STATISTICAL CHANGE DETECTION WITH MOMENTS UNDER TIME-VARYING ILLUMINATION

In this paper, an illumination-independent statistical change detectio n method is proposed. The proposed method consists of two parts. First , based on our defined circular shift moments, structural changes can be distinguished from those due to time-varying illumination in the no ise-free case, Moreover, the amount of computation is less than that o f the shading model method, Second, in the light of the characteristic s of the defined moments, a statistical decision rule is also proposed to cope with the effects of noise. The change detection problem can b e treated as one of hypothesis testing. Critical values can be chosen according to the desired level of significance. Experimental results i ndicate that the proposed method detects changes accurately in the tim e-varying illumination case.