JP4360369B2 - Electronic camera - Google Patents

Description

  The present invention relates to an electronic camera having a human face recognition function.

Conventionally, a technique has been disclosed in which feature points of a person's face are extracted from a photographed image of the person's face, and face recognition is performed based on the degree of coincidence between face recognition data related to the person to be recognized and the above feature points. (See Patent Document 1). Such face recognition technology requires that face recognition data be registered in advance in an electronic camera or the like. In particular, in order to improve the accuracy of face recognition, more face recognition data is required, so it is necessary to generate face recognition data from a plurality of frames of captured images.
JP-A-6-259534

By the way, when a large number of faces are detected in the shooting screen or when the total number of face recognition data used for the face recognition determination is large, the amount of calculation of the face recognition processing becomes very large. In the above case, since the calculation time of the face recognition process becomes long, there is still room for improvement in that the user's comfortable shooting is hindered and the possibility of missing a shutter chance increases.
The present invention is to solve the above-described problems of the prior art, and an object of the present invention is to provide an electronic camera that can appropriately adjust the calculation amount of face recognition processing according to the situation.

An electronic camera according to a first invention includes an image sensor, a face detection unit, a recording unit, a control unit, and a face recognition unit. The imaging device generates an image signal by photoelectrically converting a subject image obtained by the photographing optical system. The face detection unit detects a face area of the shooting screen based on the image signal, and extracts a feature point of the photographed person's face from the face area. The recording unit records face recognition data generated in advance based on the facial feature points of the registered person. The control unit has an evaluation value corresponding to the number of face recognition data based on the number of face recognition data included in a face recognition data group in which a plurality of face recognition data related to a registered person as a face recognition target is collected. The face recognition data for determination is selected from the face recognition data group so as to be lower than the threshold value defined by the processing capability . The face recognition unit determines whether the photographed person is a registered person for face recognition based on the feature points extracted by the face detection unit and the determination face recognition data. In addition, when there are a plurality of registered persons to be face-recognized, the control unit weights and selects determination face recognition data based on the priority order set for each registered person.

An electronic camera according to a second invention includes an imaging device, a face detection unit, a recording unit, a control unit, and a face recognition unit. The imaging device generates an image signal by photoelectrically converting a subject image obtained by the photographing optical system. The face detection unit detects a face area of the shooting screen based on the image signal, and extracts a feature point of the photographed person's face from the face area. The recording unit records face recognition data generated in advance based on the facial feature points of the registered person. The control unit has an evaluation value corresponding to the number of face recognition data based on the number of face recognition data included in the face recognition data group in which a plurality of face recognition data related to the registered person of the face recognition target is collected. The face recognition data for determination is selected from the face recognition data group so as to be lower than the threshold defined by the processing capability. The face recognition unit determines whether the photographed person is a registered person for face recognition based on the feature points extracted by the face detection unit and the determination face recognition data. In addition, when there are a plurality of registered persons to be face-recognized, the control unit preferentially selects the registered person with a large number of recorded face recognition data and selects the face recognition data for determination.
According to a third aspect, in the first or second aspect, the evaluation value is a value calculated based on the number of face recognition data and the number of face areas.

In a fourth aspect based on any one of the first to third aspects, the control unit prioritizes face recognition data having a new acquisition time when a plurality of face recognition data are recorded for the same registered person. And selecting judgment face recognition data.

According to a fifth invention, in any one of the first to fourth inventions, when the plurality of face recognition data having different shooting conditions at the time of acquisition are recorded for the same registered person, the control unit The determination face recognition data is selected by giving priority to face recognition data whose shooting conditions approximate the current shooting conditions.
According to a sixth aspect of the present invention, in any one of the first to fifth aspects, the control unit corresponds to the face front when a plurality of face recognition data having different face orientations are recorded for the same registered person. The determination face recognition data is selected with priority given to the face recognition data.

In a seventh aspect based on any one of the first to sixth aspects, the recording unit further records photographing number data indicating the number of photographings for each registered person. Then, the control unit selects the face recognition data for determination with priority on the registered person with a large number of photographing times based on the photographing number data.
Here, a representation of the configuration related to the present invention converted into a method, a recording medium, a computer program, and the like is also effective as a specific aspect of the present invention.

  In the electronic camera of the present invention, the amount of calculation of face recognition processing can be adjusted appropriately according to the situation.

FIG. 1 is a block diagram showing the configuration of the electronic camera of this embodiment.
The electronic camera includes a photographing optical system 11, a first lens driving unit 12, a swing sensor unit 13, an image blur correction lens 14, a second lens driving unit 15, an image sensor 16, and an analog signal processing unit 17. The A / D converter 18, the built-in memory 19, the image processor 20, the card I / F 21, the input / output I / F 22, the operation member 23 and the liquid crystal monitor 24, the flash light emitter 25, and the CPU 26. And a face registration memory 27 and a data bus 28. The internal memory 19, the image processing unit 20, the card I / F 21, the input / output I / F 22, and the CPU 26 are connected via a data bus 28.

The photographing optical system 11 includes a plurality of lens groups including a zoom lens 11a and a focusing lens 11b for adjusting a focus position. Each lens constituting the photographing optical system 11 is driven and adjusted in the optical axis direction by the first lens driving unit 12.
The swing sensor unit 13 includes a vertical angular velocity sensor that detects the pitch of the electronic camera and a horizontal angular velocity sensor that detects the roll of the electronic camera. The swing sensor unit 13 outputs camera swing data indicating the swing of the electronic camera to the CPU 26. This camera swing data is used to calculate the amount of movement of the image blur correction lens 14.

  The image blur correction lens 14 is configured to be swingable in the direction perpendicular to the optical axis. The second lens driving unit 15 includes a vertical swing unit that swings the image blur correction lens 14 in the vertical direction and a horizontal swing unit that swings the image blur correction lens 14 in the horizontal direction. The second lens driving unit 15 swings the image blur correction lens 14 based on the movement amount of the image blur correction lens 14 calculated by the CPU 26 and executes blur correction.

  The image sensor 16 photoelectrically converts the light beam that has passed through the photographing optical system 11 to generate an analog image signal of the subject image. The image sensor 16 also exposes the subject at predetermined intervals even when the shutter is not released (during shooting standby), and outputs an analog image signal (through image signal) by thinning readout. The through image signal is used for AF calculation, AE calculation, face detection operation, and finder image generation, which will be described later.

  The analog signal processing unit 17 includes a CDS circuit that performs correlated double sampling, a gain circuit that amplifies the output of the analog image signal, a clamp circuit that clamps the waveform of the input signal to a certain voltage level, and the like. The A / D converter 18 converts the analog image signal of the image sensor 16 into a digital image signal. The built-in memory 19 temporarily stores data in the pre-process and post-process of image processing in the image processing unit 20. The built-in memory 19 can be provided with a recording area in which captured image data described later can be recorded.

  The image processing unit 20 performs predetermined image processing on the digital image signal. The image processing unit 20 generates captured image data based on the image signal at the time of release. Further, the image processing unit 20 sequentially generates through image data based on the through image signal. The resolution of the through image data is set lower than the resolution of the captured image data. The through image data is used for face recognition processing by the CPU 26 and also used as a display image (preview image) on the liquid crystal monitor 24. Furthermore, the image processing unit 20 also performs a process of compressing captured image data. Furthermore, the image processing unit 20 can also perform an electronic zoom process that cuts out a part of the captured image and converts the resolution.

In addition, the image processing unit 20 generates video data for various displays (such as a frame display during face recognition) according to instructions from the CPU 26, superimposes the video data on the through image data, and outputs the video data to the liquid crystal monitor 24. As a result, the various displays described above are combined with the preview image.
The card I / F 21 is formed with a connector for connecting the recording medium 29. The recording medium 29 is composed of a known semiconductor memory or the like. The card I / F 21 controls writing / reading of photographed image data with respect to the recording medium 29.

  An operation member 23 and a liquid crystal monitor 24 are connected to the input / output I / F 22. The operation member 23 includes an input button, a release button, and the like. The input button of the operation member 23 accepts, for example, a switching input of a photographing mode (normal photographing mode, reproduction mode, face registration mode, face recognition photographing mode, etc.) of the electronic camera. The release button of the operation member 23 accepts an AF operation instruction input by half pressing and a release input by full pressing.

  On the other hand, the liquid crystal monitor 24 is mainly disposed on the back surface portion of the electronic camera. The liquid crystal monitor 24 displays a playback image of the captured image data, a setting screen for changing various settings of the electronic camera, and the like. The setting screen includes an editing screen for editing face recognition data described later. In addition, the preview image can be displayed as a moving image on the liquid crystal monitor 24 in the shooting mode. Therefore, the user can perform framing of the subject while confirming the preview image on the liquid crystal monitor 24.

  Furthermore, an external connection terminal is formed in the input / output I / F 22. The input / output I / F 22 controls data transmission / reception with a computer or the like connected via an external connection terminal in accordance with a serial communication standard such as USB (Universal Serial Bus) or IEEE1394. Further, it is possible to connect a communication unit 30 which is an expansion device for performing wireless communication with the outside to the external connection terminal of the input / output I / F 22.

The flash light emitting unit 25 includes a xenon arc tube, a main capacitor for storing light emission energy, a reflector and a lens member for efficiently irradiating a subject with flash light, a light emission control circuit for adjusting light emission timing and light quantity, and the like. .
The CPU 26 controls the operation of each part of the electronic camera according to a sequence program stored in a ROM (not shown). For example, the CPU 26 executes a known contrast AF (autofocus) calculation or a known AE (automatic exposure) calculation based on the through image signal. Further, the CPU 26 calculates the movement amount of the image blur correction lens 14 based on the camera swing data.

  Further, the CPU 26 functions as a face detection unit 31 and a face recognition unit 32. The face detection unit 31 extracts feature points from through image data or captured image data, and detects a face area, a face size, and the like of the subject. For example, the face detection unit 31 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, and lip end points, face contour points, head vertices, and chin lower end points.

In the “face registration mode”, the face recognition unit 32 generates face recognition data based on the detected feature points of the face. For example, the face recognizing unit 32 generates face recognition data of a registered person from the position of the detected feature point of the face, the size of the face part obtained from the feature point, the relative distance of each feature point, and the like.
Further, in the “face recognition shooting mode”, the face recognition unit 32 performs face recognition processing for determining whether or not a person's face in the shooting screen is a face of a registered person in the face recognition data. Specifically, the face recognition unit 32 first calculates the position of the feature point of the photographed person's face, the size of each face part, the relative distance of each feature point, and the like based on the detected feature points of the face. Next, the face recognition unit 32 compares the above calculation result with the face recognition data, and obtains the similarity between the face of the predetermined registered person and the face of the photographed person. Then, the face recognizing unit 32 determines that the photographed person matches a predetermined registered person when the similarity is greater than the threshold value.

Note that the CPU 26 can execute a change in shooting conditions or image processing settings, a release instruction to the image sensor 16, and the like based on the result of the face recognition process.
In the face registration memory 27, face recognition data generated by the face recognition unit 32 of the CPU 26 is recorded. In the face registration memory 27, a group folder is generated for each registered person. In the group folder, face recognition data of each registered person is recorded. That is, the face recognition data in the face registration memory 27 is grouped for each registered person by the group folder. For example, in the group folder, a plurality of face recognition data with different face orientations, presence / absence of glasses, photographing conditions, and the like can be recorded for the same registered person. Then, by designating the registered person, the CPU 26 can collectively read the face recognition data in the group folder from the face registration memory 27.

The face registration memory 27 stores data of “name of registered person” and “processing setting for face recognition”. The “registered person name” and “processing setting at the time of face recognition” are associated with the group folder of each registered person.
Here, in the “processing setting at the time of face recognition”, settings of various controls of the electronic camera that are applied when a registered person is recognized are recorded. For example, “processing setting for face recognition” includes (1) setting of shooting conditions at the time of shooting, (2) setting of image processing of shot image data, (3) setting of a recording destination of shot image data, (4 And (5) setting for transmission of photographed image data, and the like.

The setting of the shooting condition in (1) includes an exposure correction value when shooting each registered person, setting of the depth of field (shallow, standard selection), and the like. It is also possible to set the CPU 26 to automatically take a picture when a registered person is detected. The image processing settings in (2) above include settings for presence / absence of soft focus processing of captured image data, settings for suppressing edge enhancement, and the like.
In the setting of the recording destination in (3) above, the directory path of the recording folder in the recording medium 29 that is the recording destination of the photographed image data can be designated for each registered person. That is, the CPU 26 can record the photographed image data in a different recording folder for each registered person by referring to the data in the face registration memory 27.

In the print designation setting in (4) above, for example, it is possible to designate, for each registered person, whether or not to print photographed image data according to a standard such as DPOF (Digital Print Order Format). In the transmission setting in (5) above, it is possible to designate whether or not the captured image data needs to be transmitted when the registered person is photographed, and address information (e-mail address) of the transmission destination.
The face registration memory 27 stores data of “face registration image”, “face recognition data registration time”, and “shooting conditions at the time of face recognition data acquisition”. The above-mentioned “face registration image”, “face recognition data registration time”, and “imaging conditions when acquiring face recognition data” are associated with individual face recognition data in the group folder.

  The above “face registration image” is photographed image data at the time of face recognition data acquisition, and is resized to a predetermined pixel size. Then, the “face registration image” is displayed on the liquid crystal monitor 24 in the above editing screen. That is, the “face registration image” functions as an index image that enhances the identifiability and searchability of the registered person or face recognition data. For example, the “face registration image” allows the user to check on the editing screen which face recognition data has been generated from what kind of image (face front, landscape orientation, etc.).

  “Registration time of face recognition data” is data of photographing date and time of face registration image data that is the basis of the face recognition data. Further, “shooting conditions when acquiring face recognition data” is data of shooting conditions of face registration image data. The above photographing conditions include, for example, focal length data of the photographing optical system 11, white balance and exposure conditions at the time of photographing, presence / absence of flash emission at the time of photographing, and the like. Note that the data of “registration time of face recognition data” and “shooting conditions when acquiring face recognition data” are also used for determination when the CPU 26 narrows down the face recognition data used in the face recognition processing.

The face registration memory 27 also records shooting count data indicating the shooting count for each registered person. The number of times of photographing of the number of times of photographing data is incremented by the CPU 26 according to the result of face recognition in the face recognition photographing mode. Note that the number of times of shooting may be manually input by the user on an editing screen or the like.
The operation of the “face recognition shooting mode” of the electronic camera of this embodiment will be described below.

FIG. 2 is a flowchart showing the operation of the “face recognition photographing mode” of the present embodiment. This “face recognition shooting mode” is a shooting mode in which, when a designated registered person exists on the shooting screen, the electronic camera performs predetermined control to acquire shot image data.
Step 101: First, the user sets the photographing mode of the electronic camera to the “face recognition photographing mode” using the operation member 23.

  For example, in a state where the electronic camera is set to “normal shooting mode” or “portrait shooting mode”, the user performs an operation for calling a mode selection screen. As a result, the CPU 26 displays a mode selection screen (see FIG. 3) on the liquid crystal monitor 24. Then, the user operates the selection frame on the mode selection screen with the operation member 23 and instructs the CPU 26 to activate the “face recognition photographing mode”. Thereby, the CPU 26 starts a series of operations in the “face recognition photographing mode”.

Step 102: The CPU 26 displays a registration person selection screen (see FIG. 4) on the liquid crystal monitor 24. Then, the CPU 26 receives a designation input of a registered person to be face recognized (a registered person who performs face recognition at the time of shooting). It is also possible for the user to designate and input a registered person for face recognition into the CPU 26 in advance before photographing.
Here, in the “registered person selection screen”, the CPU 26 can accept designation of one or more registered persons. For example, the CPU 26 can accept all designations of registered persons in the face registration memory 27. In addition, the group folder may be linked in advance with a predetermined category (for example, family, circle, etc.), and the user may input a batch designation of registered persons linked to the CPU 26 by designating and inputting the category on the selection screen. it can.

Step 103: The CPU 26 drives the image sensor 16 to acquire a through image signal. The image sensor 16 outputs a through image signal by thinning readout at predetermined intervals. The image processing unit 20 generates through image data based on the through image signal. A preview image is displayed as a moving image on the liquid crystal monitor 24 based on the through image data.
Step 104: On the other hand, the face detection unit 31 of the CPU 26 performs a known face detection process on the through image data to detect a face area in the shooting screen.

Step 105: The CPU 26 determines whether or not a face is detected from the shooting screen in S104. If a face is detected (YES side), the process proceeds to S106. On the other hand, when the face is not detected (NO side), the process proceeds to S114.
Step 106: The CPU 26 performs AF calculation with reference to the detected position of the face area. If a plurality of face areas are detected, the CPU 26 prioritizes the face area located at the center of the shooting screen or the face area located closest to the shooting screen, and executes the AF calculation.

Step 107: The CPU 26 counts the number of all face recognition data ( the number of face recognition data included in the face recognition data group ) related to the registered person as the face recognition target. Specifically, in the face registration memory 27, face recognition data recorded in the group folder of the registered person to be face recognition is counted.
Step 108: The CPU 26 multiplies the number of face areas detected in the photographing screen (S104) and the number of face recognition data included in the face recognition data group (S107), and performs the maximum calculation of the current face recognition process. An evaluation value indicating the quantity is obtained. The above evaluation value corresponds to the amount of calculation when performing face recognition processing by applying all the face recognition data included in the face recognition data group to all the face regions detected this time.

Step 109: The CPU 26 obtains a difference value obtained by subtracting the evaluation value (S108) from a predetermined threshold value. The threshold value is set based on the amount of face recognition processing that can be executed by the face recognition unit 32 within a predetermined time limit. Therefore, the specific numerical value of the threshold depends on the calculation capability of the face recognition unit 32.
Step 110: The CPU 26 determines whether or not the difference value (S109) is negative. When the difference value is negative (YES side), the process proceeds to S111. On the other hand, when the difference value is 0 or more (NO side), the process proceeds to S112. In the case of NO in S110, the face recognition unit 32 uses all face recognition data in the face recognition data group as determination face recognition data.

Step 111: In this case, the evaluation value (S108) exceeds the threshold, and if all the face recognition data of the face recognition data group is applied, there is a high possibility that the face recognition process will not be completed within the time limit. Therefore, the CPU 26 narrows down determination face recognition data to be applied to the current face recognition process from the face recognition data group.
Specifically, first, the CPU 26 calculates the upper limit value of the determination face recognition data based on the difference value (S109). Next, the CPU 26 selects the face recognition data for determination from the face recognition data group so as to be less than the upper limit value by the following algorithms (1) and (2).

(1) When there are a plurality of registered persons for face recognition, the CPU 26 sets the distribution number of the face recognition data for determination for each registered person based on the above upper limit value. Basically, the CPU 26 sets the same number of distributions for the registered persons for face recognition.
However, the CPU 26 may weight the number of distributions based on the priority order set for each registered person so that the number of face recognition data for determination increases in proportion to the priority level. In this case, a registered person with a higher priority is subjected to face recognition processing based on more face recognition data, so that a registered person with a higher priority can be recognized with high accuracy. For example, the CPU 26 can set a higher priority for a registered person with a higher number of shootings based on the shooting number data in the face registration memory 27.

Further, the CPU 26 may compare the number of face recognition data of each group folder, and may set so that the number of distribution is higher for registered persons whose face recognition data number is a predetermined value or more. This is because a registered person with a larger number of registered face recognition data is more likely to be a registered person who wants to preferentially shoot for the user.
(2) The CPU 26 selects the face recognition data for determination from the face recognition data of the group folder for each registered person to be face recognized within the range of the distribution number set in (1) above. Here, when the face recognition data of the group folder is equal to or less than the distribution number, the CPU 26 selects all the face recognition data in the group folder as determination face recognition data. In this case, the CPU 26 may redistribute the surplus distribution number to other registered persons.

On the other hand, when the face recognition data of the group folder exceeds the distribution number, the CPU 26 selects the face recognition data for determination based on any of the following criteria (or a combination thereof).
First, the CPU 26 refers to the data of “registration time of face recognition data” in the group folder, and selects the determination face recognition data in order of the new registration time. This is because it is considered that the face recognition data with the older registration time is less reliable in this shooting.

  Second, the CPU 26 refers to the data of “shooting condition when acquiring face recognition data” in the group folder. Then, the CPU 26 selects the face recognition data for determination by giving priority to the face recognition data whose shooting condition at the time of acquisition approximates the current shooting condition from among the plurality of face recognition data. This is because the face recognition data acquired under the conditions close to the current photographing condition contributes to the improvement of face recognition accuracy.

  For example, a table for scoring the amount of change in each parameter of the shooting conditions is created in advance, and the CPU 26 obtains the score of the shooting conditions of the face recognition data based on the current shooting conditions. Then, the CPU 26 compares the above scores and selects determination face recognition data. As a result, when flashing is performed at the time of shooting, the CPU 26 preferentially selects face recognition data with flash emission and selects face recognition data for determination. On the other hand, when the flash is not emitted at the time of shooting, the CPU 26 selects the face recognition data for determination with priority on the face recognition data without the flash emission. Similarly, the CPU 26 preferentially selects face recognition data for which exposure conditions and white balance settings are closer, and selects determination face recognition data.

  Thirdly, the CPU 26 selects the face recognition data for determination with priority given to the face recognition data corresponding to the front of the face among the face recognition data of the group folder. This is because usually the face is often photographed from the front. Here, the detection of the face orientation of the face recognition data is performed as follows. For example, identification information regarding the orientation of the face may be recorded in advance in the face recognition data, and the CPU 26 may determine the orientation of the face with reference to the identification information. Further, the CPU 26 may infer the orientation of the face from the extracted state of feature points such as eyes, nose and mouth in the face recognition data.

  Step 112: The face recognition unit 32 of the CPU 26 executes face recognition processing using the determination face recognition data for the detected face area (S104). Then, the face recognition unit 32 determines whether or not the recognition target face (S102) is included. If the face to be recognized is included (YES side), the process proceeds to S113. On the other hand, if the face to be recognized is not included (NO side), the process proceeds to S115.

Step 113: The CPU 26 changes the setting of the photographing condition of the electronic camera and the setting of the image processing based on the data of “processing setting at the time of face recognition” corresponding to the detected registered person. Thereafter, the process proceeds to S115. When a plurality of faces to be recognized are detected in the shooting screen, the CPU 26 determines the priority order of the registered person by a predetermined algorithm. Then, the CPU 26 changes various settings of the electronic camera based on the “processing setting for face recognition” corresponding to the registered person with the highest priority. If automatic shooting of a registered person is set in the “processing setting for face recognition”, the CPU 26 proceeds to S116 and automatically images the registered person.

Step 114: On the other hand, in this case, the CPU 26 selects the AF area by the algorithm at the time of normal shooting and executes the AF calculation in response to the half-press operation of the release button by the user.
Step 115: The CPU 26 determines whether or not the release button has been fully pressed. When the release button is fully pressed (YES side), the process proceeds to S116. On the other hand, when there is no input to the release button (NO side), the CPU 26 waits for a full pressing operation of the release button.

Step 116: The CPU 26 drives the image sensor 16 to take a subject image. Thereafter, the image processing unit 20 generates captured image data based on the image signal at the time of release. Note that when a registered person to be face-recognized is photographed, the image processing unit 20 performs the image processing set in S113 on the photographed image data.
Step 117: The CPU 26 records the photographed image data on the recording medium 29. Further, when a registered person as a face recognition target is photographed, the CPU 26 can also classify and record the photographed image data in a predetermined recording folder based on the directory path specified by the face registration memory 27.

  Further, the CPU 26 increments the number of times of photographing of the number of times of photographing in the face registration memory 27 based on the result of the face recognition in S112. At this time, the CPU 26 may increment the number of shootings of all recognized registered persons, or only the number of registered persons selected according to the priority order may be incremented. Thus, a series of operations in the “face recognition photographing mode” is completed.

  Hereinafter, the effect of the “face recognition photographing mode” of the present embodiment will be described. In the face recognition shooting mode, face recognition of the photographed person is performed based on the face recognition data (S112). When a face to be recognized is detected, various processes at the time of shooting are performed with settings corresponding to the registered person (S113, S116, S117). Therefore, the process reflecting the user's intention can be automatically performed when the registered person is photographed, and the convenience of the user who operates the electronic camera is remarkably improved.

In the face recognition photographing mode, when there is a high possibility that the face recognition process will not be completed within the time limit, the determination face recognition data is narrowed down using a predetermined algorithm (S107 to S111). Therefore, the calculation time of the face recognition process can be kept within the time limit, and the user can take a picture comfortably.
(Supplementary items of the embodiment)
As mentioned above, although this invention has been demonstrated by said embodiment, the technical scope of this invention is not limited to the said embodiment, For example, the following forms may be sufficient.

(1) The “face registration data registration time” data in the face registration memory may be sufficient to indicate when the original face registration image data was captured. Accordingly, the “registration time of the face recognition data” may be terminated by recording at any of the shooting year, shooting month, and shooting date, and recording of detailed time and the like may be omitted.
(2) In the face recognition process (S112) in the “face recognition shooting mode”, the face recognition unit 32 corrects the face recognition data according to the degree of distortion calculated from the focal length data of the shooting optical system 11. May be.

  (3) The electronic camera of the present invention does not necessarily match the configuration of the electronic camera of the embodiment. For example, the built-in memory 19 and the face registration memory 27 may be configured by a common memory. The face registration memory 27 may be connected to the CPU 26 via the data bus 28. Further, the optical blur correction mechanism including the swing sensor unit 13, the image blur correction lens 14, and the second lens driving unit 15 may be omitted. Furthermore, the communication unit 30 may be built in the electronic camera.

(4) The CPU 26 may record the data in the face registration memory 27 in the recording medium 29 so that the face registration memory 27 can be backed up. Further, the CPU 26 may acquire face recognition data or the like generated in advance by another electronic camera from the recording medium 29 so that the data in the face registration memory 27 can be shared between the electronic cameras.
(5) In the above embodiment, determination face recognition data is selected from the face recognition data group based on the number of face recognition data and the number of face regions. However, in the present invention, determination face recognition data is selected from the face recognition data group based only on the number of face recognition data related to the selected registered person without considering the number of detected face regions. May be.

The block diagram which shows the structure of the electronic camera of this embodiment Flow chart showing the operation of the “face recognition shooting mode” of the present embodiment Figure showing an example of the mode selection screen The figure which shows an example of the selection screen of a registered person

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 ... Shooting optical system, 16 ... Imaging device, 26 ... CPU, 27 ... Face registration memory, 31 ... Face detection part, 32 ... Face recognition part

Claims (7)

An image sensor that photoelectrically converts a subject image by a photographing optical system to generate an image signal;
A face detection unit that detects a face area of the shooting screen based on the image signal and extracts a feature point of the face of the photographed person from the face area part;
A recording unit that records face recognition data generated in advance based on feature points of a registered person's face;
An evaluation value corresponding to the number of face recognition data is based on the number of face recognition data included in a face recognition data group in which a plurality of face recognition data related to a registered person of face recognition is collected. A control unit that selects face recognition data for determination from the face recognition data group so as to fall below a threshold defined by the processing capability ;
A face recognition unit for determining whether the photographed person is a registered person for face recognition based on the feature points extracted by the face detection unit and the determination face recognition data;
Equipped with a,
The control unit weights and selects the determination face recognition data based on a priority set for each registered person when there are a plurality of registered persons for face recognition . An image sensor that photoelectrically converts a subject image by a photographing optical system to generate an image signal;
A face detection unit that detects a face area of the shooting screen based on the image signal and extracts a feature point of the face of the photographed person from the face area part;
A recording unit that records face recognition data generated in advance based on feature points of a registered person's face;
An evaluation value corresponding to the number of face recognition data is based on the number of the face recognition data included in a face recognition data group in which a plurality of the face recognition data related to a registered person as a face recognition target is collected. A control unit that selects face recognition data for determination from the face recognition data group so as to fall below a threshold defined by the processing capability;
A face recognition unit for determining whether the photographed person is a registered person for face recognition based on the feature points extracted by the face detection unit and the determination face recognition data;
With
The control unit selects the determination face recognition data by giving priority to a registered person with a large number of recorded face recognition data when there are a plurality of registered persons for face recognition .   The electronic camera according to claim 1, wherein the evaluation value is a value calculated based on the number of the face recognition data and the number of the face regions. The control unit, when a plurality of the face recognition data is recorded for the same registered person, selects the face recognition data for determination with priority on face recognition data having a new acquisition time. The electronic camera according to any one of claims 1 to 3 . In the case where a plurality of face recognition data having different shooting conditions at the time of acquisition are recorded for the same registered person, the control unit gives priority to face recognition data whose shooting conditions at the time of acquisition approximate to the current shooting conditions. electronic camera according to any one of claims 1 to 4, characterized by selecting said determination facial recognition data. The control unit selects the determination face recognition data by giving priority to the face recognition data corresponding to the front of the face when a plurality of the face recognition data having different face orientations are recorded for the same registered person. electronic camera according to any one of claims 1 to 5, characterized in that. The recording unit further records shooting number data indicating the number of shootings for each registered person,
The control unit, on the basis of the number of times of photographing data, any one of claims 1 to 6 which in favor of the number of times of photographing are often registered person and selects the determination face recognizing data Electronic camera as described in.

Images