Hierarchical palmprint feature extraction and recognition based on multi‐wavelets and complex network

• DOI: 10.1049/iet-ipr.2017.0520
• Publication Date: 2018-01-22T15:24:01Z
• Authors: Lijian Zhou; Chen Zhang; Zuowei Wang; Ying Wang; Zhe‐Ming Lu
• Keywords: Feature (linguistics)

This study presents a hierarchical palmprint feature extraction and recognition approach based on multi‐wavelet complex network (CN) since they can effectively decrease redundant information enhance key points of main lines wrinkles. The is first pre‐filtered decomposed once using multi‐wavelet. Three components (LL 1,2,3 ) corresponding to the pre‐filter except for diagonal component are extracted as elementary features. Second, binary images (BLL obtained by average window method different thresholds. Third, three series dynamic evolution CN models (the 1st, 2nd, 3rd CNs) constructed from global local, which mosaiced BLL , 1 four equally divided sub‐images respectively. Fourth, statistical features networks, in degree standard deviation degrees 1st CNs 2nd CNs. Fifth, fisher linear discriminate analysis method. Finally, nearest neighbourhood classifier used recognise palmprint. Based CASIA Palmprint Image Database, experimental results show that proposed with good robustness overcome problem small training samples number.