Conversion of an image represented by a field of pixels in a gray scale to a field of pixels in binary scale

• US 4,685,145 A
• Filed: 12/07/1984
• Issued: 08/04/1987
• Est. Priority Date: 12/07/1984
• Status: Expired due to Fees

• Current Assignee: Secure Technologies Group, Inc.
• Original Assignee: Secure Technologies Group, Inc.
• Inventors: Schiller, Michael
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Couso, Jose L.
• Application Number:
  US06/679,622
• Time in Patent Office: 970 Days
• Field of Search: 382/5, 382/50, 382/52, 382/54, 358/282, 358/284
• US Class Current: 382/272
• CPC Class Codes: G06V 10/28 Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
• CPC Class Codes: G06V 40/1347 Preprocessing; Feature extraction
• CPC Class Codes: G07C 9/37 using biometric data, e.g. fingerprints, iris scans or voice recognition

Abstract

A gray scale image of a fingerprint composed of a field of pixels is converted to a binary image composed of a field of pixels by a technique which takes into account the directivity of the ridge and valley structure. Three intermediate binary images are developed, one by the use of a vertical filter, one by the use of a horizontal filter and a reference image by the use of a filter which is not directionally biased. Corresponding subfields around each pixel in each of the three images are compared. If the subfield for the vertically derived image is closer to that of the reference image then is the subfield for the horizontally derived image, then the binary value for the pixel from the vertically derived image is used in the final image; and vice versa. In this fashion, a fourth a final binary image is derived from a combination of the vertically derived image and horizontally derived image which includes the best imagery from each of those two intermediate images.

Claims

1. The method of providing a fingerprint image represented by a field of binary pixels comprising the steps of:

a. optically imaging a finger surface or a continuous fingerprint to provide an input gray scale image of said fingerprint, said input image being a field of gray scale pixels,

b. generating a first intermediate binary image using a first elongate mini-field by the method comprising the steps of;

for each pixel in said gray scale image, positioning said first elongate mini-field in a predetermined association with said pixel, comparing an average gray scale value of said mini-field with the gray scale value of said pixel, and assigning a binary value to a corresponding pixel of said first intermediate binary image, said binary value being a function of the relative values of said average gray scale value and said pixel gray scale value,

c. generating a second intermediate binary image using a second elongate mini-field, the elongate axes of said first and second elongate mini-fields being at substantial angles to one another, by the method comprising the steps of;

for each pixel in said gray scale image, positioning said second elongate mini-field in a predetermined association with said pixel, comparing an average gray scale value of said mini-field with the gray scale value of said pixel, and assigning a binary value to a corresponding pixel of said second intermediate binary image, said binary value being a function of the relative values of said average gray scale value and said pixel gray scale value,

d. generating a reference intermediate binary image, said reference intermediate binary image having directional acuity in any given locality that is intermediate between the directional acuity of said first and second binary images in said given locality,

e. associating each pixel of each of said three intermediate binary images with a predetermined subfield to establish first and second correlation values respectively, said first correlation value indicating correlation between the pixels in said subfield in said first binary image and the pixels in said subfield in said reference binary image, said second correlation value indicating correlation between the pixels in said subfield in said second binary image and said pixels in said subfield in said reference binary image, and

f. assigning a binary value to each pixel in a fourth binary image, said binary value in said fourth image being the binary value for the corresponding pixel from the one of said first or second intermediate binary images whose correlation value indicates the greater correlation.

2. The method of claim 1 wherein:

said first and second mini-fields are the outer ends of first and second elongate zones respectively, the center portions of said first and second elongate zones being omitted from said first and second mini-fields respectively.

3. The method of claim 2 wherein:

the length of each of said elongate zones is approximately the average ridge to ridge spacing of a fingerprint image represented by said gray scale image.

4. The method of claims 1, 2 or 3 wherein:

said elongate axes of said first and second elongate mini-fields are substantially perpendicular to one another.

5. The method of claims 1, 2 or 3 wherein:

said first and second correlation values are a count of the number of pixels in said subfield of said first and second binary images respectively that have the same binary value as the pixel in said reference binary image which corresponds to the pixels from said first and second binary images for which said subfield correlation values are established, thereby establishing first and second binary value counts associated respectively with corresponding pixels from said first and second intermediate binary images, and

said step of assigning a binary value to pixels in said fourth binary image employs the higher one of said binary value counts as indicating the greater correlation.

6. The method of claim 5 wherein:

said elongate axes of said first and second elongate mini-fields are substantially perpendicular to one another, one of said axes being parallel to a line of said field of pixels and the other of said axes being parallel to a column of said field of pixels.

7. The method of claim 4 wherein:

said pixel associated with each of said mini-fields is substantially centered in the corresponding one of said mini-fields, and

said pixel associated with said subfield is substantially centered in said subfield.

8. The method of claim 4 wherein:

a reference mini-field consists substantially of the zone defined by said first elongate mini-field and the zone defined by said second elongate mini-fields.

9. The method of claims 1, 2 or 3 wherein:

said first correlation value is a count of the number of lines of said subfield in said first binary image which have a larger number of individual pixel binary value correlations with corresponding pixels from said reference binary image than do corresponding lines from second binary image, and said second correlation value is a count of the number of lines of said subfield in the second binary image which have a larger number of individual pixel binary value correlation with corresponding pixels from said reference binary image than do corresponding lines from said first binary image, and

said step of assigning a binary value for pixels in said fourth binary image employs the higher correlation line count as the indicator of greater correlation.

10. A system for providing an image represented by a field of binary pixels comprising:

a. an optical imaging means to provide an input gray scale image of an object, said input image represented by a field of gray scale pixels,

b. means to generate and scan first, second and reference predetermined mini-fields respectively across said gray scale input image, said first and second mini-fields being substantially elongate and having their elongate axes at substantial angles to one another, said reference mini-field including areas of both of said first and second mini-fields,

c. means to position said first, second and reference mini-fields over each pixel in turn, to compare the gray scale value for a given pixel with an average gray scale value of the mini-field positioned over said given pixel and to assign a binary value to each pixel in first, second and reference intermediate binary images respectively based on that comparison,

d. means to associate each pixel of each of said three intermediate binary images with a predetermined subfield to establish first and second correlation values respectively, said first correlation value indicating correlation between the pixels in said subfield in said first binary image and the pixels in said subfield in said reference binary image, and second correlation value indicating correlation between the pixels in said subfield in said second binary imnage and said pixels in said subfield in said reference binary image, and

e. means to assign a binary value to each pixel in a fourth binary image, said binary value in said fourth image being the binary value for the corresponding pixel from the one of said first or second intermediate binary images whose correlation value indicates the greater correlation.

11. The system of claim 10 wherein:

said first and second mini-fields are the outer ends of first and second elongate zones respectively, the center portion of said first and second elongate zones being omitted from said first and second mini-fields respectively.

12. The system of claim 11 wherein:

the length of each of said elongate zones is approximately the average ridge to ridge spacing of a fingerprint image represented by said gray scale image.

13. The system of claims 10, 11 or 12 wherein:

said elongate axes of said first and second elongate mini-fields are substantially perpendicular to one another.

14. The system of claim 10, 11 or 12 wherein:

said first and second correlation values are a count of the number of pixels in said subfield of said first and second binary images respectively that have the same binary value as the pixel in said reference binary image which corresponds to the pixels from said first and second binary images for which said subfield correlation values are established, thereby establishing first and second binary value counts associated respectively with corresponding pixels from said first and second intermediate binary images, and

said step of assigning a binary value to pixels in said fourth binary image employs the higher one of said binary value counts as indicating the greater correlation.

15. The system of claim 14 wherein:

.

said elongate axes of said first and second elongate mini-fields are substantially perpendicular to one another, one of said axes being parallel to a line of said field of pixels and the other of said axes being parallel to a column of said field of pixels.

16. The system of claim 13 wherein:

said pixel associated with each of said mini-fields is substantially centered in the corresponding one of said mini-fields,

said pixel associated with said subfield is substantially centered in said subfield.

17. The system of claim 13 wherein:

a reference mini-field consists substantially of the zone defined by said first elongate mini-field and the zone defined by said second elongate mini-field.

18. The system of claims 10, 11 or 12 wherein:

said first correlation value is a count of the number of lines of said subfield in said first binary image which have a larger number of individual pixel binary value correlations with corresponding pixels from said reference binary image than do corresponding lines from said second binary image, and said second correlation value is a count of the number of lines of said subfield in said second binary image which have a larger number of individual pixel binary value correlations with corresponding pixels from said reference binary image than do corresponding lines from said first binary image, and

said step of assigning a binary value for pixels in said fourth binary image employs the higher correlation line count as the indicator of greater correlation.

19. The method of providing an image represented by a field of binary pixels comprising:

optically imaging a gray scale image comprising a field of pixels having values in a gray scale,

for a pixel of interest from said gray scale field, determining the local direction at said pixel of interest of the image represented by said field of pixels,

selecting an elongate mini-field having an elongate axis substantially orthogonal to said local direction, said mini-field being associated with said pixel of interest,

determining the average gray scale value of said elongate mini-field,

comparing said average gray scale value of said elongate mini-field with the gray scale value of said pixel of interest,

assigning to said pixel of interest, a binary value that is determined by the relative magnitudes of said average gray scale value of said mini-field and said gray scale value of said pixel of interest, and

repeating said steps of determining, selecting, determining comparing and assigning for each pixel of interest in said field.

20. A system for providing an image represented by a field of binary pixels comprising:

an optical imaging means to provide an input gray scale image of an object, said input image represented by a field of gray scale pixels,

means to generate and scan first, second and reference predetermined mini-fields respectively across said gray scale input image, said first and second mini-fields being substantially elongate and having their major axes substantially orthogonal to one another, said third mini-field including areas of both of said first and second mini-fields,

means to position said first, second and third mini-fields over each pixel in turn, to compare the gray scale value for a given pixel with an average gray scale value of the pixels within the mini-field positioned over said given pixel and to assign a binary value to each pixel in first, second and reference intermediate binary images respectively based on that comparison,

means to position each pixel of each of said three intermediate binary images within a predetermined subfield, said pixel being located substantially in the center of said subfield to establish first and second counts associated with said pixel of the number of pixels within the subfield from said first and second intermediate binary images respectively that have the same binary value as the corresponding pixel in said reference binary image, and

means to assign a binary value to each pixel in a fourth binary image, said binary value in said fourth image being the binary value for the corresponding pixel from the one of said first or second intermediate binary images whose count is greater.

21. A system for providing an image represented by a field of binary pixels comprising:

an optical imaging means to provide an input gray scale image of an object, said input image represented by a field of gray scale pixels,

means to generate and scan at least first and second predetermined mini-fields respectively across said input gray scale image, said first and second mini-fields being substantially elongate and having their elongate axes at substantial angles to one another,

means to position said first and second mini-fields in association with each pixel in turn, to compare the gray scale value for a given pixel with an average gray scale value of the pixels within the associated mini-field to assign a binary value to each pixel in first and second intermediate binary images respectively based on that comparison,

means to provide a reference intermediate binary image having directional acuity in any given locality that is intermediate between the directional acuity of said first and second binary images in that locality,

means to associate each pixel of each of said three intermediate binary images with a predetermined subfield to establish first and second correlation values respectively, said first correlation value indicating correlation between the pixels in said subfield in said first binary image and the pixels in said subfield in said reference binary image, said second correlation value indicating correlation between the pixels in said subfield in said second binary image and said pixels in said subfield in said reference binary image, and

means to assign a binary value to each pixel in a fourth binary image, said binary value in said fourth image being the binary value for the corresponding pixel from the one of said first or second intermediate binary images whose correlation value indicates the greater correlation.