Technical Name IDENTIFYING METHOD FOR FINGERPRINTS AND PALM PRINT
Project Operator National Pingtung University of Science and Technology
Project Host 陳念慈
Summary
The main object of the present invention is to provide a method for identifying dual features of fingerprints and palm print. It sets several characteristic regions in the acquiring palm print and fingerprints images. Use image recognition to compare with the established feature database of palm print and fingerprints. The invention has the functions of improving execution speed and recognition effect. 
An identifying method for fingerprints and palm print comprise steps of acquiring an image of fingerprints and palm print; treating the image by means of a gray scale transformation method according to a selected reference line for outputting a plurality of gray scale distribution diagrams corresponding to the image; setting a plurality of characteristic region in accordance with the gray scale distribution diagrams; and comparing the charactertisic regions with a fingerprints and palm print database by using an identifying technique.
Scientific Breakthrough
The present invention uses the finger and palm features as the basis of the identification system. In the finger feature part, each finger acquires 3 feature areas, so there are 9 feature areas in the three fingers; the palm feature part uses the main line close to the palm as the midpoint, Take 5 feature areas from each palm, and add up to 14 feature areas. Finally, the correlation coefficient method is used to compare the similarity between the feature area of the identified image and the feature area of the database image. If the feature area of the identified image has more than 12 feature areas, the comparison score is higher than the set comparison score. It is judged as the correct user; otherwise, it is judged that there is no such user.
Industrial Applicability
"Contactless scanning", as the name suggests, customers no longer need to take out credit cards or smart phones when checking out, just need to reach out the palm of the hand and put it under the infrared of the scanner to check out.
Keyword Biometrics identification technology fingerprint Palm print Image acquire Image conversion Wavelet transform Correlation coefficient method Image feature non-contact scanning system FRR (False Rejection Rate) FAR (False Acceptance Rate)
other people also saw