System and method for determining if a fingerprint image contains an image portion representing a smudged fingerprint impression

• US 5,883,971 A
• Filed: 10/23/1996
• Issued: 03/16/1999
• Est. Priority Date: 10/23/1996
• Status: Expired due to Term

• Current Assignee: Pendragon Networks LLC (Pendrell Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Bolle, Rudolf Maarten, Pankanti, Sharathchandra Umapatirao, Yao, Yi-Sheng
• Primary Examiner(s): Kelley, Christopher S
• Assistant Examiner(s): CHAWAN, SHEELA C
• Application Number:
  US08/735,718
• Time in Patent Office: 874 Days
• Field of Search: 382/115, 382/124, 382/125, 382/126, 382/116, 382/274, 382/266, 382/263, 382/272
• US Class Current: 382/124
• CPC Class Codes: G06T 2207/20021 Dividing image into blocks, subimages or windows
• CPC Class Codes: G06T 2207/30168 Image quality inspection
• CPC Class Codes: G06T 7/0002 Inspection of images, e.g. flaw detection
• CPC Class Codes: G06V 40/1347 Preprocessing; Feature extraction

Abstract

A fingerprint image is divided into blocks of pixels. The blocks are determined to be within the foreground or background of the image. Poor quality images are examined to determine if the imaging process is corrupted, e.g. by a wet finger, by determining if a smudginess measure is above a smudginess threshold. For each foreground block, a mean intensity (μ) of pixels is computed for those pixels whose intensities are smaller than the mean intensity of all pixels within the respective foreground block. Further the standard deviation (∛) of intensities of all pixels within the respective foreground block is computed. For a foreground block with good contrast, μ is small and ∛ is large. But for a block with low contrast due to smudginess, μ is small and ∛ is small. Subsequently, to measure the contrast (a contrast measure) within a block, a function of the product (C_S) of corresponding μ and corresponding ∛ is determined. A comparison is then made to determine whether a block is a smudged block. If the contrast measure is smaller than a smudginess threshold STH1, then the block is classified as a smudged block. If not, the block is classified as a non-smudged block. Embodiment, STH1 is set to 0.12. Lastly, the smudginess measure as the ratio of the number of smudged blocks to total number of foreground blocks is computed. If the resulting ratio is larger than a second threshold, STH2, a smudged impression is reported.

Claims

1. A system for determining whether fingerprint images contain smudged impressions comprising:

a computer comprising at least one central processing unit, at least one memory, and at least one input device;

at least one fingerprint image having a foreground and a background and being acquired by the input device and stored in the at least one memory, the at least one fingerprint image further comprising a plurality of blocks of pixels;

a smudginess process that operates on said computer for determining a mean intensity of pixels within each foreground block of pixels, for determining a second mean intensity of pixels for those pixels within the respective foreground block of pixels whose intensities are less than the mean intensity, and for determining a standard deviation of intensities of all pixels within the respective foreground block of pixels, wherein a contrast measure is a function of a product of the determined second mean intensity and the determined standard deviation; and

a comparator that determines that the foreground block of pixels is a smudged block if the contrast measure is less than a smudginess threshold, and that further determines that the image is a smudged image if a ratio of the number of smudged blocks to a total number of foreground blocks is greater than a second threshold.

2. A system, as in claim 1, wherein said at least one input device is comprised of at least one of a camera, an image scanner, and a second computer connected to a network in common with the system.

3. A system, as in claim 1, where the blocks of pixels comprise down-sampled subsets of pixels from an area on an original fingerprint image.

4. A system, as in claim 1, where the ratio of the number of smudged blocks to the total number of foreground blocks is less than the second threshold, but wherein the fingerprint image is identified as failing a quality criteria.

5. A system, as in claim 1, where the mean intensity is comprised of a mean ridge intensity.

6. A system, as in claim 1, where the contrast measure is normalized by a maximum value of the contrast measure.

7. A system, as in claim 6, where the smudginess threshold is between 0 and 0.2.

8. A system, as in claim 6, where the smudginess threshold is 0.12.

9. A system, as in claim 1, where the second threshold has a value in the range of less than 0.4 to approximately 0.4.

10. A method of detecting a smudged fingerprint image comprising the steps of:

selecting at least one foreground block of pixels in a fingerprint image;

determining a mean intensity of pixels for each selected foreground block of pixels;

determining a second mean intensity of all the pixels of the selected foreground block that have an intensity below the determined mean intensity;

determining a standard deviation of the intensities of pixels of the selected foreground block;

marking the foreground block as a smudged block if a contrast measure is below a smudginess threshold, the contrast measure being a function of the product of the determined second mean intensity and the determined standard deviation; and

identifying the fingerprint image as a smudged fingerprint image if a ratio of the number of smudged blocks to a total number of foreground blocks is larger than a second threshold.

11. A method for processing an image of a fingerprint, comprising steps of:

obtaining an image of a fingerprint, the image being comprised of pixels;

processing the fingerprint image to partition the image into blocks of pixels each specified to be one of a background block or a foreground block;

further processing foreground blocks to obtain regions each comprised of set of contiguous foreground blocks wherein the pixels of a block exhibit a prominent directionality;

weighting blocks as a function of distance from a common reference point within the fingerprint image;

determining a quality measure for the fingerprint image by computing a ratio of a summation of total weights of blocks exhibiting a prominent directionality to a summation of the weights of the foreground blocks; and

directing further processing of the fingerprint image based on the determined quality measure by comparing the determined quality measure to a quality threshold;

wherein if the determined quality measure is less than the quality threshold, the step of directing further processing is comprised of a step of performing a smudginess analysis of the fingerprint image;

wherein the smudginess analysis comprises steps of,

determining a mean intensity of pixels for individual ones of foreground blocks of pixels;

determining a second mean intensity of pixels for the individual ones of the foreground blocks that have an intensity below the determined mean intensity;

determining a standard deviation of the intensities of pixels for the individual ones of the foreground blocks;

marking a given foreground block as being a smudged block if a contrast measure is below a smudginess threshold, the contrast measure being a function of the product of the determined second mean intensity and the determined standard deviation; and

identifying the fingerprint image as a smudged fingerprint image if a ratio of the number of smudged blocks to a total number of foreground blocks is larger than a second threshold.

12. A method as in claim 11, wherein the common reference point is a centroid of the regions of foreground blocks.

13. A method as in claim 11, where the contrast measure is normalized by a maximum value of the contrast measure.

14. A method as in claim 11, where the smudginess threshold is in a range from approximately 0 to approximately 0.2.

15. A method as in claim 11, where the smudginess threshold is approximately 0.12.

16. A method as in claim 11, where the second threshold is approximately 0.4.