Fake finger determination device

1. A fake finger determination device, comprising:

acquisition unit for acquiring line width information related to a line width of a ridge or a line width of a valley line of a fingerprint based on a fingerprint image; and

determination unit for determining whether or not the finger corresponding to the fingerprint image is a real finger or a fake finger based on a relation of the line width of the ridge and the line width of the valley line,

wherein the acquisition unit further comprises:

splitting unit for splitting the fingerprint image into small regions;

deciding unit for deciding, for each of the split small regions, a line direction of the ridge or the valley line in the region;

gradient derivation unit for obtaining a maximum value and a minimum value of a gradient of changes in a pixel value by scanning each of the small regions in a direction that is perpendicular to the decided line direction in the region; and

line width calculation unit for obtaining the line width information by setting, as a line width starting point, a point where the gradient of changes in the pixel value obtained by scanning each of the small regions in the direction perpendicular to the line direction takes the maximum value, setting a point where the gradient of changes takes the minimum value as a line width end point, and counting the number of pixels from the line width starting point to the line width end point.

2. The fake finger determination device according to claim 1 ,

wherein the acquisition unit obtains both pieces of information of the line width information related to the line width of the ridge and the line width information related to the line width of the valley line, and

the determination unit determines that the finger corresponding to the fingerprint image is a real finger upon determining that the line width of the ridge is greater than the line width of the valley line in the fingerprint image based on the both pieces of information.

3. A fake finger determination device, comprising:

acquisition unit for acquiring line width information related to a line width of a ridge or a line width of a valley line of a fingerprint based on a fingerprint image; and

determination unit for determining whether or not the finger corresponding to the fingerprint image is a real finger or a fake finger based on a relation of the line width of the ridge and the line width of the valley line,

wherein the acquisition unit further comprises:

binary image generation unit for binarizing the fingerprint image and outputting a binary image;

image processing unit for performing contraction processing or expansion processing on the binary image a preset number of times; and

counting unit for obtaining the line width information by counting the number of white pixels in the binary image remaining after the contraction processing has been performed the preset number of times or by counting the number of black pixels in the binary image remaining after the expansion processing has been performed the preset number of times.

4. The fake finger determination device according to claim 3 ,

wherein the acquisition unit obtains both pieces of information of the line width information related to the line width of the ridge and the line width information related to the line width of the valley line, and

the determination unit determines that the finger corresponding to the fingerprint image is a real finger upon determining that the line width of the ridge is greater than the line width of the valley line in the fingerprint image based on the both pieces of information.

5. A fake finger determination device, comprising:

acquisition unit for acquiring line width information related to a line width of a ridge or a line width of a valley line of a fingerprint based on a fingerprint image; and

determination unit for determining whether or not the finger corresponding to the fingerprint image is a real finger or a fake finger based on a relation of the line width of the ridge and the line width of the valley line,

wherein the acquisition unit further includes:

binary image generation unit for binarizing the fingerprint image and outputting a binary image;

image processing unit for repeatedly performing contraction processing on the binary image until the total number of white pixels after the contraction processing becomes a ridge threshold or less, or for repeatedly performing expansion processing on the binary image until the total number of black pixels after the expansion processing becomes a valley line threshold or less; and

counting unit for obtaining the line width information by counting the number of times the contraction processing has been performed or the number of time the expansion processing has been performed.

6. The fake finger determination device according to claim 5 ,

wherein the acquisition unit obtains both pieces of information of the line width information related to the line width of the ridge and the line width information related to the line width of the valley line, and

the determination unit determines that the finger corresponding to the fingerprint image is a real finger upon determining that the line width of the ridge is greater than the line width of the valley line in the fingerprint image based on the both pieces of information.

7. A fake finger determination device, comprising:

acquisition unit for acquiring line width information related to a line width of a ridge or a line width of a valley line of a fingerprint based on a fingerprint image; and

determination unit for determining whether or not the finger corresponding to the fingerprint image is a real finger or a fake finger based on a relation of the line width of the ridge and the line width of the valley line,

wherein the acquisition unit further comprises:

splitting unit for splitting the fingerprint image into small regions;

deciding unit for deciding, for each of the split small regions, a line direction of the ridge or the valley line in the region; and

line width calculation unit for obtaining the line width information by binarizing the fingerprint image for each of the small regions, scanning each of the small regions in a direction that is perpendicular to the decided line direction in the region and counting the consecutive number of black pixels or the consecutive number of white pixels.

8. The fake finger determination device according to claims 7 ,

wherein the acquisition unit obtains both pieces of information of the line width information related to the line width of the ridge and the line width information related to the line width of the valley line, and

the determination unit determines that the finger corresponding to the fingerprint image is a real finger upon determining that the line width of the ridge is greater than the line width of the valley line in the fingerprint image based on the both pieces of information.

9. A fake finger determination method, comprising the steps of:

acquiring line width information related to a line width of a ridge or a line width of a valley line of a fingerprint based on a fingerprint image; and

determining whether or not the finger corresponding to the fingerprint image is a real finger or a fake finger based on a relation of the line width of the ridge and the line width of the valley line,

wherein the acquisition step further comprises the steps of:

splitting the fingerprint image into small regions;

deciding, for each of the split small regions, a line direction of the ridge or the valley line in the region;

obtaining a maximum value and a minimum value of a gradient of changes in a pixel value by scanning each of the small regions in a direction that is perpendicular to the decided line direction in the region; and

obtaining the line width information by setting, as a line width starting point, a point where the gradient of changes in the pixel value obtained by scanning each of the small regions in the direction perpendicular to the line direction takes the maximum value, setting a point where the gradient of changes takes the minimum value as a line width end point, and counting the number of pixels from the line width starting point to the line width end point.

10. A fake finger determination method, comprising the steps of:

acquiring line width information related to a line width of a ridge or a line width of a valley line of a fingerprint based on a fingerprint image; and

determining whether or not the finger corresponding to the fingerprint image is a real finger or a fake finger based on a relation of the line width of the ridge and the line width of the valley line,

wherein the acquisition step further comprises the steps of:

binarizing the fingerprint image and outputting a binary image;

performing contraction processing or expansion processing on the binary image a preset number of times; and

obtaining the line width information by counting the number of white pixels in the binary image remaining after the contraction processing has been performed the preset number of times or by counting the number of black pixels in the binary image remaining after the expansion processing has been performed the preset number of times.

11. A fake finger determination method, comprising the steps of:

acquiring line width information related to a line width of a ridge or a line width of a valley line of a fingerprint based on a fingerprint image; and

determining whether or not the finger corresponding to the fingerprint image is a real finger or a fake finger based on a relation of the line width of the ridge and the line width of the valley line,

wherein the acquisition step further includes the steps of:

binarizing the fingerprint image and outputting a binary image;

repeatedly performing contraction processing on the binary image until the total number of white pixels after the contraction processing becomes a ridge threshold or less, or repeatedly performing expansion processing on the binary image until the total number of black pixels after the expansion processing becomes a valley line threshold or less; and

obtaining the line width information by counting the number of times the contraction processing has been performed or the number of time the expansion processing has been performed.

12. A fake finger determination method, comprising the steps of:

acquiring line width information related to a line width of a ridge or a line width of a valley line of a fingerprint based on a fingerprint image; and

determining whether or not the finger corresponding to the fingerprint image is a real finger or a fake finger based on a relation of the line width of the ridge and the line width of the valley line,

wherein the acquisition step further comprises the steps of:

splitting the fingerprint image into small regions;

deciding, for each of the split small regions, a line direction of the ridge or the valley line in the region; and

obtaining the line width information by binarizing the fingerprint image for each of the small regions, scanning each of the small regions in a direction that is perpendicular to the decided line direction in the region and counting the consecutive number of black pixels or the consecutive number of white pixels.