摘要
压缩感知理论的出现,为人脸识别提供新的方法。基于稀疏表示的分类算法(SRC)就是压缩感知在人脸识别中的典型应用。针对SRC算法对遮挡人脸识别效果不佳的情况,提出基于图像分块的SRC算法,当分块方式选择恰当时,该算法可以显著地提高遮挡人脸的识别率。
Compressed sensing provides new methods for face recognition. Sparse Representation-based Classification algorithm is a typical applica- tion of compressed sensing in face recognition. In view of the poor effect of occluded face recognition using SRC algorithm, proposes SRC algorithm based on image segmentation. When the block is appropriate, this algorithm can significantly improve the recognition rate of the occluded face.
基金
西华大学研究生创新基金项目(No.ycjj201367)
关键词
压缩感知
人脸识别
SRC算法
Compressed Sensing
Face Recognition
SRC （Sparse Representation-based Classification） Algorithm
