一种去除人脸图像中的眼镜的方法

摘要

In face recognition , eyeglasses are the most common occluding object in facial images , which may greatly degrade the recognition performance .This paper describes an algorithm to remove eye-glasses from frontal facial images .The input facial image was reconstructed by two -dimensional image principle component analysis (2DIMPCA) and then the region occluded by the eyeglasses was obtained by comparing the reconstructed facial image and the input facial image and then binarization .An eyeglasses facial image was synthesized by recursive error compensation .Test results show that the method can effec-tively synthesize eyeglassless facial images and better than the method of traditional PCA .%在人脸识别中，眼镜作为人脸上常见的遮挡物对人脸图像的识别率有很大的影响。该文提出了一种去除人脸图像眼镜的方法。利用二维广义主成分分析（2 DIMPCA ）算法对戴眼镜的人脸进行重建，通过与输入的原始人脸图像的对比，经过二值化运算，提取出眼镜遮挡区域；通过误差迭代的方式对图像进行误差补偿，合成最终的无眼镜的人脸图像。实验结果表明，该算法可以得到比较自然的无眼镜人脸图像，优于传统的PCA方法。