Graphic Art

5 pages
6 views

Performance Evaluation of Feature Selection Methods for ANN Based Iris Recognition

of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
Performance Evaluation of Feature Selection Methods for ANN Based Iris Recognition
Transcript
  rlnrldonr i rlrstclrrssit dr nr i I CE TA G S 2 0 1 3 u n i q u e i n f o n m t i o n , b u t i t is v e r y c o m m o n t h a t th e y a r e tot a l l y o rp a r t i a l l y o c c l u d e d b y e y e l as h o r e y e l i d . T h i s a pp r o a c h r e d uc e s r em o v a l o f o c c l u s i o n b y e y e l a s h o r e y e l i d . F i g . 4 sh o w s t h e o u tp u t o f t h e s t r a t e g y . 6 0 : 1 0 0 I 1 5 0 I 20 0 ; 25 0 i 1 0 0 20 0 30 0 ( a ) w e 1 2 <J t o o - P i g . 3 . I r i s e d g e d e t e c t i o n ( a ) O r i g i n a l i m ag e ( b ) H i s t og r a m e q u a l i z a t i o n i m ag e , H o r i zo nt al l i n e p u ss e s t l i   o u g h . r , , ,   ce n t e r o f p u p il ) o f i m age , z i g - z a g l i l i e s h o w s p i x e li n t en s i t i e s a l o ng ho r i z o n t a l li n e , ( c ) A v e r age d i n t e n s i t i e s o f p i x e l s a l o n g hor i z o n t a l li n e r ,, , f o r r i g h t s i d e o f t h e p u p i l ( d ) Ir i s i m ag e w i t h b o u n da r i e s . ( h ) Fe at u r e S e l e c t i o n o f l r i s P a t t e r n : D u r i n g I ri s P a tt e   n f o r m a t i o n , d i m e ns i o n r e d uc t i o n w a s a l r e a d y a ch i e v e d b y is o l a t in g t h e a re a o f t h e i m a g e t h a t c o n t a i n s t h e c h a r a c t e r i s t i c i n   o r m a t i o n a b o u t an in d i v i d u a l . F o r i n s t a nc e , w e d e c r ea s e t h e a u t nb e r O f e l em e n t s i n t h e i n p u t v e c t o r f r o m 8 9 , 6 00 ( 3 2 0x 2 8 0 i ma g e ) t o 1 60 0 ( 4 0 x 4 0 i m a g e s ) , a 9 8 % re d uc t i o n i n s i z e . S t i l l a l a r g e p e r c e n t a g e o f t h e s e p i x e l s c o n t a i n r e d un d an t i n   o r ma t i o n , ma t h e m a t i c a l a n d s t a t i s t ic a l m e t h o d s c a n b e u t i l i z e d t o re d u c e t h e d i me n s i o n ali t y o f t h e p r ob l e m . T h i s p r o b l em i s a l s o k now na s F e a t u r e S e l e c t i o n , f r o m a s e t o f n f e a t u r e s , se l e c t a s u b s et o f m fea t u r e s t h a t l ea d s t o m i n i m u m c l a s s i f ic a t i o n e r r o r . S in g l e V a l u e D e co m p o si t i o n ( S V D) , { P r i nc i p a J C o m p o n e n t A n a l y s i s ( P C A ) a n d I n d e p e nd e n t C o m p o n e n t A n a l y s i s ( I C A ) a r e a pp l i e d t o I r i s P a tt e m f ort h i s p ur p os e . ( i ) S i n g l e V a l u e D e c o m p os i t i o n ( S VD ) : S i n g l e V a lu e D e c o m p o s i t i o n i s a p o w e r f u l ma t r i x t e c hn i q u e w i t h ma n y u s e f u l a p p l i c a t i o n s . T h e m a i n co n c e p t b e h i n d i t i s t o e x po s e t h e h i dd e ng e o m e t r y oft h e m a t r i x . W e a l s o b e n e   i t f r o mS V D u s i n g i t a s a d i me n s i o n r e d u c t i o n t o o l . T h e b a s i c o p e r a t i o n o f SV D r e l i e s o n t h e f a c t o r i za t i o n o f a n m x n m a t   i x - ( m > n ) I n t o t h r e e o t he r m a t r i c e s o f t h e   o l l o w i n g f o r m [ 2 1 ] : A = U S V T ( 1 ) w h e r e t h e s up e r scri p t   T   d e no t e s m a t r i x t r a n s po s e . U i s a n i ? i x m o   t ho gon a l m a t r i x , V i s a n n x n o r t h ogon a l m a t r i x a n d S i s a n m x n d i a gon a l m a t ri x w i t h S i j ~ 0 1 1   a n d S u > S i m i . T h e c o l u m n s o f V a r e r i g h t s i n g u l a r v e c t o r s a n d t h e c o l u m n s o f U a r e l e f t si n g u l a rv e c t or s . E i t h er t h e s i n g u l a r v e c t o r s o f t h e s in g ul ar v al u e s ( o r b o t h ) a r e c a l l e d p   i n c i p l e c om po n e n t s . S V D r e d u c e s t h e d i m e n s i o n o f i n p u t p a tt e   n f r o m m x n t o o n l y a v e c t o r o f n e l e m en t s   T h e   i r s t n e l e m e n t s 1 0 20 3 0 4 0 507 0 8 09 01 0 0 1 10 ( b ) P i g . 4 . I r i s P a l t e r n ( a) I r i s w i t h l i i s P att e r n ( b) E x t r a c t e d I   i s P m l m c o n t a i n s ub s t a n t i a l i n f o r m a t i o n , a n d w e c a n cr o p th e v e c t o r t a i l wi th ou t sig ni   i c a nt l o s s o f g e n e r a l it y . ( U ) P r i n c ip a l C o m p o n e n t A na l y s i s ( PC A ) : PC A i s a n uns u p e r v i s e d l i n e a r f e a t u r e e x t r a c t i o n a l go r i t h m ; i t i sa l i n e a r m a p p i n g w h i c h u s e s t h e e i g e n v e c t o r s wi t h th e l a r g e s t e i g en v a l u e s . T h e b a s i c i d e a f or P C A i s t o   i n d t h e co m po n e n t s s ; , s j , s s o t h a t t h e y e x p l a i n t h e ma x i m u m am o u n t o f v a r i a nc e p o s s i b l e by n l i n ea rl y t r a n s f or me d c o mp o n e n t s . T h e me t ho d o f s t u d y i n g t h e e i g e n v a l u e d e c o m p o s i t i o n o f th e c o v a r i a n c e m a t r i x i s k n ow n a s P r i n c i p al C o m pon e n t s A n a l y si s ( P C A ) [ 22 ]   L e t n m - d i m e n s i o n a] v e c t o r s X j i s a l i g n e d i n t h e d a t a m a t r i x X ( m * a ) a n d C i s t h e c o v a r i a nc e m a t   i x o f X . T h e n , S VD ( C ) w i l l b e t h e P C A o f X . L e t A b ea s q u a r e s y m m e t   i c n x n m a t r i x , a n d c o n s i d e r i t s e ig e nv a l u e d e c o m po s i t i o n w h e r e P i s a n o   t h o g o n a l ma t r i x w i t h c o l u m n s . . . , p an d A i s a d i a go n a l m a t r i x w i t h e i g e n v a l u e s X i , l > , -   l w h i c h a r e s o   t e d i n d e sc e n d i n g or d e r . / ; , i s t h e e i g e n v e c t o r c o rr e s p o n d i n g t o e i g e n v a l u e 1 , . ( H i ) I nd e p e nd e n t c o m p on e n t a n a l y s i s   I C A ) : I nd e p e nd e n t c o m p o n e n t a n a l y s i s ( I C A ) i s a s t a t is t i c a l t e c hn i q u e t h a t r e p r e s e n t s a m u l t i d i m e n s i o na l r a n do m v e c t o r a s a l i n ear co m b i n a t io n o f n o n - g a us s i an r a n do m v a r i a b l e s (   i n d e p e n d e n tc o m p o n e n t s   ) t h a t a r e a s i n d e p e n d e n ta s po ss i b l e . I C Ai s a no n - g a uss i a n v e r s i o n o f f a c t o ra n a l y s i s . I C A h a s m a n y a p p li c a t i on s i n d a t a a n a l y s i s , s o ur c e s e p a r a t i o n , a n d f e a t u r e ex t r a c t i o n . A J t h o u g h I n d e p e n d e n t C o m pon e n t A na l y s i s ( I C A ) a n d i t s m a n y v a ri a n t s w e r e o r i g in a l l y c o nc e i v e d f o r b lin d s o u r c e s e p a r a t i o n ( B SS ) , m a n y s t u d i e s w e r e c ond u c t e d o n t h e a p p l i c a t i o n o f I CA t o f ea t u r e e x t r a c t i o n f r o m t i m e s e r i e s a n d i m a g e s [ 2 3 ] . I t a i m s t o f i n d a s e t o f i n d e p e n d e n t s o u r c e s t h a t c a p t u r e t h e u n d e r l y i n g r a ndo m n e s s o f t h e ob s e r v e d s i g n a l s . - 2 1 0 -
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks