light Scanning Face Contour

作　　者：李琼[1] LI Qiong(Electronic Information Engineering College,Wuhan Polytechnic,Wuhan430074,China)

机构地区：[1]武汉职业技术学院电子信息工程学院,湖北武汉430074

出　　处：《武汉职业技术学院学报》2020年第5期112-115,共4页Journal of Wuhan Polytechnic

基　　金：2017年湖北省教育厅科学技术研究项目“基于标准光谱的线结构光光条中心的提取方法研究”(项目编号:B2017515)。

摘　　要：设计了一种基于光谱匹配滤波的线结构光扫描成像系统。系统采用近红外线结构光照明,CMOS工业相机通过光谱匹配滤波捕捉反映人脸截面轮廓边缘的扭曲条纹,有效降低了杂散光和背景景物影响,利用局域高斯参数匹配实现了光条中心快速提取,结合一维扫描运动,实现了人脸轮廓的三维成像,并通过实际人脸轮廓扫描实验验证了系统设计的有效性。The scanning imaging system using line structure light based on spectral matching filter is designed for more and more demand of face profile reconstruction.The system uses near-infrared structured light illumination,and the CMOS industrial camera captures the distorted fringes reflecting the edges of the face profile through spectral matching filtering,effectively reducing the influence of stray light and background scenery.The local Gaussian parameter matching is used to realize the rapid extraction of the center of the light strip,combined with the one-dimensional scanning motion,to realize the three—dimensional imaging of the face contour,and the actual face contour scanning experiment verifies the effectiveness of the system design.

关 键 词：三维成像 线结构光 人脸轮廓 光谱匹配滤波 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]