CAJ | 학술논문

光照变化严重影响了人脸图像的外貌,这阻碍了人脸识别的过程.为了识别可变光照下的人脸图像,提出了一种基于小波的光照归一化算法,将一幅图像分解为低频成分和高频成分,对不同的频带成分进行不同的操作--对低频成分进行直方图均衡化,同时着重强调细节(高频成分),将它乘以一个标量从而增强图像边缘;对修改后的系数进行逆小波变换就得到归一化后的图像.最后,直接使用PCA方法对单训练样本条件下的人脸进行识别,在AR和FERRET人脸库上的实验结果表明,提出的方法可以显著提高人脸识别系统的识别率.
광조변화엄중영향료인검도상적외모,저조애료인검식별적과정.위료식별가변광조하적인검도상,제출료일충기우소파적광조귀일화산법,장일폭도상분해위저빈성분화고빈성분,대불동적빈대성분진행불동적조작--대저빈성분진행직방도균형화,동시착중강조세절(고빈성분),장타승이일개표량종이증강도상변연;대수개후적계수진행역소파변환취득도귀일화후적도상.최후,직접사용PCA방법대단훈련양본조건하적인검진행식별,재AR화FERRET인검고상적실험결과표명,제출적방법가이현저제고인검식별계통적식별솔.
The appearance of a face image is severely affected by illumination conditions that hinder the automatic face recognition process.To recognize faces under varying illuminations,a wavelet-based normalization method is proposed so as to normalize illuminations.An image is decomposed into its low frequency and high frequency components.Then different band coefficients are manipulated separately.Histogram equalization is applied to the approximation (low frequency) coefficients and at the same time accentuate the detail (high frequency) coefficients by multiplying by a scalar so as to enhance edges.A normalized image is obtained from the modified coefficients by inverse wavelet transform.Finally,PCA method is used to recognize normalized image with only one training sample.The experimental results obtained by testing on the AR face database and FERRET face database demonstrate the effectiveness of this method with significant improvement in the face recognition system.