Human Ear Recognition Based on Phase Congruency and Kernel Discriminant Analysis

Song Ze Lei; Qiang Zhu

doi:10.4028/www.scientific.net/AMM.241-244.1614

Paper Titles

A New Training Algorithm for RBF Neural Network based on Dynamic Fuzzy Clustering
p.1593

Weighted Association Rules Mining Algorithm Research
p.1598

The Research on Recognition of the Irregular Objects
p.1602

Finite-Time H_∞ Consensus of Multi-Agent Systems with a Leader
p.1608

Human Ear Recognition Based on Phase Congruency and Kernel Discriminant Analysis
p.1614

Research on Parameters Optimization Algorithm in Support Vector Machine Based on Immune Memory Clone Strategy
p.1618

Study on Path Based Hierarchy Aggregation in DS-TE Environment
p.1622

A Multi-Classified Method of Support Vector Machine (SVM) Based on Entropy
p.1629

Research on Interactive Visualization Clustering Method Based on the Radar Chart
p.1633

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vols. 241-244Human Ear Recognition Based on Phase Congruency...

Human Ear Recognition Based on Phase Congruency and Kernel Discriminant Analysis

Abstract:

To solve the multi-pose ear recognition problem under the different illumination condition, a novel method which combines phase congruency with kernel discriminant analysis (KDA) is proposed. The phase congruency of ear image is first calculated using Log-Gabor filter with 5 scales and 8 orientations, and then the phase congruency of different orientation is constructed as high dimensional vector including ample information. The high dimensional vector is mapped to kernel space to acquire discriminant feature. Experimental results show that the proposed method obtains higher recognition rate compared with the other related methods. The method of the phase congruency can eliminate the influence of illumination and phase congruency with KDA is effective to multi-pose ear recognition.