논문
New Finger Biometrics Method using Near Infrared Imaging
- 저자이의철
- 학술지SENSORS 11/3
- 등재유형
- 게재일자(2011)
In this research, we propose a new finger biometrics method. Infrared finger image is captured then feature extraction
is performed based on the modified Gaussian high pass filter and with a binarization, local binary pattern (LBP) or
local derivative pattern (LDP). Infrared finger image includes the multimodal features of finger-vein and finger geometry
in addition to the parts of fingerprint. Instead of extracting each feature by using different methods, the modified
Gaussian high pass filter is fully convolved. Therefore, the extracted binary pattern of finger image includes the
multimodal features of vein, fingerprint, and finger geometry. Experimental results showed that the equal error rate of
the proposed method was 0.13%.
is performed based on the modified Gaussian high pass filter and with a binarization, local binary pattern (LBP) or
local derivative pattern (LDP). Infrared finger image includes the multimodal features of finger-vein and finger geometry
in addition to the parts of fingerprint. Instead of extracting each feature by using different methods, the modified
Gaussian high pass filter is fully convolved. Therefore, the extracted binary pattern of finger image includes the
multimodal features of vein, fingerprint, and finger geometry. Experimental results showed that the equal error rate of
the proposed method was 0.13%.
In this research, we propose a new finger biometrics method. Infrared finger image is captured then feature extraction
is performed based on the modified Gaussian high pass filter and with a binarization, local binary pattern (LBP) or
local derivative pattern (LDP). Infrared finger image includes the multimodal features of finger-vein and finger geometry
in addition to the parts of fingerprint. Instead of extracting each feature by using different methods, the modified
Gaussian high pass filter is fully convolved. Therefore, the extracted binary pattern of finger image includes the
multimodal features of vein, fingerprint, and finger geometry. Experimental results showed that the equal error rate of
the proposed method was 0.13%.
is performed based on the modified Gaussian high pass filter and with a binarization, local binary pattern (LBP) or
local derivative pattern (LDP). Infrared finger image includes the multimodal features of finger-vein and finger geometry
in addition to the parts of fingerprint. Instead of extracting each feature by using different methods, the modified
Gaussian high pass filter is fully convolved. Therefore, the extracted binary pattern of finger image includes the
multimodal features of vein, fingerprint, and finger geometry. Experimental results showed that the equal error rate of
the proposed method was 0.13%.
