首页>
外国专利>
METHOD FOR COLLATING OBJECTS REFLECTED IN TWO-DIMENSIONAL IMAGE AND SPECIFIC THREE-DIMENSIONAL IMAGE
METHOD FOR COLLATING OBJECTS REFLECTED IN TWO-DIMENSIONAL IMAGE AND SPECIFIC THREE-DIMENSIONAL IMAGE
展开▼
机译:二维图像和特定三维图像中反映对象的匹配方法
展开▼
页面导航
摘要
著录项
相似文献
摘要
PROBLEM TO BE SOLVED: To provide a method and a device for narrowing down persons which efficiently extract a person (candidate) presumed to be an appropriate person, for making identification by comparing and collating a two-dimensional video image or facial photograph data with three-dimensional data on the head of a known person by superimposition (overlaying) or the like.;SOLUTION: The method for narrowing down persons obtains distances designated arbitrarily from three-dimensional data and two-dimensional data on a person, on the basis of arbitrary points which can be identified in terms of somatology, physiognomy, and morphology, and extracts a person presumed to be the appropriate person from the values thereof, or the method for narrowing down persons calculates an index from the obtained distances and extracts the person presumed to be the appropriate person from the values of the index. A difference between the index calculated from the three-dimensional data and the index obtained from the two-dimensional data from a surveillance camera or the like is thought to be nearly 0 (zero) in the case of the person in question, and the same goes for adding up more than once. Accordingly, a person having a small difference in the indexes is regarded as the appropriate person, and, in a short period of time, persons presumed to be the appropriate person can be narrowed down to several persons.;COPYRIGHT: (C)2005,JPO&NCIPI
展开▼