New finger biometric method using near infrared imaging.

Area of Science:

• Biometrics
• Image Processing
• Pattern Recognition

Background:

• Biometric systems rely on unique physiological or behavioral characteristics.
• Finger-based biometrics offer a balance of uniqueness and user acceptance.
• Existing methods often require separate feature extraction for different biometric modalities.

Purpose of the Study:

• To propose a novel, unified finger biometric method.
• To enhance feature extraction efficiency by integrating multimodal information.
• To achieve high accuracy in finger-based identification.

Main Methods:

• Capture infrared finger images.
• Apply a modified Gaussian high-pass filter for feature extraction.
• Utilize binarization, Local Binary Pattern (LBP), and Local Derivative Pattern (LDP) for pattern analysis.

Main Results:

• The proposed method successfully extracts multimodal features (veins and geometry) using a single filter.
• The integrated feature extraction demonstrates effectiveness.
• An experimental error rate of 0.13% was achieved.

Conclusions:

• The novel infrared finger biometric method offers a unified approach to feature extraction.
• The method demonstrates high accuracy and efficiency.
• This technique holds promise for secure and reliable biometric identification.