摘要

We propose a new method for image construction of 2D-LPP. As far as matrix data, for 2D-LPP and LPP, they are often vectorized to find the k-nearest neighbors, between columns of information to be lost. Our method refines the process seeking the k-nearest neighbors. So the corresponding columns information is well preserved. Our method can well show the spatial structure information of the original image matrices. To evaluate the performance of our method, several experiments, conducted on the ORL face database, the Yale face database and the AR face database, show the effectiveness of the proposed method.

  • 出版日期2012

全文