A new face recognition method based on robust correlation analysis and sparse representation of complex matrices
Yutao Cao, Aidi Wu · Proceedings of the 2020 4th International Conference on Electronic Information Technology and Computer Engineering · 2020
In order to improve the effect of face recognition, a robust typical correlation analysis algorithm based on sparse representation and context constraints of complex matrix is proposed. First, the face images are decomposed by two different wavelets, then the high frequency subbands and low frequency subband of the wavelet decomposition are fused into complex matrices. Secondly, category information and context constraint related information are added to the framework of robust typical correlation analysis. Finally, sparse representation is used to classify and test respectively in ORL and AR databases. The experimental results show that the recognition effect is better than the traditional method.