System and method for personal identification based on multimodal biometric information

1. A system for recognizing a person, comprising:

a housing including a surface for a person to place a finger thereon;

a first light source disposed beneath the surface at one end of the housing, the first light source being configured to emit a first radiation to a first portion of the finger;

a prism disposed between the first light source and the surface, wherein the first radiation is transmitted through the prism and reflected by the finger;

a magnifying lens configured for magnifying the first radiation reflected by the finger;

a mirror configured for reflecting the first radiation magnified by the magnifying lens;

a minifying lens configured for minifying the first radiation reflected by the mirror;

a second light source including a pair of opposing and parallel light plates and disposed above the surface of the housing, wherein the finger is placed between the light plates;

a first sensor configured for acquiring a first image of a fingerprint of the person based on sensing the first radiation minified by the minifying lens and transforming the sensed first radiation into electrical signals to acquire the first image;

a second sensor configured for acquiring a second image of a finger vein of the person based on the second light source; and

an identification/verification module, implemented on at least one processor, configured for recognizing the person based on a first set of candidates identified based on the first image and a second set of candidates identified based on the second image.

2. The system of claim 1 , wherein the identification/verification module comprises:

a fingerprint image processing unit, implemented on at least one processor, configured for generating a fingerprint template based on the first image;

a finger vein image processing unit, implemented on at least one processor, configured for generating a characteristic finger vein image based on the second image; and

an identification/verification unit, implemented on at least one processor, configured for recognizing the person based on the fingerprint template and the characteristic finger vein image.

3. The system of claim 2 , wherein the identification/verification unit comprises:

a fingerprint based matching unit, implemented on at least one processor, configured for comparing the fingerprint template with a plurality of stored fingerprint templates to identify the first set of candidates; and

a finger vein based matching unit, implemented on at least one processor, configured for comparing the characteristic finger vein image with each of a plurality of stored characteristic finger vein images to identify the second set of candidates.

4. The system of claim 1 , wherein:

the first set of candidates includes one or more candidates each of which is associated with a confidence score;

the second set of candidates includes one or more candidates each of which is associated with a confidence score; and

the person is recognized based on the confidence scores.

5. The system of claim 4 , wherein:

the person is recognized as a common candidate in the first and second sets of candidates; and

if there are multiple common candidates in the first and second sets of candidates, the person is recognized based on a combination of the confidence scores for each of the common candidates.

6. The system of claim 5 , wherein the combination of the confidence scores includes at least one of: a sum of, a weighted average of, a maximum of, and a minimum of the confidence scores.

7. The system of claim 3 , wherein the finger vein image processing unit comprises:

a finger vein segmentation unit configured for segmenting a region in the finger vein image where the finger vein is present to generate a segmented image;

a finger vein image normalizer configured for normalizing the segmented image to generate a normalized image;

a finger vein extractor configured for identifying blood vessels based on the normalized image to generate a binary image;

an image enhancement unit configured for enhancing the binary image to generate an enhanced image; and

an image compression unit configured for compressing the enhanced image to generate the characteristic finger vein image.

8. The system of claim 7 , wherein the finger vein image normalizer is configured for normalizing at least one of shape, size, direction, and gray level of the segmented image.

9. The system of claim 7 , wherein the finger vein extractor is configured for determining, for each pixel in the normalized image, whether the pixel represents a portion of a blood vessel based on a predetermined threshold.

10. The system of claim 3 , wherein comparing the characteristic finger vein image with each stored characteristic finger vein image comprises:

for each offset value of a plurality of offset values,

aligning the characteristic finger vein image with the candidate characteristic finger vein image based on a next offset value;

computing a matching score associated with the offset value based on a match between the characteristic finger vein image and the candidate characteristic finger vein image aligned using the offset value;

selecting a matching score from a plurality of matching scores corresponding to matches between the characteristic finger vein image and the candidate characteristic finger vein image using the plurality of offset values; and

designating the selected matching score as a confidence score of the candidate characteristic finger vein image.

11. A method, implemented on a machine having at least one processor and storage for recognizing a person, comprising:

illuminating, upon a finger of the person being placed on a surface of a housing, the finger by a first light source disposed beneath the surface at one end of the housing and a second light source disposed above the surface of the housing, wherein the finger is placed between a pair of opposing and parallel light plates included in the second light source, and wherein, the first light source is configured to emit a first radiation to a first portion of the finger;

transmitting, via a prism disposed between the first light source and the surface, the first radiation to the finger;

magnifying the first radiation reflected by the finger;

reflecting the first radiation;

minifying the first radiation;

acquiring, by a first sensor, a first image of a fingerprint of the person based on transforming the minified first radiation into electrical signals to acquire the first image;

acquiring, by a second sensor, a second image of a finger vein of the person based on the second light source, wherein the second sensor is spaced apart from the first sensor;

determining a first set of candidates based on the first image;

determining a second set of candidates based on the second image; and

recognizing the person based on the first and/or second set of candidates.

12. The method of claim 11 , wherein determining the first set of candidates comprises:

generating a fingerprint template based on the first image; and

comparing the fingerprint template with a plurality of stored fingerprint templates to determine the first set of candidates.

13. The method of claim 11 , wherein determining the second set of candidates comprises:

generating a characteristic finger vein image based on the second image; and

comparing the characteristic finger vein image with each of a plurality of stored characteristic finger vein images to determine the second set of candidates.

14. The method of claim 11 , wherein:

the first set of candidates includes one or more candidates each of which is associated with a confidence score;

the second set of candidates includes one or more candidates each of which is associated with a confidence score; and

the person is recognized based on the confidence scores.

15. The method of claim 14 , wherein:

the person is recognized as a common candidate in the first and second sets of candidates; and

if there are multiple common candidates in the first and second sets of candidates, the person is recognized based on a combination of the confidence scores for each of the common candidates.

16. The method of claim 15 , wherein the combination of the confidence scores includes at least one of: a sum of, a weighted average of, a maximum of, and a minimum of the confidence scores.

17. The method of claim 13 , wherein generating a characteristic finger vein image comprises:

segmenting a region in the finger vein image where the finger vein is present to generate a segmented image;

normalizing the segmented image to generate a normalized image;

identifying blood vessels based on the normalized image to generate a binary image;

enhancing the binary image to generate an enhanced image; and

compressing the enhanced image to generate the characteristic finger vein image.

18. The method of claim 17 , wherein normalizing the segmented image comprises normalizing at least one of shape, size, direction, and gray level of the segmented image.

19. The method of claim 17 , wherein identifying blood vessels comprises determining, for each pixel in the normalized image, whether the pixel represents a portion of a blood vessel based on a predetermined threshold.

20. The method of claim 13 , wherein the step of comparing the characteristic finger vein image with each stored characteristic finger vein characteristic image comprises:

for each offset value of a plurality of offset values,

aligning the characteristic finger vein image with the candidate characteristic finger vein image based on the next offset value;

computing a matching score associated with the offset value based on a match between the characteristic finger vein image and the candidate characteristic finger vein image aligned using the offset value;

selecting a matching score from a plurality of matching scores corresponding to matches between the characteristic finger vein image and the candidate characteristic finger vein image using the plurality of offset values; and

designating the selected matching score as a confidence score of the candidate characteristic finger vein image.

21. A method, implemented on a machine having at least one processor and storage for identifying a person, comprising:

illuminating, upon a finger of the person being placed on a surface of a housing, the finger by a first light source disposed beneath the surface at one end of the housing and a second light source disposed above the surface of the housing, wherein the finger is placed between a pair of opposing and parallel light plates included in the second light source, and wherein, the first light source is configured to emit a first radiation to a first portion of the finger;

transmitting, via a prism disposed between the first light source and the surface, the first radiation to the finger;

magnifying the first radiation reflected by the finger;

reflecting the first radiation;

minifying the first radiation;

receiving a fingerprint image acquired from the person via a first sensor based on the first light source, wherein the first sensor transforms the minified first radiation into electrical signals to acquire the fingerprint image;

receiving a finger vein image acquired from the person via a second sensor based on the second light source, wherein the second sensor is spaced apart from the first sensor;

generating a characteristic finger vein image based on the finger vein image;

comparing the characteristic finger vein image with each of a plurality of candidate characteristic finger vein images stored in a database so that each of the candidate characteristic finger vein images is associated with a confidence score;

selecting one or more matching candidates based on the confidence scores associated therewith; and

identifying the person based on the confidence score of each of the one or more matching candidates and the fingerprint image.

22. The method of claim 21 , wherein the step of generating comprises:

segmenting a region in the finger vein image where the finger vein is present to generate a segmented image;

normalizing the segmented image to generate a normalized image; and

identifying blood vessels of the finger vein based on the normalized image to generate a binary image.

23. The method of claim 22 , wherein the step of normalizing is directed to at least one of shape, size, direction, and gray level of the segmented image.

24. The method of claim 22 , wherein the step of identifying includes determining, for each pixel in the normalized image, whether the pixel represents a portion of a blood vessel based on a predetermined threshold.

25. The method of claim 22 , further comprising:

enhancing the binary image to generate an enhanced image; and

compressing the enhanced image to generate the characteristic finger vein image.

26. The method of claim 21 , wherein the step of comparing the characteristic finger vein image with each candidate characteristic finger vein image comprises:

for each offset value of a plurality of offset values,

aligning the characteristic finger vein image with the candidate characteristic finger vein image based on the next offset value;

computing a matching score associated with the offset value based on a match between the characteristic finger vein image and the candidate characteristic finger vein image aligned using the offset value;

selecting a matching score from a plurality of matching scores corresponding to matches between the characteristic finger vein image and the candidate characteristic finger vein image using the plurality of offset values; and

designating the selected matching score as a confidence score of the candidate characteristic finger vein image.

27. The method of claim 21 , wherein the step of selecting comprises:

retrieving a pre-determined criterion to be used for selecting the one or more matching candidates; and

determining the one or more matching candidates based on the confidence scores associated therewith and the pre-determined criterion.

28. The system of claim 1 , wherein the first sensor and the second sensor are disposed beneath the surface of a housing of the system.

29. The system of claim 1 , further comprising:

a first presence detecting sensor disposed beneath the surface at a first end of the housing, and configured for sensing presence of a first portion of the finger when the person places the finger on the surface; and

a second presence detecting sensor disposed beneath the surface at a second end of the housing, and configured for sensing presence of a second portion of the finger, the second end being opposite the first end of the housing.