A Novel Rank-Level Fusion for Multispectral Palmprint Identification System

An efficient palmprint recognition method based on block dominat orientation code

“…To demonstrate the efficacy of the proposed method, it was compared with those methods representing the state-of-the art: Competitive Code based on Wavelet Fusion (CCWF) [24], Competitive Code based on Score Level Fusion (CCSLF) [26], Joint Palmprint and Palm-vein Method (JPPM) [23], and Rank-Level Fusion (RLF) [28]. Table III.…”

“…Tahmasebi et al [28] used a rank-level fusion scheme for multispectral palmprint identification. Guo et al [29] present a study of feature band selections by analysing hyperspectral palmprint data.…”

Feature extraction in palmprint recognition using spiral of moment skewness and kurtosis algorithm

“…The goal of this step is to capture the most distinctive information from the region of interest (ROI) that has been identified. There are several algorithms for this purpose: structural-based approaches such as the local line direction patterns, Sobel/Canny edge lines or minutiae determined by the scale-invariant feature transform (SIFT) or Speeded Up Robust Features (SURF) [4]; appearance-based approaches, e.g., sub-space based on the EigenPalm and principal component analysis (PCA) [5]; statistical approaches such as Gabor and wavelet [2,5]; codingbased methods (e.g., palm-code) [6]; and hybrid techniques such as 2D FLPP [7,8].…”

“…Based on a normalized 2D Gabor filter, Zhang et al devised an effective palmcode algorithm [6]. Zuo et al [9] and Kong et al [5] used competitive coding to devise the multiscale-orientated 2D log-Gabor filters and encode the dominant orientation, respectively. Jia et al [3] used the line orientation code founded on a modified finite Radon transform, which is similar to the competitive code.…”

“…Jia et al [3] used the line orientation code founded on a modified finite Radon transform, which is similar to the competitive code. Zhang et al [5] described a multispectral approach, where the features are extracted by competitive coding using six oriented 2D Gabor filters for each spectral band; a score fusion at the recognition stage completes the process. Gabor filters were also used by Tahmasebi et al [5] to create a RankLevel fusion for a multispectral palmprint system.…”