Research On 3D Face Data Processing Algorithm And System Design

Face data contains abundant characteristic information,it is the only easy to see easy to obtain and characteristics of the human body contains rich differentiation characteristics of the data,thus has important application value in more areas,including face recognition in the field of safety and crowd detection and target tracking,facial prediction and reconstruction in the field of forensic medicine,anthropology and evaluation,such as level of aging,appearance of inversion or prediction,diseases,environmental and genetic factors in the field of bioinformatics fund association,etc.And 3D face data from 2D available information more rich,it contains information such as geometry,topology,with the popularity of 3D data acquisition device 3D face data acquisition is becoming more and more convenient,use 3D scanners can quickly get to the three-dimensional point cloud data,and even the use of mobile phones also can get the 3D face data,therefore,more and more application of 3D face data as the gold standard.At present,digital geometric processing technology is developing rapidly,among which point rendering,grid reconstruction and feature extraction technologies have been very mature and are widely used in graphics.These technologies can be used to mine the important information contained in 3D face data and obtain rich and in-depth face phenotypic characteristics.In today's most researchers only when applying 3D face data using a single discrete point information,part of the application of the 3D face point cloud only as a three dimensional discrete point set,using statistical methods to obtain characteristics,is more essential geometrical topology characteristics using less or inadequate,so the use of digital geometry processing technology in 3D face data contains the essence of information is of great significance to the related application field,so the design and development procedure of dealing with a complete and unified framework of the system is very necessary,can improve the efficiency of application field study.Therefore,I intend to generate and convert 3D point cloud data into triangular grid data,design and study a set of face grid data processing algorithm,and make a contribution to the generation of a complete set of digital geometric processing system.The specific process is as follows: first,select some key points in the face of data,and comprehensive use of Delaunay triangulation with triangular mesh subdivision algorithm,to obtain scientific and reasonable triangular mesh data,or by splitting of 3D face data point cloud data,after using a similarity algorithm based on area,distance,dihedral Angle similarity comparison,using HSV color gamut visual similarity degree.This algorithm can basically satisfy the comparison of facial mesh similarity.Finally,the original point cloud data is returned,and the 3D point cloud data is segmented with feature points as the center and Voronoi diagram as the reference.Finally,these algorithms are summarized and designed into a system.