Abstract:

In order to promote the recognition effect of the sparse representation classification method,a monogenicfeature-based sparse representation classification ( MSRC) method is proposed.As compared with the Gabor feature,the monogenic feature can be used to extract the image phase information which is not sensitive to light,so theMSRC method can improve the image robustness to light.As compared with the multi-scale and multi-direction ofthe Gabor feature,the monogenic feature can reduce the processing time spent on the feature.Experimental resultsshow the proposed method is applicable with higher recognition rate and speed.

Key words: face recognition, sparse representation, monogenic feature, Gabor feature