JP2009110147A - Image processor and method therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor capable of specifying easily attributes of a person concerned with a face image in a picked-up image. <P>SOLUTION: A face detecting part 101 detects face image data from picked-up image data, during image pick-up by an imaging device 1, and a face recognition part 102 finds similarity between it and registered face image data registered preliminarily in a face registration memory 56. Then, a person corresponding to registered face image data is specified based on the similarity, and meta-data such as a name associated therewith is displayed in association with the face image, during reproduction. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing apparatus and method for detecting a face image from captured data.

Digital still cameras and digital video cameras that use facial recognition technology have been commercialized.
In Patent Document 1, a feature point of a person's face is extracted from a captured image obtained by capturing a person's face, and face recognition is performed based on the degree of coincidence between the extracted feature point and registered face image data related to a person registered in advance. Techniques for performing are disclosed.

JP-A-6-259534

  By the way, there is a strong demand to record the growth of children with digital still cameras and digital video cameras. At this time, it is convenient if the relationship between the child image in the captured image and the attribute such as the age of the child can be easily specified. In addition, when there are a plurality of children, it is easy to search and manage the recorded images if the children are photographed. Other than that, it is generally convenient if the attribute of the person can be easily specified for the face image in the captured image. Conventionally, there has been no device that satisfies such a demand.

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide an image processing apparatus and method for easily specifying an attribute of a person with respect to a face image in a captured image. It is in.

  In order to achieve the above object, an image processing apparatus of the present invention shows face detection means for detecting face image data in imaging data, registered face image data of a person registered in advance, and at least the attribute of the person Based on the storage means for storing attribute data in association with each other, the face image data detected by the face detection means, and the registered face image data recorded in the storage means, the face detection means detects the Face recognition means for determining whether or not face image data corresponds to a person in the registered image data; and the face image data detected by the face detection means corresponds to a person in the registered image data. If the face recognition means determines that there is, the attribute data of the registered person read from the storage means and the image data or the face image detected by the face detection means And a control means for processing for associating the data.

  The image processing method of the present invention is based on a face detection process for detecting face image data in imaging data, the face image data detected in the face detection process, and registered face image data registered in advance. A face recognition step for determining whether the face image data detected in the face detection step corresponds to a person in the registered image data; and the face image data detected in the face detection step When it is determined in the face recognition step that the image data corresponds to a person, the attribute data of the registered person is associated with the image data or the face image data detected in the face detection step. And a control process for performing processing.

  According to the present invention, it is possible to provide an image processing apparatus and a method thereof that make it possible to easily specify the attributes of a person in a face image in a captured image.

FIG. 1 is a block diagram illustrating a configuration example of an imaging apparatus (digital video camera) 1 according to an embodiment of the present invention.
The imaging apparatus 1 includes, for example, an imaging optical system 10, an imaging element 12, a microphone 14, an analog signal processing unit 16, an A / D converter 18, a built-in memory 20, an audio / image processing unit 24, and an HDD.
I / F 26, HDD 28, card I / F 30, card type recording medium 32, input / output I / F 34, liquid crystal monitor 36, speaker 38, operation unit 40, output terminal 42, built-in battery 50, calendar function clock 52, CPU 54, A face registration memory 56 and an external battery 58 are provided.
The built-in memory 20, sound / image processing unit 24, HDD I / F 26, card I / F 30, input / output I / F 34, and CPU 54 are connected to the data bus 22.

The face detection unit 101 corresponds to the face detection unit of the present invention, the face recognition unit 102 corresponds to the face recognition unit of the present invention, the control unit 100 corresponds to the control unit of the present invention, and the face registration memory 56 corresponds to the storage unit of the present invention.
Further, the imaging data of the present embodiment is the imaging data of the present invention, the face image data of the present embodiment is the face image data of the present invention, the registered face image data of the present embodiment is the registered face image data of the present invention, and the present embodiment. The metadata corresponds to the attribute data of the present invention.

As shown in FIG. 1, a liquid crystal monitor 36, a speaker 38, an operation unit 40, and an output terminal 42 are connected to the input / output I / F 34.
The liquid crystal monitor 36 displays a playback image corresponding to the captured data, a setting screen for changing various settings of the digital video camera, and the like. The setting screen includes an editing screen for editing face recognition data described later.
The operation unit 40 includes an input button, a recording button, and the like. The input button of the operation unit accepts, for example, a switching input of various processes (face registration setting process, face recognition imaging process, face registration update process, etc.) of the digital video camera. In addition, the recording button of the operation unit 40 receives an instruction input for starting and ending recording. A television monitor or the like is connected to the output terminal 42.

  For example, the sound / image processing unit 24 performs predetermined image processing on the digital imaging signal read from the built-in memory 20. Further, the audio / image processing unit 24 performs data compression processing on the digital imaging signal and the digital audio signal based on a predetermined format, and generates recording data associated with the metadata.

The sound / image processing unit 24 generates image data for various displays (such as a frame display during face recognition) according to instructions from the CPU 54, and superimposes the image data on the original image data read from the built-in memory 20. Output to the monitor 36. As a result, the screen of the liquid crystal monitor 36 is a combination of the above various displays.
The audio / image processing unit 24 performs audio compression processing on the digital audio signal read from the built-in memory 20 and outputs it to the HDD I / F 26,
The digital audio signal read from the I / F 26 is decoded. The audio / image processing unit 24 outputs an analog audio signal corresponding to the decoded digital audio signal to the speaker 38 during reproduction.

  When the control unit 100 of the CPU 54 determines that the calendar function-equipped clock 52 is not set, for example, when the image pickup apparatus 1 is newly purchased or when the internal battery for the clock is replaced, the calendar function is displayed on the liquid crystal monitor 36. A display prompting the user to set the clock 52 is performed. The user sets the current date using the input button of the operation unit 40. Note that the calendar function-equipped watch 52 is driven by, for example, power supplied from the internal battery 50. Further, modules other than the internal battery 50 of the imaging device 1 are driven by power supplied from the external battery 58, for example.

Hereinafter, various processes of the imaging apparatus 1 will be described.
<Face registration setting process>
FIG. 2 is a flowchart for explaining face registration setting processing by the imaging apparatus 1 shown in FIG.
When the user wants to register a face image of a new person, the face registration mode is designated (selected) with the input button of the operation unit 40. When determining that the face registration mode has been designated (step ST11), the control unit 100 of the CPU 54 displays an instruction “please image the target person” on the liquid crystal monitor 36 (step ST12).
The function of the control unit 100 is realized by the CPU 54 executing a program stored in a ROM (not shown). In addition, the control unit 100 executes each operation of the imaging apparatus 1, for example, functions such as AF (autofocus) calculation and AE (automatic exposure) calculation according to the execution of the program.

  Then, according to the display on the liquid crystal monitor 36 in step ST12, the user starts imaging (imaging) (step ST13). Thereby, the light that has passed through the imaging optical system 10 is photoelectrically converted by the imaging device 12 to generate an analog imaging (image) signal of the subject image, which is amplified by the analog signal processing unit 16, and then A / D converted. The digital image data is converted by the device 18 and written into the built-in memory 20. The external sound is converted into an electric signal by the microphone 14 to be an analog sound signal, which is amplified by the analog signal processing unit 16 and then converted into digital sound data by the A / D converter 18 to be stored in the built-in memory 20. Is written to.

Next, the face detection unit 101 of the CPU 54 searches for a face area in the captured image based on the digital image data read from the built-in memory 20 (step ST14). The face detection unit 101 is a function realized by the CPU 54 executing the above program.
At this time, the face detection unit 101 extracts feature points from the digital image data and detects the face area, the face size, and the like of the subject. For example, the face detection unit 101 extracts a face region by a feature point extraction process described in Japanese Patent Laid-Open No. 2001-16573. Examples of the feature points include eyebrow, eye, nose, lip end points, face contour points, head vertices, and lower chin points. Note that the algorithm used by the face detection unit 101 for face area detection is not limited to that described above, and an arbitrary algorithm can be used.

When the face detection unit 101 detects a face area in the captured image (“Y” in step ST14), the sound / image processing unit 24 uses the instruction from the control unit 100 to mark a mark image near the detected face area. New image data to be displayed superimposed on the image of the original imaging data is generated, and a screen corresponding thereto is displayed on the liquid crystal monitor 36 (step ST15). As a specific example, the screen 200 shown in FIG. 3 is displayed on the liquid crystal monitor 36.
When the face detection unit 101 detects a plurality of face regions, the sound / image processing unit 24 displays a plurality of mark images 202 and 204 in response to an instruction from the control unit 100, and one of the mark images is displayed. The color 202 is made different from the others and blinked (step ST15). If the face detection unit 101 cannot identify the face area, the control unit 100 ends the face registration setting process after a predetermined time (“N” in step ST14).

When the user looks at the screen displayed on the liquid crystal monitor 36 in step ST15 and the person image with the blinking mark image is that of the person he wants to register, for example, the operation unit 40 shown in FIG. When “Register” is selected with the input button (“Register” in step ST16) and another person is to be registered, “Select another person” is selected with the input button of the operation unit 40 shown in FIG. “Select another person” in step ST16).
When “select another person” is selected using the input button of the operation unit 40 in step ST16, the sound / image processing unit 24 blinks the mark image of the next unselected face area (step ST17). .

On the other hand, when “Register” is selected using the input button of the operation unit 40 in step ST16, the face recognition unit 102 extracts the feature points of the face image in the face area corresponding to the blinking mark image. Based on this, registered face image data is generated (step ST18). For example, the face recognition unit 102 uses registered face image data of a registered person indicating the position of the feature point of the face image detected by the face detection unit 101, the size of the face part obtained from the feature point, the relative distance of each feature point, and the like. Generate.
The face recognizing unit 102 determines whether or not the generated registered face image data is sufficient data for recognition based on a predetermined standard, and when determining that the registered face image data is sufficient data, “Registration OK” is displayed on the liquid crystal monitor 36.

Next, the control unit 100 of the CPU 54 displays a message “Do you want to register a name?” On the liquid crystal monitor 36. When the user wants to register, the user selects “Yes” with the input button of the operation unit 40 (“Y” in Step ST19), and further inputs a name (Step ST20).
Subsequent to step ST20 or when “No” is selected in step ST19, the control unit 100 displays a message “Do you want to register the date of birth?” On the liquid crystal monitor 36 (step ST21). When the user wants to register the date of birth, the user selects “Yes” with the input button of the operation unit 40 (“Y” in step ST21), and further inputs the date of birth (step ST22).

After completing the above-described series of operations / processes, the CPU 54 uses the registered face recognition data generated in step ST18, the name data or / and date of birth data input in steps ST20 and ST22, and the calendar function clock 52. The acquired date data (current registration date data) is associated and recorded in the face registration memory 54 (step ST23). As the association, for example, data is stored as one file.
The face registration memory 56 is composed of, for example, a non-volatile semiconductor memory. As described above, the face registration memory 56 is associated with the registered face image data, the person's name data and date of birth data, and the person's registered face. Record metadata such as registered date data that updated the image data. Further, the date of birth data and the date of registration are included up to the date of birth. However, if the year and date of birth are sufficient, the date and date may be omitted.

  As described above, according to the imaging apparatus 1, a face image is automatically detected from digital imaging data, the registered face image data is generated, and the input name data and date of birth data are automatically generated. Register in association. Therefore, registration of registered face image data can be easily performed with a simple operation.

<Face recognition imaging process>
FIG. 4 is a flowchart for explaining the face recognition imaging process in the imaging apparatus 1 shown in FIG.
When the user wants to capture an image in the face recognition imaging mode, the face recognition imaging mode is selected with the input button of the operation unit 40. When detecting that the face recognition imaging mode has been selected (“Y” in step ST31), the control unit 100 of the CPU 54 starts imaging in the face recognition imaging mode. As a result, digital imaging data corresponding to the imaging result is written into the built-in memory 20.

The face detection unit 101 of the CPU 54 detects (searches) a face area in the captured image based on the digital imaging data read from the built-in memory 20 (step ST32). The face area detection process is performed, for example, by the same method as in step ST14 described above.
Then, the face recognition unit 102 of the CPU 54 determines whether or not the face image data in the face area detected in step ST <b> 32 matches any registered face image data stored in the face registration memory 56. Face recognition processing is performed (step ST33). Specifically, the face recognition unit 102, based on the feature points of the face detected by the face detection unit 101, the position of the feature point of the face image of the captured person, the size of each face part, and the relative of each feature point Calculate the distance. Next, the face recognition unit 102 calculates the similarity between the face of a predetermined registered person and the face of the imaged person by comparing the above calculation result with the registered face image data. Then, the face recognition unit 102 determines that the imaged person matches a predetermined registered person when the similarity is greater than the threshold value. The algorithm based on the similarity is not limited to the above-described algorithm, and various algorithms can be used.
The function of the face recognition unit 102 is realized by the CPU 54 executing a program stored in a ROM (not shown).

  If the control unit 100 of the CPU 54 determines that they match in step ST33 ("Y" in step ST34), it is associated with the registered face image data used for the match determination in the face registration memory 56. Metadata such as stored name data and / or date of birth data is read (step ST35).

  The control unit 100 reads the current date data from the calendar function clock 52. And the control part 100 calculates the age data of the person from the date-of-birth date data read in step ST35 and the date data (imaging date data) at that time acquired from the calendar function clock 52. (Step ST36).

  In addition, when the control unit 100 determines that the school age is based on the calculated age data, the control unit 100 generates grade data such as the third grade of elementary school, for example. In addition, the control unit 100 determines whether the date data acquired from the calendar function-equipped clock 52 indicates the birthday of the person. These determinations are made using a calculation function and a determination function provided in the CPU 54.

Based on the instruction from the control unit 100, the sound / image processing unit 24 reads the name data which is the metadata read from the face registration memory 56 as described above, the age data generated as described above, the grade data, the birthday At least one of the data indicating that the image is a region other than the face region of the screen being imaged and is superimposed on the region near the face region to generate new image data to be displayed (step ST37). .
The new image data is output to at least one of the liquid crystal monitor 36, the speaker 38, the operation unit 40, and the output terminal 42 via the input / output I / F 34 shown in FIG.
Thereby, for example, as shown in FIG. 5, in the liquid crystal monitor 36, in the vicinity of each face image area, the name of the person corresponding to the face image, the age, and information indicating whether it is a birthday or not Is displayed.

The control unit 100 determines whether or not a predetermined period (for example, 1 minute) has elapsed since the start of imaging and the face recognition processing was completed, and when determining that it has elapsed ("Y" in step ST38) ) Terminates the process.
On the other hand, when determining that the time has not elapsed in step ST38 ("N" in step ST38), the control unit 100 determines whether the face recognition processing has been completed for all detected face regions (step ST39). If it is determined that the process has not been completed, the process returns to step ST33 to start face recognition processing for an unprocessed face area.
In addition, when the control unit 100 determines that the processing has been completed in step ST39 (“N” in step ST38), for example, the superimposed image data generated in step ST37 is written in the HDD 28 or the card type recording medium 32 ( Step ST40).
Writing to the HDD 28 is performed via the HDD I / F 26. Further, writing to the card type recording medium 32 is performed via the card I / F 30. The card type recording medium 32 is configured by a known semiconductor memory or the like.

At this time, the control unit 100 outputs a display for confirming to the user whether or not the superimposed image data can be directly written to the HDD 28 or the card type recording medium 32, and the user may superimpose the data. Writing may be started when the selection is made. At this time, the superimposed image data may be displayed on the liquid crystal monitor 36 to allow the user to confirm.
When the user selects writing without superimposing, the control unit 100 associates the imaging data before superimposition with metadata such as name data and age data individually and associates with the HDD 28 or the card-type recording medium. Write to 32.
Note that the control unit 100 may perform writing in any format without selection from the user.

The audio / image processing unit 24 reproduces the image data read from the HDD 28 or the card type recording medium 32 at the time of reproduction according to an instruction from the control unit 100, but superimposes metadata associated with the image data on the reproduction screen. Whether or not to do so may be selected according to a user instruction.
Note that the control unit 100 may record birth date data and date data at the time of imaging as metadata instead of age data. In this case, at the time of reproduction, the control unit 100 may extract birth date data and date data at the time of imaging from the metadata, and calculate age data of each person at the time of imaging. .

  As described above, in the face recognition imaging process, the imaging apparatus 1 can automatically detect a registered person's face image from the imaging data and display the metadata of the person. Therefore, it is possible to immediately know the attributes such as the age and name of the person displayed in the reproduced image to the user based on the displayed metadata. In addition, when the reproduction date is a birthday, the imaging device 1 can perform a playful and interesting reproduction by displaying a decoration image indicating the birthday on the person's face image. . .

<Face registration update process>
FIG. 6 is a flowchart for explaining face registration update processing in the imaging apparatus 1 shown in FIG.
When the control unit 100 of the CPU 54 detects that the power of the imaging apparatus 1 is turned on (“Y” in step ST51), the control unit 100 acquires date data at that time from the calendar function clock 52 (step ST52).
The control unit 100 reads metadata such as birth date data associated with each registered face image data registered in the face registration memory 56 and registered date data of the registered face image data.
And the control part 100 calculates the age data of the person corresponding to the said registration face image data from the read birth date data and the present date data read from the clock 52 with a calendar function (step). ST53).

  The control unit 100 calculates elapsed date data after registering the registered face image data from the read registration date data and the current date data acquired in step ST52 (step ST54).

The control unit 100 acquires a threshold value corresponding to the age data calculated in step ST53 from the face registration memory 56.
A plurality of threshold values are preliminarily defined at the time of shipment corresponding to the age. For example, 2 months up to 6 months after birth, 6 months up to 2 years old, 12 months up to 12 years old, 24 up to 18 years old. If it is more than a month, it is a numerical value such as 999 months (substantially no update).
The control unit 100 compares the elapsed day data calculated in step ST54 with the acquired threshold value (step ST55). If the controller 100 determines that the number of days indicated by the elapsed day data is large for at least one registered face image data (if face registration update is required: “Y” in step ST56), the control unit 100 proceeds to step ST57. move on. On the other hand, when it is determined that the number of days indicated by the elapsed days data is not large for all the registered face image data (when face registration update is not required: “N” in step ST56), the control unit 100 performs face registration. The update process ends.

When face registration update processing is required (“Y” in step ST56), control unit 100 displays an instruction “capture the person to be updated” on liquid crystal monitor 36 (step ST57). Then, the user starts imaging, and digital imaging data corresponding to the imaging result is written in the built-in memory 20 (step ST58).
Then, the face detection unit 101 detects a face area in the captured image based on the digital image data read from the built-in memory 20 (step ST59). The detection of the face area is performed by the process described in step ST14 shown in FIG.

When the face detection unit 101 detects a face area in the image being captured (“Y” in step ST59), the audio / image processing unit 24 superimposes a mark image in the vicinity of the detected face area. Data is generated and a screen corresponding to the data is displayed on the liquid crystal monitor 36. As a specific example, a screen 220 shown in FIG. 7 is displayed on the liquid crystal monitor 36.
When the face detection unit 101 detects a plurality of face regions, the sound / image processing unit 24 displays a plurality of mark images 220 and 222, and the color of the mark image 222 near one unselected face region. Is changed from others and blinks (step ST60). At that time, the name of the person may be displayed near the face area of the blinking mark image 222.

The voice / image processing unit 24 updates the face image data generated based on the feature points of the face images of the plurality of face regions (the elapsed day data after registration is larger than the threshold value). The mark image of the face area of the face having the highest similarity with the registered face image data stored in the registration memory 56 is first blinked.
For example, as shown in FIG. 7, the audio / image processing unit 24 displays mark images 220 and 222 in the vicinity of the face image of the person to be updated in the liquid crystal monitor 36, respectively. Flash on.
When the person to whom the blinking mark image is added is the person who wants to update the registration, the user selects “registration update” with the input button of the operation unit 40 (“registration update” in step ST61). If the person to be registered / updated is another person, the user selects “select another person” (“select another person” in step ST61). When “select another person” is selected, the sound / image processing unit 24 changes the color of the mark near one face area that has not been selected and blinks (step ST62). This one face area is a face area having the next highest similarity.

When “registration update” is selected with the input button of the operation unit 40 (“registration update” in step ST61), the face recognition unit 102 generates new face image data generated based on the detected face feature points. Thus, the registered face image data already stored in the face registration memory 56 is replaced and recorded (step ST63). In addition, the control unit 100 replaces the registered date data recorded in the face registration memory 56 with the current date data acquired from the calendar function clock 52 and records the data.
The control unit 100 determines whether or not face registration update has been completed for all registered face image data determined to require face registration update in step ST56, and determines that it has ended (in step ST64). “Y”) terminates the process, otherwise, the color of the mark image near one unselected face area is changed from the other and blinked (step ST60).

In addition to the above, the face registration update process may be performed as shown below.
The face registration update process described above was started with the power-on of the imaging apparatus 1 as a trigger as shown in FIG. In the example shown below, face registration update processing is performed during face recognition imaging processing.
That is, the face detection unit 101 detects a face area from digital image data obtained during imaging. Then, the similarity between the face recognition unit 102 and the face image data generated by executing the face recognition process on the detected face area among the registered face image data registered in the face registration memory 56 is similar. The highest one is specified, and the birth date data and registration date data of the person indicated by the metadata corresponding to the registered face image data are read out.

And the control part 100 acquires the date data at the time from the timepiece 52 with a calendar function. Next, the control unit 100 calculates age data from the acquired date data and the birth date data of the person corresponding to the detected face area, and the date data and the registration date data are calculated. Elapsed days data since registration (or registration renewal) is calculated.
Then, the control unit 100 determines a predetermined index value based on the calculated similarity, age data, and elapsed day data by a predetermined calculation process. For the determination, an arithmetic expression may be used, or table data prepared in advance may be used. Here, for example, the index value is determined to be larger as the similarity is lower, larger as the age is lower, and larger as the elapsed days are longer.
Thereby, the face registration update process can be performed more easily as the change of the face is larger, the age is lower, and the elapsed days are longer.

When the determined index value exceeds a predetermined threshold, the control unit 100 generates the face image data generated by the face recognition process, the date data when it is acquired, Metadata (name data or the like) corresponding to registered face image data having a high degree of similarity is recorded in a predetermined area of the face registration memory 56 as an update target.
When the determined index value exceeds the threshold value, the control unit 100 displays “Registration update candidate available. Do you want to enter the update mode” on the liquid crystal monitor 36 at the timing of the end of imaging or the like. When the user selects “Yes” with the input button of the operation unit 40, the audio / image processing unit 24 reproduces and outputs the image data recorded in the face registration memory 56, and further, the audio / image processing unit 24. A mark image and name data corresponding to the mark image are displayed in the vicinity of the face area corresponding to the registered face image data to be updated. Then, as illustrated in FIG. 7, the control unit 100 displays a screen 220 indicating “Do you want to update?” On the liquid crystal monitor 36.

  When the user selects “Yes” with the input button of the operation unit 40, the control unit 100 updates the registered face image data of the registered person recorded in the face registration memory 56. That is, the registered face image data of the registered person is replaced with the face image data stored in the predetermined area, and the face image data in which the registered date data is stored in the predetermined area is acquired. Replace with date data.

  According to the imaging device 1, as described above, the necessity of updating the registered face image data is automatically determined, and when the update is necessary, an operation screen that prompts the update is displayed, so that it grows up like a child or the like. For a person whose facial features change along with this, registered face image data suitable for the face authentication process can be used, and the possibility of a recognition error occurring in the face recognition process can be reduced.

The present invention is not limited to the embodiment described above.
That is, those skilled in the art may make various modifications, combinations, subcombinations, and alternatives regarding the components of the above-described embodiments within the technical scope of the present invention or an equivalent scope thereof.

  In the above-described embodiment, as the metadata of the present invention, the name data of the person corresponding to the registered face image data, the date of birth data, the date of registration data, and the like are exemplified. You may use what shows the attribute of a person or an image, such as hobbies and sex.

  Further, in the above-described embodiment, in the face recognition imaging process, the case where the metadata obtained by the face authentication process is superimposed on the vicinity of the detected face area, that is, the case where the metadata is associated with the face image data is exemplified. The invention may associate metadata with imaging data.

  In the above-described embodiment, a digital video camera is exemplified as the imaging device 1, but the imaging device 1 may be a digital still camera, for example. Further, the image processing apparatus of the present invention may be realized as a function for a computer to execute a program in addition to the imaging apparatus.

  In the above-described embodiment, the imaging apparatus 1 that can execute all the processes of the face registration process, the face recognition imaging process, and the face registration update process is illustrated. However, at least one of the face recognition imaging process and the face registration update process is exemplified. Processing may not be performed.

  In the above-described embodiment, the case where metadata such as a name is displayed in association with the face image recognized in the imaging data is exemplified. However, the metadata is the imaging data or a predetermined face image in the imaging data. It may be used to search for included frames or to manage such imaging data and frames for a predetermined purpose.

  Further, in the above-described embodiment, the case where the calendar function clock 52 generates year / month / day data is exemplified, but year data or year / month data is generated, and processing is performed in units of year or month. Also good.

  In the above-described embodiment, the case where the CPU 54 executes the functions of the control unit 100, the face detection unit 101, and the face recognition unit 102 has been exemplified. However, at least one function may be realized by dedicated hardware. Further, the function of the sound / image processing unit 24 may be realized by the CPU 54 executing a program.

  The present invention is applicable to a system that processes a face image included in imaging data.

1 is a schematic diagram illustrating a configuration example of an imaging apparatus (digital video camera) according to an embodiment of the present invention. 3 is a flowchart for explaining face registration setting processing by the imaging apparatus shown in FIG. 1. It is a figure for demonstrating the screen used for the face registration process of the imaging device shown in FIG. 3 is a flowchart for explaining face recognition imaging processing in the imaging apparatus 1 shown in FIG. 1. It is a figure for demonstrating the screen used for the face recognition imaging process of the imaging device shown in FIG. 3 is a flowchart for explaining face registration update processing in the imaging apparatus 1 shown in FIG. 1. It is a figure for demonstrating the screen used for the face registration update process of the imaging device shown in FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Imaging device 10 ... Imaging optical system 12 ... Imaging device 14 ... Microphone 16 ... Analog signal processing part 18 ... A / D converter 20 ... Built-in memory 22 ... Data bus 24 ... Audio | voice / image processing part 26 ... HDD I / F
28 ... HDD
30 ... Card I / F
32 ... Card type recording medium 34 ... Input / output I / F
36 ... Liquid crystal monitor 38 ... Speaker 40 ... Operation unit 42 ... Output terminal 50 ... Built-in battery 52 ... Clock 54 with calendar function ... CPU
56 ... Face registration memory 58 ... External battery 100 ... Control unit 101 ... Face detection unit 102 ... Face recognition unit

Claims (4)

Face detection means for detecting face image data in the image data;
Storage means for storing the registered face image data of a person registered in advance and attribute data indicating at least the attribute of the person in association with each other;
Based on the face image data detected by the face detection means and the registered face image data recorded in the storage means, the face image data detected by the face detection means corresponds to a person of the registered image data. Face recognition means for determining whether or not an object,
The attribute data of the registered person read from the storage means when the face recognition means determines that the face image data detected by the face detection means corresponds to the person of the registered image data And an image processing apparatus comprising: a control unit that performs a process of associating the imaging data or the face image data detected by the face detection unit. The control means, when the face recognition means determines that the face image data detected by the face detection means corresponds to a person in the registered image data, the registered read out from the storage means The process of calculating age data of the person based on the date of birth data of the person and the imaging year data indicating the year in which the imaging data was generated, and the elapsed time from the registration year of the registered face image data Processing to calculate year data and to prompt the update of the registered face image data when the elapsed year indicated by the calculated elapsed year data exceeds a predetermined threshold corresponding to the calculated age data The image processing apparatus according to claim 1, wherein: The control means, when the face recognition means determines that the face image data detected by the face detection means corresponds to a person in the registered image data, the registered read out from the storage means The age data of the person is calculated based on the date of birth of the person and the imaging year data indicating the year in which the imaging data was generated, and the calculated age data is used as the imaging data or the detected face image. The image processing apparatus according to claim 1, wherein a process of displaying in association with data is performed. A face detection step of detecting face image data in the imaging data;
Based on the face image data detected in the face detection step and registered face image data registered in advance, the face image data detected in the face detection step corresponds to a person in the registered image data. A face recognition process for determining whether or not
When the face recognition step determines that the face image data detected in the face detection step corresponds to a person in the registered image data, the attribute data of the registered person and the imaging data or And a control step of performing a process of associating the face image data detected in the face detection step.