The heterogeneous face recognition algorithm is based on super resolution reconstruction and two-dimensional marginal fisher analysis. In this paper, a super resolution reconstruction algorithm by adaptive multi-dictionary learning is adopted. Compared with the traditional global dictionary learning, this algorithm spends less time on dictionary training and image reconstruction to a great extent. Firstly, a sketch is transformed to a photo by eigenface algorithm. Secondly, super resolution reconstruction by improved adaptive multi-dictionary learning is used to reconstruct the synthesized photo, which is able to enhance the quality of synthesized photo effectively. Finally, the synthesized photo is recognized by two-dimensional marginal fisher analysis. We demonstrate these ideas in practice and show how they lead to faster operation speed and ideal recognition rate.
CITATION STYLE
Zhao, J., Mao, Y., Fang, Q., Liang, Z., Yang, F., & Zhan, S. (2015). Heterogeneous face recognition based on super resolution reconstruction by adaptive multi-dictionary learning. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9428, pp. 143–150). Springer Verlag. https://doi.org/10.1007/978-3-319-25417-3_18
Mendeley helps you to discover research relevant for your work.