文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > IT/计算机 > Detection of a moving object in the image
侵权投诉

Detection of a moving object in the image

DETECTION OF A MOVING OBJECT IN THE IMAGE461

the ideal observer. The Bayes detection algorithmforms the value

Vmax/2

I=

–Vmax/2

∫

exp[L(V)]wpr(V)dV

(2.6)

and we compare it with the threshold p0/p1. If I > p0/p1,the decision on the presence of an object image in theobserved realization is made, and, otherwise, the deci-sion on the absence of the object image is made. As themaximum-likelihood detector, the Bayes detector ismultichannel. In addition, the technical implementationof the Bayes algorithm is a bit more complex than thatof the maximum-likelihood detector since it is neces-sary to integrate the function L(V) (2.2) after a nonlin-ear transformation.

3. CHARACTERISTICS

OF THE QUASI-PLAUSIBLE DETECTORUsually, the ef ciency of the detection algorithm isestimated by the values of the probabilities of the errorof the rst kind (false alarm) α and of the second kind(missing of the object) β (or the probability of correctdetection PD = 1 – β) [10–12]. Let us nd these proba-bilities for the quasi-plausible detection.

Substituting into (2.1) the realization of observeddata (1.1), we express the logarithm of the likelihoodratio functional in the form of the sum of the signal andnoise functions [13]

(V,V0)+N (V,V)/2, (V)–SL(V)=γ0S

(3.1)

where γ0 = 1, if the hypothesis H1 is true, and γ0 = 0 if

the alternative H0 holds, and2 (V,V0)=-----S

N0

T

∫∫∫[s(x–Vt,y)–ν(x,y)]

0

×[s(x–V0t,y)–ν(x,y)]×Is(x–Vt,y)Is(x–V0t,y)dxdydtis the signal function, and2 (V)=-----N

N0

T

(3.2)

∫∫∫n(x,y,t)[s(x–Vt,y)–ν(x,y)]

0

×Is(x–Vt,y)dxdydt

is the noise function that is a realization of a Gaussianrandom process with the zero mean value and correla-tion function

(V1,V2). (V2) =

Detection of a moving object in the image

S (V1)N N

(3.3)

JOURNAL OF COMPUTER AND SYSTEMS SCIENCES INTERNATIONAL Vol. 45 No. 3 2006

上一页第3页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email