基于AdaBoost及Kalman算法的人眼检测及跟踪.pdf

( ) 46 6 V o l. 46, N o. 6 JO U R N A L O F N A NJIN G U N IV ERSIT Y 2010 11 N ov.2010 ( N A T U RA L SCIEN CES) * AdaBoost Kalman ** 1 1 2 严 超 , 王元庆 , 张兆扬 ( 1. , , 2 1009 3 2 . , , 200444) : , , , . , , A daBoo st Ka lm an . ! . A daBo ost , , AdaBo ost , , K alm an , , . Window s X P, P entium I V, 5 12 M emor y, 24 G Hz , 640 ∀ 480 , 9 35% 10 m s/ , . . : , , AdaBo ost , Kalm an : T P 30 16 Eye detection andtracking basedon AdaBoost and Kalman algorithms 1 1 2 Yan Chao , Wang YuanQing , Zhang ZhaoYang ( 1. Depar tm ent of Electr onic Science and Engineer ing , Nanjing U niver sity, N anj ing, 21009 3, China 2 . K ey L abor ator y of Advanced Display and System A pplicatio n, Sh ang ha i U niver sity, M inistr y of Education , Sh ang h ai, 200444, China) Abstract: Lo cat ing and tracking the hum an eyes is o ne of the cr itical techno log ies in f ree stereosco pic display sy stem . O w ing to it s hig h academ ic and social value, ey e detectio n h as been becom ing a hot spot in p attern recog nition r ecently. Ey e det ection should be accur ate and r ealtim e. T o satisfy these technical needs, a new ey e