弱纹理人脸图像局部破损点修复方法

Weak Texture Face Image Local Damage Point Repair Method

为了增强弱纹理人脸图像的视觉效果,需要对人脸图像局部破损点修复方法进行研究。采用当前图像破损点修复方法对弱纹理人脸图像中存在的破损点进行修复时,存在弱纹理人脸图像分解精度低,修复效果差的问题。提出一种弱纹理人脸图像局部破损点修复方法,通过全变分将弱纹理人脸图像分为骨架部分和纹理部分,采用贝叶斯压缩感知得到骨架部分和纹理部分稀疏系数的分布函数,计算骨架部分和纹理部分分布函数的方差和均值,根据弱纹理人脸图像骨架部分和纹理部分的方差和均值得到修复后的人脸图像,完成弱纹理人脸图像局部破损点的修复。仿真结果表明,所提方法对弱纹理人脸图像分解时的精度高,得到的弱纹理人脸图像修复效果好。

In order to enhance the visual effect of face image with weak texture,a method to repair the local damaged point of face image with weak texture was proposed. At first,the total variation method was used to divide face image with weak texture into the skeleton part and the texture part. Then,the Bayesian compressive sensing was used to get distribution functions of sparse coefficients in the skeleton part and the texture part. Moreover,the variance and mean value of distribution function in the skeleton part and the texture part were calculated. According to the variance and mean value in the skeleton part and the texture part of face image with the weak texture,the face image after restoration was obtained. Thus,the restoration of local damaged point of face image with weak texture was completed.Simulation results show that the proposed method has high accuracy in decomposing face images with weak texture.Meanwhile,it has good effect on repairing face images with weak texture.