Discriminative histograms of local dominant orientation (D-HLDO) for biometric image feature extraction

作者：

Highlights：

• A novel biometric image feature extraction method D-HLDO is proposed.

• D-HLDO means discriminative histogram of local dominant orientation.

• LM-NNDA is used to obtain the low-dimensional and discriminative feature.

• Experimental results demonstrate the effectiveness of the proposed D-HLDO method.

摘要

•A novel biometric image feature extraction method D-HLDO is proposed.•D-HLDO means discriminative histogram of local dominant orientation.•LM-NNDA is used to obtain the low-dimensional and discriminative feature.•Experimental results demonstrate the effectiveness of the proposed D-HLDO method.

论文关键词：Feature extraction,Principal component analysis (PCA),Biometrics,Image representation,Linear discriminant analysis (LDA)

论文评审过程：Received 17 July 2012, Revised 2 March 2013, Accepted 5 March 2013, Available online 14 March 2013.

论文官网地址：https://doi.org/10.1016/j.patcog.2013.03.005