期刊名称：International Journal of Advanced Research In Computer Science and Software Engineering
印刷版ISSN：2277-6451
电子版ISSN：2277-128X
出版年度：2012
卷号：2
期号：1
出版社：S.S. Mishra
摘要：The objective of this paper is to design an automated face recognition system which can be used in application like recording of daily attendance. In this paper fully automatic recognition of frontal 2D face image of an individual has been described. Firstly, all the 2D frontal faces from FRAV3D image database are selected and by applying SUSAN (Smallest Univalue Segment Assimilating Nucleus) method all the images are converted into grayscale images for detection of the edges .After that the method which is represented here is used to extract n (n=4) feature points(which are expression, light, size invariant ). But in this method only different frontal facial expressions have been considered. For face recognition of every individual only three ratios in terms of database have been stored. Experimental results show that frontal face images may be easily recognize and this technique can be used
关键词：SUSAN; ;Segmentation; Eye corner detection; ;Nostrils detection; Face recognition.; var currentpos;timer; function initialize() { timer=setInterval("scrollwindow()";10);} function sc(){clearInterval(timer); }function scrollwindow() { currentpos=document.body.scrollTop; window.scroll(0;++currentpos); if (currentpos != document.body.scrollTop) sc();} document.onmousedown=scdocument.ondblclick=initializeVolume 2; issue 1; January 2012 www.ijarcsse.com;. 2011; IJARCSSE All Rights Reserved ;border extraction [3] has been used as an alternative to ;artificial template matching. ;In this method ; rightmost point of right eye; leftmost point of ;left eye; rightmost point of right nostril; and leftmost point of ;left nostril have been extracted usin g proposed algorithm. ;II.;PRESENT METHOD ;This method is mainly divided into three parts. Firstly (A) ;SUSAN method have been applied to convert the 2D images ;into gray scale images; after that (B) segmentation algorithm ;and (C) eye corner extraction and nose corner extraction ;algorithm have been applied. ;II.A SUSAN operation ;Here SUSAN is used mainly for edge detection of the images; ;SUSAN places a circular mask over the pixel to be tested; ;which is termed as the nucleus [1]. The circular mask is moved ;through each point of the image; the intensity of each pixel ;within the mask is compared with that of the nucleus. A Simple ;equation determined this co mparison is as follo ws: ;.;.;.;.;.;.;.;.;t;t;r;I;r;I;if;r;I;r;I;if;r;r;c;);(;);(;1;);(;);(;0;0;0;0;);(;(1) ;A more stable and reliable form of the above equation is used ;in the follo wing form: ;.;.;6;);(;);(;0;0;.;.;.;.;.;. .;.;.;t;r;I;r;I;e;r;r;c;(2) ;where; I(r;0;) is the intensity of the nucleus; I(r)is the intensity ;of any other pixel within the mask; t is the gray level ;difference threshold and c(r;r;0;) is the output of the ;comparison. This comparison is done for each pixel within the ;mask; and a running total of the outputs c(r;r;0;) are computed as ;follows: ;.;.;.;);(;0;0;0;);(;);(;r;c;r;r;r;c;r;n;(3) ;Setting area threshold as g ; and comparing it with ;USAN's area to get the initial corner resp onse function: ;R;(4) ;where; g is named the `geometric threshold' or SUSAN ;threshold . So from the above equation it can be stated that if ;the SUSAN area is small then it will get positive value for ;R(r;0;) otherwise it will be 0 as the value of g is set to be 27 ;generally[1]. This concept of each image point having ;associated with it a local area of similar brightness is the basis ;for the SUSAN principle. ;For corner detection; two further steps have been used. Firstly; ;the centroid;of the nucleus has been identified. A proper corner ;will have the centroid far from the nucleus. The second step ;insists that all points on the line from the nucleus through the ;centroid out to the edge of the mask are in the SU SAN. ;Figure 1: Four circular masks at different places on a simple image. ;Figure 2: Four circular masks with similarity colouring; USANs are shown as ;the white parts of the masks. ;After converting the image into gray scale; this method is ;applied to extract the feature points. And then vario us image of ;a person with various expressions are experimented. ;3.a ;3.a;3.b;Application of ;the SUSAN ;algorithm
