Image comparison method

• US 5,267,324 A
• Filed: 10/22/1992
• Issued: 11/30/1993
• Est. Priority Date: 07/05/1990
• Status: Expired due to Term

• Current Assignee: YOZAN, Inc.
• Original Assignee: Ezel, Inc., Sharp Corporation (Hon Hai Precision Industry Co., Ltd.)
• Inventors: Kumagai, Ryohei
• Primary Examiner(s): Razavi, Michael T.
• Assistant Examiner(s): COUSO, YON JUNG
• Application Number:
  US07/965,054
• Time in Patent Office: 404 Days
• Field of Search: 382/4, 382/5, 382/21, 356/71
• US Class Current: 382/127
• CPC Class Codes: G06V 40/1347 Preprocessing; Feature extraction
• CPC Class Codes: G07C 9/257 electronically G07C9/26 takes precedence

Abstract

A method for comparing two images including ridges with each other, comprising steps from i) to iv).

i) Calculating density value of each pixel of an image to be examined and reference image; ii) detecting a direction with minimal change between each pixel and adjacent one, concerning to each pixel; iii) dividing the image to be examined and reference image, calculating an index according to the direction with minimal change as to each block; iv) using the index, it is judged if the image to be examined and reference image substantially coincide to each other.

Claims

1. A method for identifying a ridge sample object as similar to a ridge master object comprising steps of:

illuminating the master object;

capturing a master image depicting ridges of the master object, the master image being represented by master ridge pixels;

detecting a master pixel direction having a minimum change in density from every one of a group of master ridge pixels, said master ridge pixels group including a majority of the master ridge pixels, to all pixels adjacent to and near every one of the master ridge pixels in said master ridge pixels group, thereby detecting a distribution of the master pixel directions having a minimum change in density for the entire master image based on the majority of master ridge pixels in said master ridge pixels group;

storing master pixel directions in a memory;

illuminating a sample object;

capturing a sample image depicting ridges of the sample object, the sample image being represented by sample ridge pixels;

detecting a sample pixel direction having a minimum change in density from every one of a group of sample ridge pixels, said sample ridge pixels group including a majority of the sample ridge pixels, to all pixels adjacent to and near every one of the sample ridge pixels in said sample ridge pixels group, thereby detecting a distribution of sample pixel directions having a minimum change in density for the entire sample image based on the majority of sample ridge pixels in said sample ridge pixels group;

detecting a measure of correspondence relating master pixel directions for a majority of master ridge pixels with sample pixel directions for a majority of sample ridge pixels; and

determining whether the measure of correspondence satisfies a predetermined condition, thereby identifying the sample object as similar to the master object.

2. A method as in claim 1 wherein the master object is a finger.

3. A method as in claim 1 wherein the step of detecting a master pixel direction includes a step calculating a differential value between a master ridge pixel and a master local pixel.

4. A method as in claim 1 wherein the step of detecting a sample pixel direction includes a step calculating a differential value between a sample ridge pixel and a sample local pixel.

5. A method as in claim 1 further comprising steps of:

detecting a master block direction for each of a plurality of multi-pixel master blocks in the master image;

detecting a master block variance V_M for each master block characterizing the variance of master pixel directions within the master block;

detecting a sample block direction for each of a plurality of multi-pixel sample blocks in the sample image, each sample block corresponding to a master block; and

detecting a sample block variance V_S for each sample block characterizing the variance of sample pixel directions within the sample block.

6. A method as in claim 5 wherein the step of determining whether the measure of correspondence satisfies a predetermined condition includes a step of determining whether the measure:

space="preserve" listing-type="equation">RE=cos.sup.2 (Θ

/2)-(Vm+Vs)/2

is greater than a predetermined value for each of a plurality of corresponding sample blocks and master blocks where;

Θ

is an angle between the sample block direction and the master block direction;

V_M is the master block variance; and

V_S is the sample block variance.

7. A method as in claim 5 wherein the step of detecting a master block direction includes steps of:

i) detecting a master subblock direction for each of a plurality of multi-pixel master subblocks in the master block; and

ii) determining an average value of master subblock directions as the master block direction.

8. A method as in claim 7 wherein the step of detecting a master subblock direction includes steps of:

for each of a number m or master pixel directions d measuring a count N(d) of master ridge pixels within a master subblock having the same master pixel direction;

calculating a weighted master value V(D) for each of a series of test directions D according to the formula;

##EQU7## where;

D are test directions;

n is a normalized constant

d are master pixel directions

N(d) are the number of pixels having direction d

m is a count of master pixel directions; and

selecting, as the detected master subblock direction, the minimum of values V(D).

9. A method as in claim 5 wherein the step of detecting a sample block direction includes steps of:

i) detecting a sample subblock direction for each of a plurality of multi-pixel sample subblocks in a sample block; and

ii) determining an average value of sample subblock directions as the sample block direction.

10. A method as in claim 9 wherein the step of detecting a sample subblock direction includes steps of:

for each of a number m of sample pixel directions d measuring a count N(d) of sample ridge pixels within a sample subblock having the same sample pixel direction;

calculating a weighted sample value V'"'"'(D) for each of a series of test directions D according to the formula;

##EQU8## where;

D are test directions;

n is a normalizing constant

d are sample pixel directions

N(d) are the numbers of pixels having direction d

m is a count of sample pixel directions;

selecting, as the detected sample subblock direction, the minimum of values V(D).

11. An apparatus for regulating access to a restricted region comprising:

a light source;

an optical path directing light from the light source to a measuring location at which an object having ridges can be positioned;

means for transforming an optical image to an electrical signal;

an optical path directing light from the measuring location to said transforming means;

image processing apparatus, responsive to said transforming means;

blocking means for preventing access to the restricted region; and

releasing means for disabling the blocking means when receiving a signal from the image processing apparatus;

wherein the image processing apparatus comprises;

i) means for capturing a master image depicting ridges of a master object, the master image being represented by master ridge pixels, and for capturing a sample image depicting ridges of a sample object, the sample image being represented by sample ridge pixels;

ii) means for detecting a plurality of master pixel directions, one for each of a majority of the master ridge pixels, each master pixel direction indicating a direction having a minimum change in density from the master ridge pixel to all pixels adjacent to and near the master ridge pixel;

iii) means for storing the master pixel directions;

iv) means for detecting a plurality of sample pixel directions, one for each of a majority of the sample ridge pixels, each sample pixel direction indicating a direction having a minimum change in density from the sample ridge pixel to all pixels adjacent to and near the sample ridge pixel;

v) means for detecting a measure of correspondence relating master pixel directions for a majority of master ridge pixels with sample pixel directions for a majority of sample ridge pixels; and

vi) means for signaling the releasing means when the measure of correspondence satisfies a predetermined condition.