VEIN PATTERN RECOGNITION BASED BIOMETRIC SYSTEM AND METHODS THEREOF

• US 20110169934A1
• Filed: 09/22/2009
• Published: 07/14/2011
• Est. Priority Date: 09/22/2008
• Status: Active Grant

• Current Assignee: Arti Agrawal, Govind Agrawal, Kranthi Kiran Pulluru
• Original Assignee: Arti Agrawal, Govind Agrawal, Kranthi Kiran Pulluru
• Inventors: Agrawal, Arti, Agrawal, Govind, Pulluru, Kranthi Kiran
• Granted Patent: US 8,803,963 B2
• Time in Patent Office: Days
• US Class Current: 348/77
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G06V 40/12 Fingerprints or palmprints
• CPC Class Codes: G06V 40/14 Vascular patterns
• CPC Class Codes: G06V 40/63 by static guides
• CPC Class Codes: H04L 63/0861 using biometrical features, e.g. fingerprint, retina-scan cryptographic mechanisms or cryptographic arrangements for entity authentication using biological data H04L9/3231

Abstract

Individual authentication techniques are disclosed, which relate to a vein recognition based to biometric authentication system. Systems and methods to capture the subcutaneous vein pattern at the back of the palm and use it as the biometric identity of an individual are also disclosed.

Claims

1-11. -11. (canceled)

12. A vein pattern recognition based biometric system comprising

i. a hand placement unit, and detection unit being configured to inform computing device about presence of a hand for processing;

said hand placement unit comprises

a portion to rest flat portion of the palm,

plurality of buttons to be pressed by thumb being configured to indicate left or right hand has been placed,

concavity to rest fleshy part of the palm, and

portion to guide the wrist for alignment and protrusion being adopted to separate middle finger and ring finger;

ii. illuminating and imaging unit comprises a bright object placed in the field of view of the illuminating and imaging unit being configured to illuminate region of interest with uniformly diffused near-infrared light;

iii, feature extraction unit being adopted to extract appropriate information of the vascular pattern from image captured by the imaging device; and

iv. recognition unit being configured to compare plurality of patterns and to make authentication.

13. The system as claimed in claim 12, wherein the system is configured to capture subcutaneous vein pattern at the back of the palm.

14. The system as claimed in claim 12, wherein the illuminating and the imaging unit comprises a camera, and light source.

15. The system as claimed in claim 12, wherein the bright object is reflecting mirror or an LED light source.

16. The system as claimed in claim 12, wherein the feature extraction unit and recognition unit comprises computing devices selected from a group comprising embedded system, PDA, mobile phones, computers and other related devices.

17. A method to authenticate vein pattern, said method comprising acts of

i. pressing the detection button by placing hand in a hand placement unit to inform computing device about presence of a hand for processing;

ii. illuminating region of interest of the hand placed and capturing binarized image of the hand;

iii. extracting vascular pattern from the captured image for representing the pattern as line-segment-hypothesis and binarized image;

iv. computing core segments and merging them to form additional line segment hypothesis;

v. representing the line segment hypothesis in the form of a directed graph, where each node represents a line segment and there does not exist a path between two nodes if they have a common core segment; and

vi. comparing the extracted line-segment-hypothesis and binarized image with stored line-segment-hypothesis and binarized image to authenticate the vein pattern.

18. The method as claimed in claim 17, wherein the method of authentication comprises registration process comprising acts of

i. placing the hand in the hand placement unit and assigning unique identifier for each user;

ii. capturing the images of the hands for predetermined number of times upon pressing detection button and converting the captured image into bit pattern;

iii. computing prerequisite features from the captured image;

iv. matching the computed sample with previously obtained sample or samples of same hand; and

v. indexing the matched samples in the storage for authentication,

19. A method for extracting vascular pattern comprising acts of

i. extracting image of boundary of hand and performing filtration;

ii. applying local thresholding by comparing the image with filtered image and storing thresholded image as binarized vein pattern;

iii. thinning the binarized vein pattern into single pixel thickness;

iv. computing core line segment from the thinned image; and

v. computing merged segments from the core segments for extracting vascular pattern.

20. The method as claimed in claim 19, wherein extracting image of hand boundary includes

i. resizing the image from original size to predetermined size to increase the processing speed;

ii. performing median filtering on resized image to avoid speckle noise from camera;

iii. computing threshold of the filtered image to segment hand region from background; and

iv. binarizing the thresholded image to obtain hand shape binarized image.

21. A method for comparing live pattern with recorded pattern comprising acts of computing distance between plurality of line-segment-hypothesis, wherein the computing distance comprises steps of

i. constructing a matrix with live patterns segment and recorded pattern segments;

ii. initializing cumulative haussdorf distance to zero;

iii. filling the matrix representing line segment haussdorff between predetermined segment (h) of the live pattern and the predetermined segment (k) of the recorded pattern;

iv. selecting the smallest element from the filled matrix and incrementing the haussdorff distance with the selected element;

v. filling all the elements of row h and column k of the matrix with positive and negative infinity respectively; and

vi. checking if the predetermined segments (h and k) in the live pattern and recorded pattern respectively correspond to a merged segment, if so to fill all the rows corresponding to the core segments of h and columns corresponding to the core segments of k respectively with infinity.

22. The method as claimed in claim 21, wherein the method computes

i. difference between hand widths in the live pattern and recorded pattern, if the difference is greater than a prescribed threshold then returns zero (0);

ii. difference between number of vein points in the live pattern and recorded pattern, if the difference is greater than a prescribed threshold then returns zero (0); and

iii. match between the binarized image of the live pattern and the recorded pattern, where the match is carried out using logical operation between the live pattern and the recorded pattern to determine ratio of number of white pixels in the patterns.