JP2003141512A - Face imaging system and image processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To confirm an automatically extracted face image with the most appropriate open state of the eyes and a plurality of face image candidates other than the face image on the same screen and to easily determine the open state of the eyes of a subject 20. SOLUTION: A system includes a camera 82 for taking successive images of the upper torso of the subject 20 including the face to provide a plurality of face image data D1 for each subject 20; an HDD 81 storing the face image data D1 provided by the camera 82; a CPU 84 reading the plurality of face image data D1 from the HDD 81 for each subject 20 and extracting a plurality of face image candidates including the face image with the most appropriate open state of the eyes from the face image data D1; and an operation monitor 87 displaying the plurality of face image candidates extracted by the CPU 84. The CPU 84 contracts the face image candidates except the face image with the most appropriate open state of the eyes, and displays the face image with the most appropriate open state of the eyes and the plurality of contracted face image candidates on the same screen and at the same time.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention
A facial image photographing system suitable for an automatic issuing system of ID cards such as a cash card, a membership card, an employee ID card, an employee ID card, a student ID card, an alien registration card, and various driver's licenses, and an image processing method thereof Is.

[0002]

2. Description of the Related Art In recent years, licenses such as car licenses, identification cards, passports, alien registration cards, library cards, cash cards, credit cards, employee ID cards,
ID cards such as employee ID cards, membership ID cards, medical cards, and student ID cards, as well as application forms for issuing such ID cards, are often provided with ID photographs converted into digital image information. This kind of ID photograph is obtained by photographing a face part of a subject using a face image photographing system and acquiring face image information.

Also, the face image of the person, name, address, date of birth, date of issue, expiration date, etc. are recorded on these ID cards. In general, an ID photograph is required not to deteriorate in image quality for a long period of 1 to 5 years, for example.
Furthermore, according to an ID card such as an employee ID card, a good face image with an open eye state without eye blinking is required.

[0004]

The face image capturing system according to the conventional example has the following problems. A method has been devised to automatically detect the blinking of the photographed person and display a face image that has been determined to be appropriate on the operation monitor, but at present, the face image with the best open state cannot be detected correctly in many cases. .

In the conventional method, one face image, which is taken in a non-continuous manner and which has the best opened state, is displayed on the operation monitor, and the operator confirms the face image. Therefore, when the photographed image is blinking or the facial expression is not suitable for the face image of the ID card, the photograph must be taken again.

In the card creating system, if the face image sent from the face photographing system has a problem such as blinking, it is no longer possible to create an ID card using the face image. Face image with blinking I
This is because the D card is not suitable as a means for identifying the person.

Incidentally, a method of sequentially displaying and confirming the face images obtained by continuously photographing the face part on the operation monitor can be considered, but in the method of displaying one by one on the operation monitor, the operator is displayed on the monitor in front. You have to remember the state of the face image and compare the two in your head, which makes the work efficiency very poor.

Therefore, the present invention solves the above-mentioned problems, and it is possible to confirm on the same screen a face image which is said to have the best eye-opening state and a plurality of face image candidates other than this face image. It is an object of the present invention to provide a face image capturing system and an image processing method capable of easily determining the eye open state of a subject.

[0009]

In order to solve the above problems, a face image photographing system according to the present invention continuously photographs the upper half of the body including the face of the photographed person, and obtains a plurality of face image information for each photographed person. A photographing means for acquiring the face image information, a storage means for storing the face image information acquired by the photographing means, a plurality of face image information for each photographed person from the storage means, and an eye-opened state from the face image information. The control device for extracting a plurality of face image candidates including the best face image of, and a display unit for displaying the plurality of face image candidates extracted by the control device, the control device having the best eye open state A face image candidate other than a normal face image is reduced, and the display means is controlled so that the face image with the best open state and the reduced face image candidates are simultaneously displayed on the same screen. To do.

According to the face image photographing system of the present invention, the photographing means continuously photographs the upper half of the body including the face of the photographed person to obtain a plurality of face image information for each photographed person.
The face image information acquired by this photographing means is stored in the storage means. Based on this assumption, the control device reads out a plurality of face image information from the storage means for each person to be photographed, and extracts a plurality of face image candidates including the face image with the best open state from the face image information. It

For example, the control device detects all the eye areas of the face image of the photographed person with respect to a plurality of face image information of the photographed person read out from the storage means, and then detects any one of them. The eyes of the face image are compared with the eyes of other face images to extract the larger one, and the eyes of the face image extracted here are compared with the eyes of other face images The face image with the best eye open state is selected from the plurality of face image information read out from the storage means.

This control device further reduces the face image candidates other than the face image with the best eye-opening state, and simultaneously displays the face image with the best eye-opening state and the reduced face image candidates on the same screen. The display means is controlled to display.
The display means displays a plurality of face image candidates processed by the control device.

Therefore, the operator can confirm on the same screen the face image automatically extracted by the control device, which is said to have the best eye opening state, and a plurality of face image candidates other than this face image. As a result, if the face image automatically extracted by the control device has the best eye-opening state, the face image is selected, and other face image candidates are selected than the face image automatically extracted by the control device. If the eye open state is good, this face image can be selected. That is, it is possible to widen the selection range with respect to the face image and allow the operator to freely select the face image or to select the face image that suits the person to be photographed in the automatic photographing system.

Further, the operator can easily judge the eye-opened state of a face image which could not be judged by the conventional automatic blinking detection function, and the print loss can be reduced. Moreover, since the re-imaging process and the image re-displaying process are not required, the work efficiency is improved and the imaging processing capability can be improved.

The image processing method according to the present invention is a method of processing the face image information obtained by continuously photographing the upper body including the face of the photographed person for each photographed person. Acquiring face image information, and extracting from the face image information a plurality of face image candidates including the face image in the best eye-opening state, and in the plurality of face image candidates extracted here, the best eye-opening state Face image candidates other than the normal face image are reduced, the best enlarged face image in this opened state and a plurality of reduced face image candidates are displayed on the same screen at the same time, and any eye opening is performed on this screen. The feature is that the face image in the state is selected.

According to the image processing method of the present invention, when the face image information obtained by continuously photographing the upper half of the body including the face of the photographed person is processed for each photographed person, the most automatically extracted eyes are opened. The operator can confirm the face image that is in good condition and a plurality of face image candidates other than this face image on the same screen.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A face image photographing system and an image processing method according to embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration example of a face image capturing system as an embodiment of the present invention.

In this embodiment, a control device is provided for acquiring a plurality of face image information for each person to be photographed and extracting a plurality of face image candidates including a face image with the best opened eye from the face image information. , The operator can check the automatically extracted face image with the best eye-opening condition and multiple face image candidates other than this face image on the same screen by the operator. The operator can easily determine the eye-opened state of the person to be photographed, even for a face image that cannot be determined by the detection function.

The face image photographing system 100 shown in FIG. 1 can be applied to an ID card issuing system for issuing an ID card containing a face image. Face image capturing system 10
0 is a photographing device 10, a light source 21, a background plate 70, and a chair 7.
9 is equipped. The person to be photographed 20 is photographed while sitting on the chair 79. The background plate 70 becomes the background of the person to be photographed. Background plate 7
0 has a vertical length of about 1.8 m and a horizontal length of 0.9-1.
It is about 2 m, and the front has a light blue color (blue).

The photographing apparatus 10 is provided with a camera 82 as an example of photographing means, and the person to be photographed 20 under the light source 21.
The upper body including the face is continuously photographed, and a plurality of face image information (hereinafter referred to as face image data D1) is acquired for each person 20 to be photographed. As the camera 82, a digital camera for still images, a digital camera for moving images, or the like, which can continuously capture face images at intervals of at least 1 second in one shooting, is used. This is for extracting the face image with the best open state.

A system bus 88 is connected to the camera 82. A RAM 82 as an example of a storage unit and a fixed disk drive device (hereinafter simply referred to as HDD) 81 having a predetermined storage capacity are connected to the system bus 88, and face image information (hereinafter referred to as a face image) obtained by a camera 82 is obtained. The data D1) is stored. A plurality of face image data D1 is stored for each photographed person 20. In addition to the RAM 86, an EEPROM (electrically writable and erasable read-only memory) 85, which is an example of a storage unit, is connected to the system bus 88, and a plurality of color correction LUTis for color correction of a face image are provided. Remembered.

A CPU (Central Processing Unit), which is an example of a control device, is provided in the HDD 81 through a system bus 88.
84 is connected, and a plurality of face image data D1 for each person 20 to be photographed is read from the HDD 81 to the RAM 86, and a plurality of face image candidates including the face image with the best open state are selected from the face image data D1. It is made to extract.

In the photographing apparatus 10, the CPU 84 is the HDD 8
With respect to the plurality of face image data D1 of the subject 20 read out from No. 1, all the eye regions of the face image of the subject 20 are detected, and then the eyes of any one face image and others are detected. Of the face image is compared with the eyes of another face image, and the eyes of the face images extracted here are compared with the eyes of other face images. The face image with the best eye open state is selected from the plurality of face image data D1 read from the.

In this photographing apparatus 10, the eye blink is automatically detected from the face image candidates. For example, a change in brightness in a region of interest including eyes from the face part is detected. Then, the area of the part where the significant change occurs is obtained. And detect the blink,
Select an image other than blinking as an image with no blinking.
Alternatively, the image most distant from the blink image in time may be selected as an image without blinking (“Development of a completely unrestrained drowsy driving prevention system by multimedia sensing”, The 4th Image Sensing Symposium Symposium Shu, pp 285-288, 1998).

Further, the face image is subjected to image processing to detect the upper eyelid line and the lower eyelid line, and the interval between the upper eyelid line and the lower eyelid line detected here is measured. Then, the face with the largest interval may be determined as an image without eye blinking (“Extraction of eye feature points from face image”, IEICE Transactions, D-II, No. 6).
8 pp, 1796-1804, 1998.8). In this way, the optimum face image automatically detected is displayed clearly.

An operation monitor 87, which is an example of display means, is connected to the CPU 84 via a system bus 88, and a plurality of face image candidates extracted by the CPU 84 are displayed at once. A liquid crystal display device, a CRT device, or the like is used for the operation monitor 87. The operator is made to select one of the face image candidates. This is for selecting a face image suitable for forming on the face image-containing ID card. As the face image formed on the ID card, a face image with a good open eye state without red eyes or blinking is preferable.

For example, the CPU 84 enlarges the face image in the best opened state and reduces face image candidates other than the face image, and the enlarged face image and the reduced face image candidates are displayed on the same screen. The operation monitor 87 is controlled so as to be displayed at the same time. This is because it is easier for the operator to confirm a good face image as compared with other face images on the same screen.

Further, in order to allow the operator to verify the automatic blinking detection executed by the CPU 84, C
The PU 84 extracts an eye region from a face part with respect to a plurality of face image candidates and cuts out an eye part image, and a face image with the best open eye state and a plurality of eye part images of other face image candidates including the face image. The operation monitor 87 is controlled so that and are simultaneously displayed on the same screen.

An input tool 83, which is an example of operating means, is connected to the system bus 88, and the operation monitor 8
It is used to select an arbitrary face image from other face image candidates including the face image with the best eye-opening state displayed in 7. Although the CPU 84 of the image capturing apparatus 10 automatically detects the face image in the best eye-opened state, the operator may select the face image in the best eye-opened state by his or her own judgment.

The input tool 83 is also operated when the eye enlarged display mode is set for a plurality of face image candidates.
Here, the eye enlargement display mode refers to an operation for instructing enlargement display of an eye image. When the eye enlargement display mode is set, the CPU 84 extracts the eye region from the face portion and cuts out the eye image.

In this case, the eye image of the face image which is displayed on the screen of the operation monitor 87 and has the best eye opening state and the face image which is simultaneously displayed on the screen of the operation monitor 87 are included. The input tool 83 is operated by the operator to replace the eye image of the face image candidate. This makes it possible for the operator to easily check the blinking by displaying the optimum image and cutting out and displaying the eyes from multiple face images. Can be selected.

In addition to this, the input tool 83 is adapted to input the personal identification number ID of the person to be photographed 20. This is because the personal identification number ID is added to the face image data D1 to manage the file for each person to be photographed. The system bus 88
An I / O port 89 having a communication function is connected to. The I / O port 89 is used to connect to a host computer (not shown) or the like through a communication cable such as a LAN or USB, and the face image data D1 acquired by the face image capturing system 100 is transferred. This is to create employee ID data.

Next, an employee ID card 30 as an example of an ID card to which the face image data D1 acquired by the face image capturing system 100 is applied will be described. FIG. 2 is an image diagram showing a configuration example of the employee certificate 30. The employee ID 30 to which the face image data D1 acquired by the face image capturing system 100 is applied has, for example, a vertical length of about 6 cm, a horizontal length of about 9 cm, and a thickness in FIG. It has about 0.5 to 1.0 mm. The employee ID 3
The image display information is printed in the predetermined area P3 of 0. The image display information is, for example, “XX employee certificate”, “personal identification number”, “name”, “issue date” ...

The employee ID 30 includes a face image forming area P4.
Is assigned, and a face image of the issue requester 20 of the employee ID 30 is formed. This face image is formed by a predetermined ID card printing device based on the face image data D1. The surface of the card is covered with a protective sheet 51 so as to protect the face image and the image display information.

FIG. 3 is an image diagram showing a display example of face image candidates extracted by the CPU 84. According to the display example shown in FIG. 3, the top five face images with the best open state are extracted, and the face image F1 with the best open state among them is extracted.
Are enlarged and the other four face images F2 to F5 are reduced and displayed simultaneously on the same screen on the operation monitor 87.

In the photographing apparatus 10, the CPU 84 magnifies and displays the face image F1 with the best eye open state on the left half screen of the operation monitor 87, and displays the face images F2 to F5 other than the face image F1 in the right half. The screen is divided into four and displayed in a reduced size. This is because the five face image candidates extracted by the CPU 84 are displayed at once to make them easier to see. The operator is made to select one of the face image candidates. This is for forming the face image with the best eye open state on the ID card such as the employee ID shown in FIG.

FIG. 4 is an image diagram showing a display example of a face image and a plurality of eye part images in the best opened state. According to the display example shown in FIG. 4, the CPU using the input tool 83
When the eye enlargement display mode is set for 84, the CPU 84 extracts the eye region from the face part for the face image candidate and cuts out the eye image, and the face image F1 with the best eye open state and the face image concerned. The four eye part images of face image candidates other than F1 are displayed simultaneously on the same screen on the operation monitor 87.

In the photographing apparatus 10, the CPU 84 displays the face image F1 with the best eye open state on the left half screen of the operation monitor 87, and displays other eye image images of face image candidates other than the face image F1. The right half of the screen is displayed vertically in four rows. At this time, as compared with the eye image of the left face image F1, the right eye images of ~ are displayed in an enlarged manner. In addition, from the top of the screen to the bottom, the ones with the best eye-opening state are displayed in order (priority display).

This is because the eye image of the face image F1 which is displayed on the screen of the operation monitor 87 and has the best eye opening state, and other face image candidates which are simultaneously displayed on the screen of the operation monitor 87. This is because it is easy to replace the eye part images with. The operator can easily check the eye opening state,
It is possible to select a good face image without blinking.
In addition, P1 shown in FIG. 4 is an eye part image extraction area | region, and is an object area | region which replaces the eye part image of face image candidate.

In this example, the eye image of four face image candidates other than the face image F1 is displayed on the right half screen of the operation monitor 87.
Are arranged vertically, but the present invention is not limited to this and may be displayed in five stages including the eye image of the face image F1. This is because eye part images of the same size can be compared on the same screen.

FIG. 5 is an image showing an example of display control before the eye image of the face image F1 is replaced. In the display control example shown in FIG. 5, five face images F1 to F5 are developed in the RAM 86, eye part images are cut out from the face image F1, and the rest are similarly processed from face images F2 to F5. Face image F1 with the best eye opening
And five eye part images ~ of other face image candidates including the face image F1 are simultaneously displayed on the same screen on the operation monitor 87. Specifically, the face image data D1 related to the face image F1 and the data related to the eye image are combined from the RAM 86 to the operation monitor 87 and output as one frame of image display data.

FIG. 6 is an image showing an example of display control after the eye image of the face image F1 is replaced. In the display control example shown in FIG. 6, the eye image and the eye image of the face image F1 are exchanged. In this case, the read position (file storage address) of the face image data D1 related to the face image F1 and the face image data D1 related to the face image F2, which are loaded in the RAM 86, is changed, and the eye image is replaced. RAM by CPU84 so that enlarged display is possible
86 and the operation monitor 87 are controlled.

In this way, the operator can confirm the state of the face image before and after the eye image of the face image F1 is replaced with the eye image. The face image most suitable for the face image such as the employee ID shown in FIG. 2 can be selected.

Next, with respect to the image processing method according to the present invention, a processing example in the face image photographing system 100 will be described. FIG. 7 is a flowchart (main routine) showing a processing example in the face image capturing system 100. FIG. 8 is a flowchart (subroutine) showing an example of extraction of face image candidates, and FIG. 9 is an example of selection processing relating to face images.

In this embodiment, it is premised that the face image data D1 obtained by continuously photographing the upper body including the face of the person to be photographed 20 is processed for each person to be photographed 20. It is assumed that the CPU 84 extracts the top five face images. Of course, the face image capturing system 100 is powered on.

Under these processing conditions, in step A1 of the flow chart shown in FIG. 7, a plurality of (m) face image data D1 are acquired by continuously photographing each person to be photographed 20 using the camera 82. The face image data D1 acquired by the camera 82 is stored in the HDD 81 and / or the RAM 8 for each photographed person 20.
6 is stored.

Then, in step A2, a plurality of face image candidates including the face image with the best eye-opening state are extracted from the face image data D1. In this example, for a plurality of face image data D1 of the person to be photographed 20, the subroutine shown in FIG. 8 is called, and in step B1, the face image data D1 of the person to be photographed 20 is read from the HDD 81 by the CPU 84.
The data is sequentially read to AM86. In the RAM 86, for example, m pieces of face image data D1 obtained by continuously photographing the upper body including the face of the person to be photographed 20 from the front are developed.

Thereafter, in the CPU 84, m face image data D1 of the subject 20 developed in the RAM 86.
With regard to the above, all the eye regions of the face image of the person to be photographed 20 are detected in step B2. Then, in step B3, two arbitrary face images are extracted by the CPU 84, and 1
The eyes of one facial image are compared with the eyes of another facial image. Then, in step B4, the face image with the larger eyes is extracted.

In order to compare the eyes of the face image extracted here with the eyes of other face images, the process proceeds to step B5, in which the eye sizes of all the face images of the subject 20 are determined. The CPU 84 determines whether the comparison is made. If the size of the eyes is not compared for all the face images of the person to be photographed 20, the process returns to step B3.

A plurality of face image data D1 is obtained according to the comparison processing in step B3 and the selection processing result in step B4.
The face image with the best open condition is selected from among the above. Then, in step B6, a plurality of top face images are extracted. In this example, five face images are extracted from the face image with the best eyes open. Operation monitor 87
This is because the priority display is made with.

When five face images have been extracted with respect to the face image in the best opened state, the process returns to step A2 of the main routine shown in FIG. Then, with respect to the five face image candidates extracted previously, in step A3, face image candidates other than the face image in the best opened state are reduced, and the face image in the best opened state and a plurality of reduced faces are also reduced. The image candidate and the image candidate are simultaneously displayed on the same screen. As a result, the CPU 84
It is possible to manually select the face image with the best open state from the five face image data D1 of the subject extracted by. These face images are selected using the input tool 83.

Then, in step A4, it is checked whether or not a face image in an arbitrary open state is selected on this screen. C
If the PU 84 agrees with the extracted face image, the process directly proceeds to step A6. If the operator sees that another face image candidate is better, the process proceeds to step A5 to execute the face image selection process.

For example, regarding five face image candidates, FIG.
It is checked in step C1 whether the eye enlargement display mode for enlarging the eye image is set. When the operator sets and executes the eye enlarged display mode, the process proceeds to step C2, and the CPU 84 extracts the eye region from the face part and cuts out the eye image. Then, the process shifts to step C3, and the CPU 84 displays the face image F1 with the best open eye state and the eye image images of the other face images F2 to F5 including the face image F1 on the same image (FIG. 4, FIG. 5).

After that, the process proceeds to step C4 and the other eye image including the eye image of the face image F1 with the best open state displayed on the same screen and the face image F1 simultaneously displayed on the screen. The CPU 84 checks whether to replace the eye images F2 to F5. At this time, the operator uses the input tool 83 and designates the eye part image when the image is replaced.

The CPU 84 waits until one of the eye image is designated. If any of the eye image is designated, the process proceeds to step C6, and, for example, replacement display with the eye image is displayed. After performing the interchange display of the eye image, if the eye enlarged display mode is not selected in step C1 and if the eye images are not interchanged in step C4, the process proceeds to step C7.

At step C7, it is checked by the CPU 84 whether a face image has been selected. At this time, the operator uses the input tool 83 to select a face image. When the face image is not selected, the process returns to step C1 and the above-described processing is repeated. When the face image is selected, the process returns to step A5 of the main routine shown in FIG.

Then, the process proceeds to step A6, and the selected face image is stored in a database (not shown). Then, in step A7, the CPU 84 checks whether the face image capturing of the person to be photographed 20 is completed. When the face shooting is completed, the power-off information is detected and the control process is ended. If the face image shooting has not been completed, the process returns to step A1 and the above-described processing is repeated.

As described above, according to the face image photographing system 100 according to the embodiment of the present invention, a plurality of face image data D1 is acquired for each person 20 to be photographed, and the eye open state is selected from the face image data D1. The CPU 84 for extracting a plurality of face image candidates including the best face image of
U84 reduces the face image candidates other than the face image in the best opened state and displays the face image in the best opened state and the reduced face image candidates on the same screen at the same time. Is to control.

Therefore, the operator can confirm the face image automatically extracted by the CPU 84, which is said to have the best eye opening state, and a plurality of face image candidates other than this face image on the same screen (the book). Image processing method according to the invention).

As a result, if the face image automatically extracted by the CPU 84 has the best eye-opening state, the face image is selected and another face image candidate is selected more than the face image automatically extracted by the CPU 84. If the eye open state is good, this face image can be selected.
The selection range for face images can be widened, and the operator can freely select (face image automatic manual mixed selection function).

Further, the operator can easily judge the opened state of the face image, which cannot be judged by the conventional automatic blink detection function, and the print loss can be reduced. Moreover, since it is not necessary to perform the re-imaging process or the image re-displaying process, the work efficiency is improved and the imaging processing capability can be further improved.

In the above-described embodiment, the case of the employee ID card 30 with the face image has been described with respect to the ID card to which the face image photographing system 100 is applied, but the present invention is not limited to this, and a cash card, It can be applied to employee IDs, membership IDs, student IDs, alien registration IDs, passports, library cards and various driver's licenses. The ID card is also applied to an electronic card having an IC chip built therein.

[0063]

As described above, according to the face image photographing system of the present invention, a plurality of face image information is acquired for each person to be photographed, and the face opening information having the best eye-opening state is selected from the face image information. A control device for extracting a plurality of face image candidates including a face image is provided, and this control device reduces face image candidates other than the face image in the best opened state, and reduces the face image in the best opened state. The display means is controlled so that a plurality of face image candidates are simultaneously displayed on the same screen.

With this configuration, the operator can confirm on the same screen the face image automatically extracted by the control device, which is said to have the best eye opening state, and a plurality of face image candidates other than this face image. it can. Therefore, if the face image automatically extracted by the control device has the best eye-opening state, the face image is selected, and other face image candidates are opened than the face image automatically extracted by the control device. If the condition is good, this face image can be selected. The face image has a wider choice, and the operator can freely select it.

Further, the operator can easily determine the eye-opened state even for a face image that cannot be determined by the conventional automatic blink detection function, and print loss can be reduced. Moreover, since the re-photographing process and the image re-displaying process are not performed, the work efficiency is improved and the photographing processing capability can be improved.

According to the image processing method of the present invention, a plurality of face image information is acquired for each photographed person, and a plurality of face image candidates including the face image with the best eye-opening state are included in the face image information. Extraction, in the plurality of face image candidates extracted here, reduce the face image candidates other than the face image in the best open state, the best face image in the open state and reduced face image candidates And are simultaneously displayed on the same screen, and a face image in any open state is selected on this screen.

With this configuration, the operator can confirm, on the same screen, the face image that is automatically extracted and that the eye open state is the best, and a plurality of face image candidates other than this face image. Therefore, if the automatically-extracted face image has the best eye-opening state, then the face image is selected and the other face-image candidates have a better eye-opening state than the automatically-extracted face image. Then, you can select the face image of this person. As a result, the selection range for the face image can be widened, and the operator can freely select the face image.

The present invention is applied to an automatic issuing system for ID cards such as cash cards, membership cards, employee IDs, employee IDs, student IDs, alien registration IDs and various driver's licenses containing facial images. Very suitable.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration example of a face image capturing system 100 as an embodiment of the present invention.

FIG. 2 is an image diagram showing a configuration example of an employee certificate 30.

FIG. 3 is an image diagram showing a display example of face image candidates extracted by a CPU 84.

FIG. 4 is an image diagram showing a display example of a face image F1 and a plurality of eye part images which are the most favorable in an eye open state.

FIG. 5 is an image showing a display control example before the eye image of the face image F1 is replaced.

FIG. 6 is an image showing an example of display control at the time of replacing the eye image of the face image F1.

7 is a flowchart (main routine) showing a processing example in the face image capturing system 100. FIG.

FIG. 8 is a flowchart (subroutine) showing an example of extracting face image candidates.

FIG. 9 is a flowchart (subroutine) showing an example of selection processing related to a face image.

[Explanation of symbols]

10 Imaging device 21 light source 30 Employee ID (ID card) 81 HDD (storage means) 82 Camera (shooting means) 83 Input tool (operation means) 84 CPU (control device) 85 EEPROM (storage means) 86 RAM (storage means) 87 Operation monitor (display means) 88 system bus 89 I / O port 100 face image capturing system

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁷ identification code FI theme code (reference) H04N 7/18 H04N 7/18 UF term (reference) 2C005 HB01 HB02 HB03 HB09 JB02 KA40 LA22 LA32 LA33 LA34 LA38 5B057 BA24 CA08 CA12 CA16 CB08 CB12 CB16 CC03 CD05 CE08 CE09 CH07 CH08 CH11 CH12 DA07 DA08 DA16 DB02 DB09 DC05 DC09 DC32 DC36 5C054 CC05 CG02 FC01 FC04 FC12 FD07 FE09 FE12 FE17 FF03 GA05 GB02 HA00 5L096 CA02 FA59 FA59

Claims (10)

[Claims] 1. A photographing means for continuously photographing an upper body including a face of a photographed person to obtain a plurality of face image information for each photographed person, and a memory for storing the face image information obtained by the photographing means. A control unit that reads out a plurality of face image information for each of the photographed persons from the storage unit and extracts a plurality of face image candidates including a face image with the best open eye state from the face image information; And a display unit that displays a plurality of face image candidates extracted by the control device, wherein the control device reduces face image candidates other than the face image that is the best in the opened state, and the best opened state. Face image photographing system, characterized in that the display means is controlled to simultaneously display a different face image and a plurality of reduced face image candidates on the same screen. 2. The control device detects all eye regions of a face image of the subject with respect to a plurality of face image information of the subject read out from the storage unit, and then optionally. The eye of one face image is compared with the eye of another face image to extract the larger eye, and the eye of the face image extracted here is compared with the eye of another face image. 2. The face image having the largest open eyes is selected from the plurality of face image information read out from the storage means, and the face image with the best eye opening is selected. Face image shooting system. 3. The control device extracts an eye region from a face part of a plurality of face image candidates and cuts out an eye part image, and includes the face image in the best opened state and the face image. The face image capturing system according to claim 1, wherein the display means is controlled so that the plurality of eye image images of the face image candidates are simultaneously displayed on the same screen. 4. Another face image candidate including an eye image of the face image with the best open state displayed on the screen of the display unit and a face image displayed on the screen of the display unit at the same time. The face image capturing system according to claim 1, further comprising an operating unit that is operated to replace the eye image of the eye. 5. The operation unit is used to select an arbitrary face image from other face image candidates including the face image with the best opened eye displayed on the display unit. The face image capturing system according to claim 4. 6. A method of processing face image information obtained by continuously photographing an upper body including a face of a subject for each subject, and obtaining a plurality of face image information for each subject. Then, from the face image information, a plurality of face image candidates including the face image with the best open state are extracted, and among the extracted face image candidates, a face image other than the face image with the best open state is extracted. Face image candidates are reduced, the best face image in the opened state and a plurality of reduced face image candidates are simultaneously displayed on the same screen, and a face image in any opened state is selected on the screen. An image processing method characterized by the above. 7. With respect to a plurality of face image information of the photographed person, all eye areas of the face image of the photographed person are detected, and then the eyes of any one face image and other face images are detected. The larger eye is extracted by comparing with the eyes, the eyes of the face image extracted here are compared with the eyes of other face images, and the larger eye is extracted, and the plurality of faces are extracted. 7. The face image with the best eye open state is selected from the image information.
The image processing method described in. 8. An eye region image is extracted by extracting an eye region from a face part with respect to a plurality of the face image candidates, and the face image in the best opened state and other face image candidates including the face image are extracted. The image processing method according to claim 6, wherein a plurality of eye part images are simultaneously displayed on the same screen. 9. An eye part image of the face image with the best open state displayed on the same screen, and an eye part image of another face image candidate including the face image displayed on the screen at the same time. 9. The image processing method according to claim 8, further comprising: 10. The image according to claim 6, wherein an arbitrary face image is selected from other face image candidates including the face image with the best eye-opened state displayed on the same screen. Processing method.