基于子模式纹理分析的鲁棒人脸识别研究

Research of Robust Face Recognition Based on Sub-pattern Texture Analysis

面部识别(FR)系统可以自动识别或校验从数码相机或图像生成设备中获得的人脸图像,要从所获图像中提取面部特征,并与人脸数据库中的数据进行比对。目前,几乎所有的FR都面临与面部视角相关的障碍,包括光照不足和低分辨率,这些问题使其识别率大为降低。为了解决这个问题,提出了一种基于光照变化的人脸识别框架,该框架利用离散余弦变换的图像全变差最小化(DTV)及伽柏过滤器;并融合子模式分析(SMP)及区分性累计特征变换(DAFT),有效地解决了光照条件变化大的人脸识别问题。在AR及YaleB人脸数据库上的实验表明,与其它最先进的方法相比较,在处理光照条件变化很大的人脸识别问题时明显优于其它方法。

Face recognition （FR） system is automatically identifying or verifying a personal face acquired from a digital camera or a image generation device. In order to do this, facial features from the acquired image should be extracted and compared with a facial database. All FRs face an obstacle related to the viewing angle of the face including poor lighting and low resolution. Because of those problems, its recognition rate substantially decreases. In this paper, a new face recognition framework under an illumination change conditions is proposed, which contains the total variation minimization framework based on the discrete cosine transform image （DTV） and Gabor filters, and integration of sub-mode analysis （SMP） and distinguish the cumulativefeature Transform （DAFT）. Experiments results on AR and Yale B show that proposed method has more effective recognition accuracy on face recognition with large illustration condition changing comparing with other latest approaches.