JPH11316836A - Person identifying - Google Patents

Abstract

PROBLEM TO BE SOLVED: To surely identify a specified person at the time of performing re- identification processing, accompanying the specification of a person oneself by providing an image re-acquisition condition changing means to change image re-acquisition result of physical characteristics from those at the last time when acquiring a re-image by means of an image acquiring means. SOLUTION: When an image is acquired from an image pickup camera 12, the characteristic quantity of an image acquired by an identification processing part 23 is compared with the characteristic quantity proper to a specified person 18 stored and managed by a database 24 and collation is confirmed. When it is decided that the person can not be correctly identified from the acquired image, an acquisition condition of physical characteristic quantity is significantly changed by controlling at least one from a controller 21 and an image acquisition data is made different from the last one and is acquired through this. Thus, an acquired image is surely different this time from the last time, the preceding image acquisition condition itself in which an identification element is in short is improved and the physical characteristic quantity of the person 18 can be clearly acquired.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a personal identification device for identifying a physical feature of a person by using a collation element, and more particularly to a re-identification process that is repeatedly executed when identification cannot be performed correctly at one time when the individual cannot be identified correctly. The present invention relates to a personal identification device having a high specific success rate at the time.

[0002]

2. Description of the Related Art In general, a personal identification device of this type captures the face of a walking person with an imaging camera and identifies the person in real time, or identifies a person standing in front of the device by identifying the person. An identification function is provided. At the time of this identification, a feature amount is extracted from the face image, and the extracted feature amount is compared with registered data to obtain a similarity, thereby identifying whether or not the user is the person.

At this time, if the user cannot be correctly identified, the identification procedure is repeated to execute the re-identification process. In this case, the probability of correctly identifying the individual in one identification process is 1- (1-0.9) N if the identification success rate is 90%, for example. For example, the number of re-identification processes N = Even if the number is 3, the identification success rate is 99.99%. Thus, even if the identification success rate per one time is low, if the number of re-identification processes is increased in calculation, the identification success rate will be greatly improved.

[0004] However, in practice, if the re-identification processing images are exactly the same image, and if it is the first error, it will not succeed even if it is repeated many times. Even if the identification processing images are different from each other by several frames in time, there is no visually significant difference in the face image, and the re-identification processing is often not successful. In particular, when the brightness of the input image is not appropriate, the shadow of the face image is excessive, or the orientation of the face is not suitable for identification, good results cannot be obtained even if re-identification is performed many times. Often an error occurs.

[0005]

SUMMARY OF THE INVENTION Accordingly, the present invention provides a method of re-identifying a specific person by changing the setting to image acquisition conditions suitable for the re-identification processing so that the specific person can be reliably identified. The purpose of the present invention is to provide a personal identification device.

[0006]

According to the first aspect of the present invention,
A person identification device that acquires the physical characteristics of a person from the image acquisition means and identifies the person from the similarity between the acquired physical characteristics and the physical characteristics of the specific person specified in advance stored in the storage means. Further, when the image acquiring means acquires the re-image, a re-image acquiring condition changing means for making a re-image acquiring result of the physical feature amount different from the previous one is provided.

According to a second aspect of the present invention, a person's face image is obtained from an image obtaining means, and the person is specified based on the similarity between the obtained face image and a previously specified face image of a specific person stored in the storage means. And a re-image acquisition condition changing unit for making the re-image acquisition result of the face image different from the previous image when the image acquisition unit acquires the re-image.

According to a third aspect of the present invention, there is provided a re-image acquisition condition changing means having an illumination control means for changing and controlling a light amount.

According to a fourth aspect of the present invention, there is provided a re-image acquisition condition changing unit including an exposure time control unit for changing and controlling the exposure time of the image acquisition unit.

According to a fifth aspect of the present invention, there is provided a re-image acquisition condition changing unit including a movement adjustment control unit for adjusting at least one of an illumination control unit and an image acquisition unit.

The invention according to claim 6 is characterized in that it is a re-image acquisition condition changing means provided with a gesture action guiding means for changing the facial expression and the direction of a person.

[0012] According to a seventh aspect of the present invention, there is provided an estimating means for estimating an image acquiring condition which can obtain a changed element most effectively from image acquiring conditions which have caused an unidentified person who cannot identify a person. And a change control unit for changing and controlling the image acquisition condition.

[0013]

According to the present invention, when the physical characteristics of a person are obtained from the image obtaining means, the obtained physical characteristics and the physical characteristics of the specified person stored in the storage means are stored. The person is identified from the similarity with the quantity. At this time, if the user cannot be identified, the image is acquired again by the image acquiring means by the re-image acquiring condition changing means, and the acquisition result of the physical feature amount is acquired at the time of acquiring the re-image by making it different from the previous time.

For this reason, the identification processing image at the time of the re-identification processing is significantly different between the previous and current times, and the image acquisition condition itself in which the previous identification element was insufficient is changed. Re-identification processing can be performed by changing settings under image acquisition conditions suitable for re-identification. As a result, the identification success rate at the time of re-identification is improved, and unnecessary re-identification repetition is eliminated, so that the identification processing time can be reduced.

Further, a physical feature such as with a body may be used as an identification element, and a face image that is clearly different for each individual suitable for identification may be used as an identification element. Furthermore, when changing the image acquisition conditions when acquiring a re-image, if the illumination control means controls the change in the amount of light to the subject, it is possible to change the setting of the illuminated acquired image of the person to optimal brightness.

Similarly, when changing the image acquisition conditions for reacquisition of an image, the exposure time of the image acquisition means such as an image pickup camera is controlled by the exposure time control means so that the optimum brightness suitable for image acquisition is obtained. Settings can be changed.

Similarly, at the time of changing the image acquisition conditions when re-acquiring an image, the movement adjustment control means moves and adjusts at least one of the illumination control means and the image acquisition means to up, down, left, right, rotation, etc. If the setting is changed, the image acquisition angles are greatly different, and images from different directions can be acquired.

Similarly, when changing the image acquisition conditions for re-acquisition of an image, if the facial expression and the direction of the person are changed by the action guide means, the physical characteristic amount which is clearly different for each individual peculiar to the person. Is easily obtained.

Similarly, when changing the image acquisition condition when acquiring a re-image, the image acquisition condition that most effectively obtains the changed element is estimated from the previous image acquisition condition that caused the unclear identification of the person. If the change control unit is changed and controlled so as to satisfy the image acquisition condition estimated by the unit, the image acquisition condition can be changed to the image acquisition condition most suitable for re-identification.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below in detail with reference to the drawings. FIG. 1 shows a personal identification device 11 applied to entrance / exit management for opening and closing a room door.
Denotes an imaging camera 12, a numeric keypad 13, and a card reader 1.
A collation unit 15 integrally provided with the camera 4 is installed at a face-height wall position near the door, and an illumination light 16 and a speaker 17 for changing imaging conditions are installed above the wall.

The above-described imaging camera 12 is set in a direction in which a face of a person approaching the door is imaged, and acquires face image data.
When this face image data is acquired by the imaging camera 12, the eyes,
The face information unique to the person, such as the shape and size of each part such as the nose, mouth, etc. and the entire face, the hairstyle, the presence or absence of glasses, the color, wrinkles, the degree of makeup, etc., is acquired.

Then, the obtained face information is transmitted to the specific person 18.
When the person approaches the door, the imaging camera 12 captures the face information of the person, compares it with registered data of the specific person 18 in advance, and unlocks when the registration is confirmed. Is set as follows.

When the power of the imaging camera 12 is set to OFF, in addition to the unlocking data based on the image, the specific person 18 inputs the personal identification number (PIN) to the numeric key 13 of the collating unit 15 to unlock the camera. It can be locked and can be unlocked if the card reader 14 reads and checks the ID card of the specific person 18.

FIG. 2 is a block diagram of a control circuit of the personal identification device 11. The controller 21 controls each circuit device according to a set program, and stores the control data in a readable manner.

First, when an image is acquired from the imaging camera 12, the acquired image is taken into the image input unit 22,
The information is guided to the identification processing unit 23, and the characteristic amount of the image acquired by the identification processing unit 23 is compared with the characteristic amount unique to the specific person stored and managed in the database 24 to confirm the collation.

At this time, if it is determined that the person cannot be correctly identified from the acquired image, the controller 2
1 controls output so that an image is repeatedly acquired. In this case, the controller 21 has an image acquisition condition changing function for changing an image acquisition condition in order to acquire an image suitable for re-identification.

The image acquisition condition changing function is such that the camera control unit 25, the illumination control unit 26, and the sound output unit 27 are connected to each other so as to be remotely controllable from the controller 21, and at least one is controlled by the controller 21 to control the body. The acquisition conditions of the target feature amount are greatly different, so that the image acquisition data is acquired differently from the previous time.

For example, when controlling the camera control unit 25, the exposure time of the imaging camera 12 is changed to change the brightness of the subject. At this time, the setting can be changed to the optimum brightness suitable for acquiring the image of the physical feature amount.

Further, the image capturing camera 12 is moved and adjusted, such as up and down, left and right, and rotation, to change the image capturing position and the image obtaining angle to obtain images from different directions. At this time, the acquired image is definitely different between the previous image and the current image, and the previous image acquisition condition itself in which the identification element is insufficient is improved, so that the physical feature amount of the specific person 18 can be clearly acquired.

When controlling the illumination control section 26, the light quantity of the illumination light 16 is changed. At this time, the brightness of the entire periphery can be changed to change the setting to a brightness suitable for acquiring an image of the subject.

The image is acquired by changing the illumination position and the illumination angle by adjusting the movement of the illumination light 16 such as up and down, left and right. In this case, the image acquisition condition itself changes greatly compared to the previous time, and as a result, the physical feature amount of the specific person 18 can be clearly acquired.

When controlling the voice output unit 27, voice guidance is provided from the speaker 17 so as to change a person's actions such as changing a facial expression or a person's direction. In this case, an identification element unique to the individual that is different for each individual is obtained, and as a result, an image in which the physical features of the specific person 18 are clarified is obtained.

In addition, as an image acquisition condition changing function, the controller 21 estimates an image acquisition condition that enables the changed element to be obtained most effectively from the previous image acquisition conditions that have caused the uncertainty. One or more of the camera control unit 25, the illumination control unit 26, and the audio output unit 27 are changed and controlled so as to satisfy the condition.

For example, when extracting a characteristic amount from image data, the imaging camera 12 is moved in a correction direction by a fixed amount based on various identification factors such as a density difference of an acquired image, a face direction, a degree of inclination, and the like. It is only necessary to move and adjust the illumination light 16 in the correction direction by a fixed amount, or to provide voice guidance of the act of acting so as to change the facial expression or the direction of the person from the speaker 17. As a result, it is possible to change the setting to the image acquisition condition most suitable for the re-identification and acquire the re-image.

By operating such an image acquisition condition changing function, it is possible to reliably acquire an image of a physical characteristic amount suitable for identification at the time of re-identification, so that the identification success rate is improved. Useless re-identification repetition due to repeatedly acquiring an approximate image is eliminated, and the identification processing time can be reduced.

By the way, in the database 24, the face characteristic amount specified in advance and specified by the specific person is registered, and this is used as the collation confirmation data. The amount is compared and collated to determine whether or not the person is the principal.

FIG. 3 shows a similarity determination processing operation of the personal identification device, and a raw image obtained by acquiring face information from the imaging camera 12 is temporarily stored in the image memory 31. The face area is searched by the face detection unit 32 from the accumulated raw image, and the captured face area is detected. Here, the outline of the face and its surroundings is detected. 1. Background subtraction method for extracting the difference between the background image and the acquired image 2. Skin color detection method using color 3. Motion detection method using optical flow and frame difference The face region is detected by using either the neural network or the method of finding the face by pattern matching.

After detecting this face area, the face position detection unit 3
In step 3, the position of the face is accurately detected by matching based on the feature model of the eyes, nose, mouth, etc. When the position of the face is accurately detected and positioned, the face feature extraction unit 34 extracts the face feature amount extracted from the face image. This facial feature is calculated by extracting the difference from the average face using a statistical method such as principal component analysis,
Alternatively, it is extracted from a gray image such as eyes, nose, mouth, etc. by template matching. This extracted face feature amount,
The similarity determination unit 35 compares and matches the feature memory previously registered in the database 24 to determine the similarity of the face information.

In this case, the controller 21 has set a threshold level as a criterion for identifying the person, and has collated the feature amount of the face obtained from the input image data with the feature amount of the registered data. The similarity value at that time is calculated, and if this value is higher than the threshold level, the person is specified, and if the value is lower than the threshold level, the person is determined to be another person or an unregistered person.

The identification processing operation of the personal identification device 11 thus configured will be described with reference to the flowchart shown in FIG. Now, when face image data is acquired from the imaging camera 12, a search for a face area is performed from the image data to detect a captured face area (steps n1 to n2).
After the face area is detected, if the position of the face is accurately detected and positioned based on features such as eyes, nose, mouth, etc., the face feature amount is extracted from the face image (steps n3 to n4).

The extracted face feature amount is compared with a pre-registered feature amount to obtain a similarity (step n5).
To n6), when the similarity is equal to or higher than the threshold level, the person is recognized as a specific person, and the identification processing ends (step n7).
).

By the way, when the face feature amount is extracted from the face image, the image acquisition conditions when the image is taken by the image taking camera 12,
Illumination conditions when illuminated by illumination light 16, speaker 1
When the act is voice-guided in step 7, the voice guidance condition is taken in as a parameter of the identification condition (step n8).
), When the identification error is determined, the controller 2
1 selects a parameter suitable for re-identification, and the controller 21 changes the setting of the condition of the selected parameter and re-identifies (steps n9 to n10).

For example, as shown in FIG. 5, when an indistinguishable error occurs due to an input image of a tilted face at the time of the first image acquisition, an image of a normal frontal face is obtained. It is determined that adjusting the movement of the camera 12 is a parameter that is the most effective image acquisition condition at the time of re-identification.
Is re-identified so as to obtain a frontal face by correcting the orientation of the image. Thus, at the time of the second image acquisition, an input image close to the front face can be acquired, and by acquiring the image of the front face, the identification probability is increased, and the specific person can be clearly distinguished and identified.

As shown in FIG. 6, when an unidentifiable error occurs at the time of the first image acquisition because of an input image in which the left half of the face is dark, regular uniform illumination can be obtained. It is determined that adjusting the movement of the illumination light 16 is a parameter that is the most effective image acquisition condition at the time of re-identification, and the controller 21 corrects the direction of the illumination light 16 based on the parameter to obtain uniform illumination. The re-identification process is executed so as to obtain the same. Thus, at the time of the second image acquisition, an input image with uniform brightness suitable for identification determination can be acquired, and by acquiring the image with appropriate brightness, the identification probability increases. Therefore, a specific person can be reliably identified by several times of the identification processing operation.

As described above, at the time of re-image acquisition, since the image acquisition data of the physical characteristic amount is acquired by changing the re-image acquisition conditions so as to be different from the previous time, the identification processing images at the time of the re-identification processing are the previous and current images. The image acquisition condition itself, in which the previous discrimination element was insufficient, is improved, and the physical feature amount can be clarified and acquired. As a result, the identification success rate at the time of re-identification is improved, and unnecessary re-identification repetition is eliminated, so that the identification processing time can be reduced. Further, in addition to using a face image that is clearly different for each individual suitable for identification as an identification element, a physical feature such as with a body can be used as an identification element.

In the correspondence between the present invention and the configuration of the above-described embodiment, the image acquiring means of the present invention corresponds to the imaging camera 12 of the embodiment, and similarly, the storage means corresponds to the database 24. The image acquisition condition changing means, the estimation means and the change control means correspond to the controller 21, the illumination control means corresponds to the illumination light 16 and the illumination control unit 26, and the exposure time control means corresponds to the imaging camera 12 and the camera control. Corresponding to the unit 25, the action guidance means is a speaker 17
The present invention can be applied based on the technical idea described in the claims, and is not limited to the configuration of the above-described embodiment.

[Brief description of the drawings]

FIG. 1 is a schematic side view showing a use state of a personal identification device of the present invention.

FIG. 2 is a control circuit block diagram of the personal identification device of the present invention.

FIG. 3 is an explanatory diagram showing a similarity determination processing operation of face information of the personal identification device of the present invention.

FIG. 4 is a flowchart showing an identification processing operation of the personal identification device of the present invention.

FIG. 5 is an explanatory diagram showing an image acquisition condition changing operation in which the direction of the imaging camera according to the present invention is corrected.

FIG. 6 is an explanatory diagram showing an image acquisition condition changing operation in which the direction of the illumination light is corrected according to the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 ... Personal identification apparatus 12 ... Imaging camera 16 ... Illumination light 17 ... Speaker 18 ... Specific person 21 ... Controller 22 ... Image input part 23 ... Identification processing part 24 ... Database 25 ... Camera control part 26 ... Lighting control part 27 ... Audio output Department

Claims (7)

[Claims] 1. A method for acquiring a physical characteristic of a person from an image acquiring means, and identifying the person from the similarity between the acquired physical characteristic and a physical characteristic of a specific person specified in advance stored in a storage means. A personal identification device, comprising: a re-image acquisition condition changing unit configured to make a re-image acquisition result of a physical characteristic amount different from a previous image at the time of re-image acquisition by the image acquisition unit. 2. An identification apparatus for acquiring a face image of a person from an image acquisition means, and identifying the person based on a similarity between the acquired face image and a face image of a specified person stored in advance in a storage means. A personal identification device comprising a re-image acquisition condition changing means for making a re-image acquisition result of a face image different from a previous image at the time of re-image acquisition by the image acquisition means. 3. The personal identification device according to claim 1, wherein the re-image acquisition condition changing means is an illumination control means for changing and controlling a light amount. 4. The personal identification device according to claim 1, wherein the re-image acquisition condition changing means is an exposure time control means for changing and controlling an exposure time of the image acquisition means. 5. The personal identification according to claim 1, wherein the re-image acquisition condition changing means is a movement adjustment control means for moving and adjusting at least one of the illumination control means and the image acquisition means. apparatus. 6. The personal identification device according to claim 1, wherein the re-image acquisition condition changing means includes an action guiding means for changing a facial expression and a direction of a person. 7. A re-imaging condition changing means for estimating an image obtaining condition for obtaining a changed element most effectively from image obtaining conditions which have caused an unidentified person who cannot identify a person, and the estimating means 3. The personal identification device according to claim 1, further comprising: a change control unit configured to change and control the image acquisition condition estimated in the step (c).