结合改进Surf特征的素描人脸识别  

Sketch face recognition combining with improved Surf feature

作　　者：曹林[1] 李猛[1] 张华 CAO Lin;LI Meng;ZHANG Hua(Department of Telecommunication Engineering,Beijing Information Science and Technology University,Beijing 100101,China)

机构地区：[1]北京信息科技大学通信工程系,北京100101

出　　处：《计算机工程与应用》2018年第21期188-193,216,共7页Computer Engineering and Applications

基　　金：国家自然科学基金(No.61671069)

摘　　要：素描人脸识别属于异质人脸识别范畴,是刑侦领域的研究热点。根据素描人脸识别的特点,对已配准的人脸图像进行伪素描转化,并用Surf算法提取伪素描图像对的特征点。对经过提取后的伪素描特征点进行坐标邻域一致性优化,排除坐标邻域相对位置不一致的特征点,最后统计伪素描图像对的有效特征点,以实现识别的目的。利用现有的素描人脸库,进行实验验证,在选取50个特征点时的识别率达到99%,验证了算法的有效性。该算法经优化后,可用于素描人脸识别。The sketch face recognition belongs to the heterogeneous face recognition,and it is the research hot-spot in criminal investigation.According to the characteristics of sketch face recognition,the pseudo sketch transformation is used for the registered image,and the feature points are extracted by using Surf algorithm.Coordinate adjacent area consistency optimization have been proposed to eliminate the feature points of relative position inconsistent,and effective points of pseudo sketch image are calculated to achieve the purpose of recognition.The experiment are tested by using the existing sketch database,and the recognition rate is 99%when 50 feature points are selected.The result can prove that this algorithm is a new effective way in sketch face recognition.This method can be applied in sketch face recognition after optimization.

关 键 词：素描人脸识别 伪素描 SURF算法 坐标邻域优化 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]