基于识别的凸集投影人脸图像超分辨率重建  

Face Image Super-Resolution Reconstruction Based on Recognition and Projection onto Convex Sets

作　　者：黄华[1] 樊鑫[1] 齐春[1] 朱世华[1] 

机构地区：[1]西安交通大学电子与信息工程学院,西安710049

出　　处：《计算机研究与发展》2005年第10期1718-1725,共8页Journal of Computer Research and Development

基　　金：陕西省自然科学基金项目(2003F29)

摘　　要：人脸图像的超分辨率重建在公安、视频监控等领域有重要应用价值·基于识别的思想,对人脸灰度图像进行统计分析,得到有关人脸灰度整体特征的先验知识,将其描述为属性集合,从而利用凸集投影算法进行超分辨率图像重建·实验结果表明,重建质量较为理想,与通常的超分辨率凸集投影重建方法相比,抑制噪声的能力有显著提高,重建质量改善明显,收敛速度加快,且易于计算和实现·Face image super-resolution reconstruction （SRR） can be widely used in forensic analysis and video surveillance. With the recognition-based idea, the statistical characteristics of face images are investigated and incorporated into SRR. Based on the set theoretic formulation, a projection onto convex sets （POCS） algorithm is applied to find the solution to face image reconstruction. Compared with the traditional POCS based SRR methods, the proposed approach imposes some extra constraint sets to the solution. The experiment results on frontal face images show that the proposed approach gains a better performance both on noise suppression and reconstruction quality and has the advantage of simplicity in computation.

关 键 词：图像重建 人脸图像 超分辨率 凸集投影 

分 类 号：TN911.73[电子电信—通信与信息系统]