基于伪Zernike矩归一化的人脸特征提取方法

以改进的伪Zernike矩相关知识为基础，提出了一种新颖的人脸识别特征提取方法。该方法在识别时不需要进行几何校正，降低了人脸识别系统的算法复杂度。计算人脸图像的伪Zernike矩，对其进行归一化处理，采用最小邻近分类器进行判别。实验结果表明，该算法不受人脸尺度和旋转变化的影响，并且对表情、姿态和光照具有较强鲁棒性。进一步探讨了伪Zernike矩在人脸识别方面的应用。

Face feature extraction based on normalization of pseudo-Zernike moment

A novel face image feature extraction based on normalization of pseudo-Zernike moment is proposed in this paper.It dose not require geometry correction of face image when recognizingr,educing the algorithm complexity of face recogni-tion system.Firstlyt,he pseudo-Zernike moments of the face image are computed.Secondlyt,he result is normalized.At lastt,he nearest neighborhood algorithm is used to construct classes.Experimental results demonstrate the presented approach basically eliminates the effects of the change of face scale and rotationa,nd has better robust to variations of illumination,pose and fa-cial expression.On the basis of previous studyt,he approach makes a further discussion on the application of pseudo-Zernike moments in the aspect of face recognition.