Research On Fractal-Based Repair Method Of Finger-Vein Network

The finger-vein recognition has been receiving increasing attention for its active liveness and convenience in recent years.However,due to the high scattering of near-infrared light in biological tissue,the finger-vein images are degradated greatly,which results in breaking the network of veins.This makes finger-vein recognition performance unsatisfactory.To solve the problem of vascular network incompleteness in finger-vein IR images,a method based on fractal is proposed to repair the finger-vein network.1)A method is proposed for extracting and optimizing the fractal feature of finger veins.The finger-vein structure is analyzed based on the fractal theory.Then the ratio of branch length to the length of the parent vessel,as a kind of fractal feature,is extracted to characterize the finger-vein network.Furthermore,in order to optimize the result of fractal feature extraction,the 8-orientation images obtained by Gabor filtering are used to pre-repair the vascular structure.2)A fractal-based method is developed to repair the finger-vein network.Based on the statistic similarity of the fractal structure,the finger vein is simulated by calculating the movement probability of neighbor vessel points.And the fractal feature extracted above is used to calculate the length of the lost blood vessel.Then the finger vein is repaired and the integrity of the network is improved.3)An improved method based on HMM is proposed to repair the network.In order to simulate the finger-vein morphology effectively,a movement prediction model of vessel points based on HMM is established.Then,with the fractal feature obtained above,the finger-vein network is repaired more effectively.The experimental results show that the proposed method is effective to repair the locally incomplete region of finger-vein network.The proposed method also achieves higher accuracy in finger-vein recognition.