KR20040026905A - Evaluation apparatus and method of image quality for realtime iris recognition, and storage media having program thereof - Google Patents

Abstract

PURPOSE: A device and a method for recognizing an iris in real-time, and a recording medium storing a program are provided to recognize a user fast and correctly by discriminating only an image fit to iris recognition in real-time. CONSTITUTION: A sampler(210) samples a middle value of the image in a preset pixel area unit by receiving the image capturing an eye from the outside. An absolute evaluation tool(220) detects the iris from a sampling image output from the sampler and removes the image that the detected iris is under a preset absolute reference value. A relative evaluation tool(230) recovers the iris of the sampling image output from the absolute evaluation tool to an original size, matches the iris with the image capturing the eye, and relatively decides a quality order of the image by using a position of an eyelid on the matched image.

Description

FIELD AND METHOD OF IMAGE QUALITY FOR REALTIME IRIS RECOGNITION, AND STORAGE MEDIA HAVING PROGRAM THEREOF

The present invention screens only the image suitable for iris recognition using the pupil in real time, and then provides an image quality evaluation device and method for real-time iris recognition that provides a priority-determined image so as to identify a high quality image first, and a program thereof. It relates to a recording medium storing the.

Traditionally, a personal password or personal identification number is widely used as a method of identifying an individual. However, such conventional methods do not provide a stable and accurate personal identification method due to the risk of theft and loss.

In addition, its dysfunction can cause many side effects in society. As an alternative to supplement the disadvantages of the conventional method, biometrics are in the spotlight.

Biometrics is a method of identifying individuals based on their physical (physical) and behavioral characteristics. Fingerprints, faces, irises, palm prints, etc. may be classified as physical characteristics, and signatures, voices, etc. may be classified as behavioral characteristics.

Among these biometric methods, the iris is known to be the most excellent in terms of uniqueness, invariability, and stability. In addition, the iris is applied to a field that requires a high level of security because the false recognition rate is very low. The iris is formed before birth after age three, and is known to change forever unless there is a special trauma.

The conventional iris recognition method does not use filtering to determine the suitability of an image in real time with respect to an image input from a camera. Because of this, there is a problem that the inappropriate image lowers the recognition rate of the entire system.

In addition, there is a problem in that the user has a lot of inconvenience by forcing the user to select the correct position and operation. In addition, there is a problem that the input of a low quality image interferes with the automation of the iris recognition system.

Even if the filtering process is added, it takes a long time to calculate, and there is a problem that can be used only in a constrained environment.

The present invention has been made to solve the above problems, the present invention screens only the image suitable for iris recognition in real time, the image quality evaluation apparatus and method for real-time iris recognition that can identify individuals quickly and accurately and its program The purpose is to provide a stored recording medium.

It is another object of the present invention to provide an apparatus and method for evaluating image quality for real-time iris recognition, which provides priority to identify an image having high quality, and a recording medium storing the program.

1 is a block diagram showing an embodiment of an image quality evaluation apparatus for real-time iris recognition according to the present invention,

2 is a flowchart illustrating an embodiment of an image quality evaluation method for real time iris recognition according to the present invention;

3 illustrates brightness characteristics of each region of an eye image,

4 is an embodiment of an image inappropriate for iris recognition,

5 illustrates an embodiment of binarizing a sampling image.

6 is an embodiment in which the binarized sampling image is projected horizontally,

7 is an embodiment of matching the image of the eye by restoring a search area including a pupil to its original size,

8 shows a horizontal edge projected horizontally,

9 illustrates an example in which a quality priority of an image is determined.

<Description of the symbols for the main parts of the drawings>

100: photographing means 200: image quality evaluation apparatus

210: sampling means 220: absolute evaluation means

222: pupil detection unit 224: absolute evaluation unit

226: removal unit 230: relative evaluation means

232: search area unit 234: restoration unit

236: eyelid detection unit 238: relative evaluation unit

300: Iris recognition system

In order to achieve the above object, according to an embodiment of the present invention, sampling means for sampling an intermediate value of the image by a predetermined pixel region by receiving an image of the eye from the outside; Absolute evaluation means for detecting a pupil from the sampled image output from the sampling means, and removing the image whose detected pupil is equal to or less than a predetermined absolute reference value; And a relative evaluation means for restoring the pupil of the sampling image outputted from the absolute evaluation means to the original size to match the image of the eye, and relatively determining the quality priority of the image using the eyelid position of the matched image. It provides an image quality evaluation apparatus for real-time iris recognition comprising a.

Further, in the present invention, the absolute evaluation means determines the threshold of the pupil using a histogram and binarizes the sampling image based on the determined threshold of the pupil, and projects the binarized sampling image horizontally and vertically to determine the size and position of the pupil. Pupil detection unit to obtain; An absolute evaluation unit for absolute evaluation of the quality of an image having the size and position of the pupil obtained by the pupil detection unit; And an elimination unit for removing an image whose sum of absolute values evaluated by the absolute evaluation unit is less than or equal to the absolute reference value.

In addition, the present invention, the absolute evaluation unit

Presence evaluation module for evaluating the size of the pupil using the presence evaluation function (S _p ),

Where R _mean = average of pupil size,

x _radius = pupil's x-axis radius, y _radius = pupil's y-axis radius;

Position evaluation module for evaluating the position of the pupil using a position evaluation function (S _l ),

Where I _width = _{width of the} image, I _height = length of the image,

x _c : x coordinate of image center, y _c : y coordinate of image center,

x: x coordinate of detected pupil center, y: y coordinate of detected pupil center; And

Integrity evaluation module for evaluating the aspect ratio of the pupil using the integrity evaluation function (S _s ),

(when x _radius ≥y _radius ),

(when x _{radius <} y _radius ),

It solves the above-mentioned problem as an image quality evaluation apparatus for real-time iris recognition comprising a.

Further, in the present invention, the relative evaluation means comprises: a search area unit for setting a search area including a pupil from the sampled image output from the absolute evaluation means; A restoration unit for restoring the search area to its original size and matching the pupil of the restored area to an image of the eye; An eyelid detector configured to detect a horizontal edge from the matched image by the restoration unit and to project a horizontal edge, and then detect a position of the eyelid using the horizontal edge projected horizontally; And a relative evaluation unit for determining the priority of the quality of the image according to the position of the eyelid and the distance of the pupil, and solves the above-mentioned problems as an image quality evaluation apparatus for real time iris recognition.

In addition, in the present invention, the search area solves the above-mentioned problem as an image quality evaluation device for real-time iris recognition, characterized in that the rectangular region circumscribed to the pupil as a search area.

In addition, in the present invention, the eyelid detection unit solves the above-mentioned problems as an image quality evaluation apparatus for real-time iris recognition, characterized in that for detecting the horizontal edge using the Sobel-Y operator.

In addition, in the present invention, the eyelid detection unit solves the above-mentioned problems as an image quality evaluation apparatus for real-time iris recognition, characterized in that the position where the horizontal component appears the most, as the position of the eyelid.

In another embodiment of the present invention to achieve the above object an input step for receiving an image of the eye taken from the outside; Sampling the intermediate value of the image in units of a predetermined pixel region; An absolute evaluation step of detecting a pupil from the sampled image and removing an image of which the detected pupil is equal to or less than a predetermined absolute reference value; And a relative evaluation step of restoring the pupil of the sampled image to its original size to match the image of the eye and relatively determining the quality priority of the image using the position of the eyelid of the matched image. It provides an image quality evaluation method.

In addition, in the present invention, the absolute evaluation step (a) determining the threshold of the pupil using the histogram; (b) binarizing the sampled image based on the determined threshold of the pupil; (c) projecting the binarized sampling image horizontally and vertically to obtain the size and position of the pupil; (d) absolute evaluation of the quality of the image having the size and position of the pupil; And (e) removing the image whose sum of the absolute evaluation values is less than or equal to the absolute reference value. The above-described problem is solved as an image quality evaluation method for real-time iris recognition.

In the present invention, the step (d) is

(d1) evaluating the size of the pupil using the presence evaluation function (S _p ),

Where R _mean = average of pupil size,

x _radius = pupil's x-axis radius, y _radius = pupil's y-axis radius;

(d2) evaluating the position of the pupil using a position evaluation function S _l ,

Where I _width = _{width of the} image, I _height = length of the image,

x _c : x coordinate of image center, y _c : y coordinate of image center,

x: x coordinate of detected pupil center, y: y coordinate of detected pupil center; And

(d3) evaluating the aspect ratio of the pupil using the integrity evaluation function (S _s ),

(when x _radius ≥y _radius ),

(when x _{radius <} y _radius ),

It solves the above-mentioned problem as an image quality evaluation method for real-time iris recognition comprising a.

Also, in the present invention, the relative evaluation step may include: (f) setting a search region including a pupil from the sampled image; (g) restoring the search region to its original size and matching the pupil of the restored region to the image of the eye; (h) detecting a horizontal edge from the matched image; (i) projecting the horizontal edge horizontally; (j) detecting the position of the eyelid using a horizontal edge projected horizontally; And (k) determining the quality priority of the image according to the position of the eyelid and the distance of the pupil. The above-described problem is solved as an image quality evaluation method for real time iris recognition.

In addition, in the present invention, step (f) solves the above-described problems as an image quality evaluation method for real-time iris recognition, characterized in that the quadrangle with the pupil is set as a search region.

In addition, in the present invention, step (h) solves the above-mentioned problems as an image quality evaluation method for real-time iris recognition, characterized in that the horizontal edge is detected using the Sobel-Y operator.

In addition, in the present invention, step (j) solves the above-mentioned problems as the image quality evaluation method for real-time iris recognition, characterized in that the position where the horizontal component appears the most, as the position of the eyelid.

In order to achieve the above object, another embodiment of the present invention comprises an input process for receiving an image of the eye from the outside; A sampling process of sampling an intermediate value of the image in units of a predetermined pixel region; An absolute evaluation process of detecting a pupil from the sampled image and removing an image whose detected pupil is equal to or less than a predetermined absolute reference value; And a relative evaluation process of restoring the pupil of the sampling image to its original size to match the image of the eye, and relatively determining the quality priority of the image using the position of the eyelid of the matched image. It provides a recording medium storing the image quality evaluation program for the.

In addition, in the present invention, the absolute evaluation process includes: (a) a process of determining the threshold of the pupil using the histogram; (b) binarizing the sampled image based on the determined threshold of the pupil; (c) a process of projecting the binarized sampling image horizontally and vertically to obtain the size and position of the pupil; (d) a process of absolute evaluation of the quality of the image with the size and position of the pupil; And (e) a process for removing an image in which the sum of the absolute evaluation values is equal to or less than the absolute reference value.

In the present invention, the (d) process is

(d1) a process of evaluating the size of the pupil using the presence evaluation function (S _p ),

Where R _mean = average of pupil size,

x _radius = pupil's x-axis radius, y _radius = pupil's y-axis radius;

(d2) a process of evaluating the position of the pupil using the position evaluation function S _l ,

Where I _width = _{width of the} image, I _height = length of the image,

x _c : x coordinate of image center, y _c : y coordinate of image center,

x: x coordinate of detected pupil center, y: y coordinate of detected pupil center; And

(d3) a process for evaluating the aspect ratio of the pupil using the integrity evaluation function (S _s ),

(when x _radius ≥y _radius ),

(when x _{radius <} y _radius ),

The above-described problem is solved by a recording medium comprising a.

Also, in the present invention, the relative evaluation process includes the steps of: (f) setting a search region including a pupil from the sampled image; (g) restoring the search region to its original size and matching the pupil of the restored region to the image of the eye; (h) detecting a horizontal edge from the matched image; (i) a process of projecting the horizontal edge horizontally; (j) detecting the position of the eyelid using a horizontal edge projected horizontally; And (k) a process of determining the quality priority of the image according to the position of the eyelid and the distance of the pupil.

Further, in the present invention, the process (f) solves the above-described problem as the recording medium, characterized in that the quadrangle that circumscribes the pupil is set as the search area.

In addition, in the present invention, the (h) process solves the above-described problems as a recording medium, characterized by detecting the horizontal edge using the Sobel-Y operator.

Further, in the present invention, the (j) process solves the above-described problems as the recording medium, which detects the position where the horizontal component appears most as the position of the eyelid.

The objects, features and advantages of the present invention will be more readily understood by reference to the accompanying drawings and the following detailed description.

The present invention preferably implements a method for determining the priority so as to first check the image suitable for iris recognition using a pupil and screening it in real time, and then confirms the image having a high quality, and an image quality evaluation apparatus implemented to support the same. Suggest as an example.

Preferred embodiments of the invention include embodiments such as computer systems and computer program products programmed to carry out the methods of the invention. According to an embodiment of a computer system, a set of instructions for carrying out a method resides in one or more memories, which may be stored as a computer program product on a recording medium such as a CD-ROM or the like.

1 is a block diagram showing an embodiment of an image quality evaluation apparatus for real-time iris recognition according to the present invention. Referring to FIG. 1A, the image quality evaluation apparatus 200 includes a sampling unit 210, an absolute evaluation unit 220, and a relative evaluation unit 230.

The sampling unit 210 receives an image of the eye from the outside and samples an intermediate value of the image in units of a predetermined pixel region. In the embodiment of FIG. 1, the sampling means 210 is configured to receive an image of an eye photographed from the photographing means 100.

The absolute evaluation means 220 detects a pupil from the sampled image output from the sampling means 210, and removes an image whose detected pupil is equal to or less than a predetermined absolute reference value.

The absolute evaluation unit 220 includes a pupil detector 222, an absolute evaluation unit 224, and a removal unit 226.

The pupil detector 222 determines the threshold of the pupil using the histogram and binarizes the sampled image based on the determined threshold of the pupil. In this case, the histogram is preferably an intensity histogram. Then, the binarized sampling image is projected horizontally to obtain the size and position of the pupil.

The absolute evaluation unit 224 has an absolute evaluation of the quality of the image with the size and position of the pupil obtained by the pupil detection unit 222, the removal unit 226 is the sum of the absolute value evaluated in the absolute evaluation unit 224 is absolute The image below the reference value is removed.

The absolute evaluation unit 224 includes an existence evaluation module 224a, a position evaluation module 224b, and a completeness evaluation module 224c as shown in FIG. 1B.

Present evaluation module (224a) is present the evaluation function (S _p), the evaluation function exists and the size of the pupil using the (S _p) is defined as equation (1).

R _mean = _mean pupil size, x _radius = pupil's x-axis radius, y _radius = pupil's y-axis radius

In this case, the average R _mean of the pupil sizes is a value obtained by averaging the pupil sizes in advance from the recognizable images.

Position evaluation module (224b) is located in the evaluation function _(l S) to assess the position of the pupil, where the evaluation function (S _l) used is defined as in equation (2).

I _width = the _{width of the} image, I _height = the _{height of the} image,

x _c : x coordinate of image center, y _c : y coordinate of image center,

x: x coordinate of detected pupil center, y: y coordinate of detected pupil center

Integrity evaluation module (224c) is evaluating the aspect ratio of a pupil by using the integrity evaluation function (S _s), and the integrity evaluation function (S _s) is defined as equation (3).

,

(when x _{radius <} y _radius )

The relative evaluation means 230 restores the pupil of the sampled image output from the absolute evaluation means 220 to the original size to match the image of the eye, and uses the position of the eyelid of the matched image to prioritize the quality of the image. Determine relatively.

The relative evaluation means 230 includes a search area unit 232, a restoration unit 234, an eyelid detection unit 236, and a relative evaluation unit 238.

The search area 232 sets a search area including the pupil from the sampled image output from the absolute evaluation means 220. At this time, the search area 232 preferably sets a rectangle that circumscribes the pupil as the search area.

The restoration unit 234 restores the search region to its original size, and matches the pupil of the restored region to the image of the eye. Here, the image of the eye is received from the photographing means (100).

The eyelid detector 236 detects the horizontal edge from the matched image by the restoration unit 234 and projects the horizontal edge, and then detects the position of the eyelid using the horizontal edge projected horizontally. The eyelid detector 236 preferably detects the horizontal edge using the Sobel-Y operator, and detects the position where the horizontal component appears most as the position of the eyelid.

The relative evaluation unit 238 determines the priority of the image quality according to the position of the eyelid and the distance from the pupil.

The original image that has passed the quality priority determined by the relative evaluation means 230 and the absolute evaluation process is output to the iris recognition system 300, and the iris recognition system 300 recognizes the iris from the high priority image.

2 is a flowchart illustrating an embodiment of an image quality evaluation method for real time iris recognition according to the present invention. Referring to FIG. 2, the present invention provides an input step (S410) of receiving an image of an eye from an external source, a sampling step (S420) of sampling an intermediate value of a predetermined pixel area unit, and a pupil from a sampling image. Absolute evaluation step (S430) of detecting and removing the image of the detected pupil is less than the predetermined absolute reference value (S430), restoring the pupil of the sampling image to its original size to match the image of the eye and using the eyelid position of the matched image Relative evaluation step (S440) to relatively determine the priority of the image quality.

Absolute evaluation step (S430) is a step (a) of determining the threshold of the pupil using the histogram (S431), (b) step (S432) of binarizing the sampled image based on the determined threshold of the pupil, and the binarized sampling (C) step (S433 to S434) of obtaining the size and position of the pupil by projecting the image horizontally and vertically; (d) (S435) of absolute evaluation of the quality of the image with the size and position of the pupil; And (e) step S436, which removes an image whose sum of values is equal to or less than an absolute reference value.

Although not shown in the drawing, step (d) uses step (d1) of evaluating the size of the pupil using the existence evaluation function (S _p ) of Equation 1 and position estimation function (S ₁ ) of Equation (2). (D2) evaluating the position of the pupil, and (d3) evaluating the aspect ratio of the pupil using the integrity evaluation function S _s of Equation 3.

Relative evaluation step (S440) is a step (f) of setting the search area including the pupil from the sampled image (S441), restoring the search area to the original size and matching the pupil of the reconstructed area to the image of the eye (g) step S442, step (h) S443 of detecting a horizontal edge from the matched image, step (i) (S444) of horizontally projecting the horizontal edge, and horizontal edge projected horizontally (J) step S445 of detecting the position of the eyelid using the step, and (k) step 446 of determining the priority of the image quality according to the position of the eyelid and the distance between the pupil.

(f) Step S441 sets a rectangle that circumscribes the pupil as the search area, and (h) Step S443 detects the horizontal edge using the Sobel-Y operator. (j) Step S445 detects the position where the horizontal component appears most as the position of the eyelid.

In the present invention configured as described above, after the image quality evaluation apparatus 200 evaluates only the image suitable for iris recognition in real time and determines in real time, the priority of the image quality is first determined so as to provide the iris recognition system. The process will be described in detail with reference to FIGS. 1 to 8 as follows.

The sampling unit 210 receives an image of the eye from the photographing unit 100 (S410), and samples an intermediate value of the image in units of a predetermined pixel region (S420).

For example, if the image is divided into 4 pixel areas in the horizontal and vertical directions, and the median is taken in each area, the overall size is reduced to 1/16. The reason for doing this is to shorten the processing time.

Then, the absolute evaluation means 220 detects the pupil from the sampled image output from the sampling means 210, and removes the image of the detected pupil is less than or equal to the predetermined absolute reference value (S430). Detailed description of the S430 process is as follows.

3 illustrates brightness characteristics of each region of an eye image. In the recognizable histogram of the eye image, the brightness of the pupil area, the iris area and the sclera area is divided into three parts. Among them, in particular, the brightness of the pupil region occupies the lowest portion and has a characteristic of being separated from other portions.

However, among these images, an image that is not suitable for iris recognition exists as shown in FIG. 4. Comparing FIG. 3 with FIG. 4, the peak of the pupil region does not appear well in the histogram. 4A shows a case of blinking eyes, and FIG. 4B shows a case in which miscellaneous images have been entered. 4C illustrates a case in which the eyelashes or eyelids invade the pupil, and FIG. 4D illustrates a case in which the center of the pupil largely deviates from the center of the image due to the movement of the user.

Such images should be excluded before the iris recognition system to improve the efficiency and recognition rate of the iris recognition process.

In order to use this feature, the pupil detector 222 determines the threshold of the pupil using the histogram (S431). The minimum value is determined between the two peaks P1 and P2 shown in FIG. 3 to determine a threshold. This is because the image showing the peak of the pupil area can be completely separated by the threshold determined in step S431, but the image that is not is difficult to separate.

When the sampling image is binarized in black and white based on the determined threshold of the pupil (S432), it is as shown in FIG. 5 illustrates an embodiment of binarizing a sampling image.

Next, the pupil detector 222 obtains the size and position of the pupil by projecting the binarized sampling image horizontally (S433) (S434). 6 illustrates an embodiment in which the binarized sampling image is horizontally projected.

The absolute evaluation unit 224 has an absolute evaluation of the quality of the image with the size and position of the pupil obtained by the pupil detection unit 222 (S435), and the process is as follows.

The presence evaluation module 224a evaluates the size of the pupil using the existence evaluation function S _p of Equation 1. Then, the position evaluation module 224b evaluates the position of the pupil using the position evaluation function S ₁ of Equation 2. In addition, the integrity evaluation module 224c evaluates the aspect ratio of the pupil using the integrity evaluation function S _s of Equation 3.

The values evaluated in the modules 224a, 224b, and 224c are summed in the removal unit 226, and then the image whose sum is less than or equal to the absolute reference value is removed (S436). Here, the absolute reference value means the fitness of utilization of the image, and the intensity of the fitness of fitness may be changed by adjusting the absolute reference value.

In other words, the image for registration should adopt a good quality, so the absolute reference value is applied higher to improve the accuracy, and the absolute reference value is applied lower in the actual recognition stage. In this way, if the absolute reference value is applied low in the actual recognition step, the image of poor quality can be used as the image for iris recognition. In addition, since the user does not need to photograph the iris several times due to the rejection of the image adoption, the user's convenience can be achieved.

The relative evaluation means 230 restores the pupil of the sampled image passing through the absolute evaluation means 220 to its original size to match the image of the eye, and uses the position of the eyelid of the matched image to prioritize the quality of the image. Determine relatively. This process is explained in more detail as follows.

The search area 232 sets a search area including the pupil from the sampled image output from the absolute evaluation means 220 (S441), and the recovery unit 234 restores the search area to its original size. The restoration unit 234 matches the pupil of the restored region with the image of the eye (S442). Here, the image of the eye is received from the photographing means (100).

FIG. 7 illustrates an embodiment in which the search region including the pupil is restored to its original size to match the image of the eye. This process is for extracting accurate location information, and the search area is preferably a quadrangle around the pupil.

For example, if the starting and ending points in the x-direction, and the starting and ending points in the y-direction of the pupil extracted from the sampling image are four times, respectively, the positions of the pupils are matched to positions in the original image. The reason for 4 times is because the median value is taken in units of 4 pixels in the sampling process.

Next, the eyelid detector 236 detects a horizontal edge from the matched image in step S442 (S443), and then horizontally projects the horizontal edge (S444). Although there are various methods of detecting the horizontal edge, it is preferable to use the Sobel-Y operator in the present invention.

The eyelid detector 236 detects the position of the eyelid using the horizontal edge projected horizontally (S445). 8 shows a horizontal edge projected horizontally, the most horizontal component appears on the horizontal line where the eyelid is located. Therefore, the eyelid detector 236 preferably detects the position where the horizontal component appears most as the position of the eyelid.

The relative evaluation unit 238 determines the quality priority of the image according to the position of the eyelid and the distance to the pupil 446, and examples of determining the quality priority of the image are illustrated in FIG. 9. As the distance between the eyelid and the pupil increases, the area of the iris that can be recognized increases, so that the image having a long distance is preferentially selected.

In addition, the quality priority determined by the relative evaluation unit 238 is output to the iris recognition system 300 together with the image passing the absolute evaluation process.

The present invention is not limited to the embodiments described above, and various modifications and changes can be made by those skilled in the art, which are included in the spirit and scope of the present invention as defined in the appended claims.

As described above, the present invention, by selecting only the image suitable for iris recognition in real time to provide to the iris recognition system, the iris recognition system helps to recognize the individual quickly and accurately. And the iris recognition system can be fully automated.

In addition, the histogram can be used to detect the pupil under the constraint of the external environment, and the efficiency of preprocessing can be improved by using the detected pupil information.

In addition, the iris recognition system provides the iris recognition system with a priority determined using the position of the eyelids so that the image having high quality can be identified first.

In addition, there is no need for an administrator's instruction to adopt a high quality image in a real security system. By adopting a recognizable image of a series of consecutively taken images for a given time, and automatically removing images that are not, it becomes easy for a user to acquire an image by himself. And, there is an advantage that can maximize the convenience of the user.

Claims (21)

Sampling means for receiving an image of the eye from the outside and sampling a median value of the image in a predetermined pixel area unit; Absolute evaluation means for detecting a pupil from the sampled image output from the sampling means, and removing the image whose detected pupil is equal to or less than a predetermined absolute reference value; And Relative evaluation means for restoring the pupil of the sampling image output from the absolute evaluation means to the original size to match the image of the eye, and relatively determine the quality priority of the image using the eyelid position of the matched image; Image quality evaluation device for real-time iris recognition, including. The method of claim 1, wherein the absolute evaluation means A pupil detector determining a threshold of the pupil using a histogram and binarizing the sampled image based on the determined threshold of the pupil, and projecting the binarized sampling image horizontally to obtain a pupil size and position; An absolute evaluation unit for absolute evaluation of the quality of an image having the size and position of the pupil obtained by the pupil detection unit; And And an elimination unit for removing an image whose sum of absolute values evaluated by the absolute evaluation unit is less than or equal to the absolute reference value. The method of claim 2, wherein the absolute evaluation unit Presence evaluation module for evaluating the size of the pupil using the presence evaluation function (S _p ), Where R _mean = average of pupil size, x _radius = pupil's x-axis radius, y _radius = pupil's y-axis radius; Position evaluation module for evaluating the position of the pupil using a position evaluation function (S _l ), Where I _width = _{width of the} image, I _height = length of the image, x _c : x coordinate of image center, y _c : y coordinate of image center, x: x coordinate of detected pupil center, y: y coordinate of detected pupil center; And Integrity evaluation module for evaluating the aspect ratio of the pupil using the integrity evaluation function (S _s ), (when x _radius ≥y _radius ), (when x _{radius <} y _radius ), Image quality evaluation apparatus for real-time iris recognition, comprising a. The method according to any one of claims 1 to 3, wherein the relative evaluation means A search area unit for setting a search area including a pupil from the sampled image output from the absolute evaluation means; A restoration unit for restoring the search area to its original size and matching the pupil of the restored area to an image of the eye; An eyelid detector configured to detect a horizontal edge from the matched image by the restoration unit and to project a horizontal edge, and then detect a position of the eyelid using the horizontal edge projected horizontally; And And a relative evaluation unit to determine the priority of the quality of the image according to the position of the eyelid and the distance between the pupil. The method of claim 4, wherein the search region unit Image quality evaluation apparatus for real-time iris recognition, characterized in that for setting the quadrangle to the pupil and the circumference as a search area. The method of claim 4, wherein the eyelid detection unit An image quality evaluation apparatus for real time iris recognition, characterized by detecting the horizontal edge using the Sobel-Y operator. The method of claim 6, wherein the eyelid detection unit An image quality evaluation apparatus for real-time iris recognition, characterized in that the position where the most horizontal component appears as the position of the eyelid. An input step of receiving an image of the eye from the outside; Sampling the intermediate value of the image in units of a predetermined pixel region; An absolute evaluation step of detecting a pupil from the sampled image and removing an image of which the detected pupil is equal to or less than a predetermined absolute reference value; And Real-time iris recognition including a relative evaluation step of restoring the pupil of the sampling image to the original size to match the image of the eye, and relatively determining the quality priority of the image using the position of the eyelid of the matched image. Image quality evaluation method. The method of claim 8, wherein the absolute evaluation step (a) determining a threshold of the pupil using the histogram; (b) binarizing the sampled image based on the determined threshold of the pupil; (c) projecting the binarized sampling image horizontally and vertically to obtain a pupil size and position; (d) absolute evaluation of the quality of the image having the size and position of the pupil; And and (e) removing the image whose sum of the absolute evaluation values is less than or equal to the absolute reference value. The method of claim 9, wherein step (d) (d1) evaluating the size of the pupil using the presence evaluation function (S _p ), Where R _mean = average of pupil size, x _radius = pupil's x-axis radius, y _radius = pupil's y-axis radius; (d2) evaluating the position of the pupil using a position evaluation function S _l , Where I _width = _{width of the} image, I _height = length of the image, x _c : x coordinate of image center, y _c : y coordinate of image center, x: x coordinate of detected pupil center, y: y coordinate of detected pupil center; And (d3) evaluating the aspect ratio of the pupil using the integrity evaluation function (S _s ), (when x _radius ≥y _radius ), (when x _{radius <} y _radius ), Image quality evaluation method for real-time iris recognition, characterized in that it comprises a. The method according to any one of claims 8 to 10, wherein the relative evaluation step (f) setting a search region including a pupil from the sampled image; (g) restoring the search region to its original size and matching the pupil of the restored region to the image of the eye; (h) detecting a horizontal edge from the matched image; (i) projecting the horizontal edge horizontally; (j) detecting the position of the eyelid using a horizontal edge projected horizontally; And (k) determining the quality priority of the image according to the position of the eyelid and the distance of the pupil; image quality evaluation method for real-time iris recognition. The method of claim 11, wherein step (f) The image quality evaluation method for real-time iris recognition, characterized in that the quadrangle the pupil and the circumference is set as a search region. The method of claim 11, wherein step (h) An image quality evaluation method for real-time iris recognition, characterized by detecting a horizontal edge using Sobel-Y operator. The method of claim 13, wherein step (j) An image quality evaluation method for real-time iris recognition, characterized by detecting the position where the horizontal component appears most frequently as the position of the eyelid. An input process for receiving an image of the eye from the outside; A sampling process of sampling an intermediate value of the image in units of a predetermined pixel region; An absolute evaluation process of detecting a pupil from the sampled image and removing an image whose detected pupil is equal to or less than a predetermined absolute reference value; And Real-time iris recognition including a relative evaluation process of restoring the pupil of the sampling image to its original size to match the image of the eye, and relatively determining the quality priority of the image using the position of the eyelid of the matched image. Recording medium storing image quality evaluation program. The method of claim 15, wherein the absolute evaluation process (a) a process of determining the threshold of the pupil using the histogram; (b) binarizing the sampled image based on the determined threshold of the pupil; (c) a process of projecting the binarized sampling image horizontally and vertically to obtain the size and position of the pupil; (d) a process of absolute evaluation of the quality of the image with the size and position of the pupil; And (e) a process of removing an image whose sum of the absolute evaluation values is equal to or less than the absolute reference value. 17. The process of claim 16, wherein the process (d) (d1) a process of evaluating the size of the pupil using the presence evaluation function (S _p ), Where R _mean = average of pupil size, x _radius = pupil's x-axis radius, y _radius = pupil's y-axis radius; (d2) a process of evaluating the position of the pupil using the position evaluation function S _l , Where I _width = _{width of the} image, I _height = length of the image, x _c : x coordinate of image center, y _c : y coordinate of image center, x: x coordinate of detected pupil center, y: y coordinate of detected pupil center; And (d3) a process of evaluating the aspect ratio of the pupil using the integrity evaluation function (S _s ), (when x _radius ≥y _radius ), (when x _{radius <} y _radius ), Record medium comprising a. 18. The process of any of claims 15 to 17, wherein the relative evaluation process is (f) setting a search area including a pupil from the sampled image; (g) restoring the search region to its original size and matching the pupil of the restored region to the image of the eye; (h) detecting a horizontal edge from the matched image; (i) a process of projecting the horizontal edge horizontally; (j) detecting the position of the eyelid using a horizontal edge projected horizontally; And (k) a process for determining the priority of the quality of the image according to the position of the eyelid and the distance of the pupil. 19. The process of claim 18, wherein (f) And a rectangle that circumscribes the pupil as a search area. 19. The process of claim 18, wherein (h) A recording medium characterized by detecting the horizontal edge using the Sobel-Y operator. The process of claim 20 wherein the process (j) A recording medium characterized by detecting the position where the horizontal component appears most as the position of the eyelid.