JP5664559B2 - Imaging device and specific image detection program - Google Patents

Description

  The present invention relates to an imaging apparatus having a function for detecting a specific face during shooting, and a specific image detection program.

  In recent years, digital cameras equipped with a face detection function during shooting have been introduced. A digital camera having this type of function stores information for specifying a face image in advance, and whether or not a face image that can be specified by the information is included in captured images that are sequentially displayed through in the shooting mode. Determine whether. If it is included, when the half-press of the shutter key is detected, the image area including the face image is specified, and the focusing process and the AE process are performed on the image area ( For example, see Patent Document 1).

JP 2007-10898 A

  However, in the digital camera having the above-described function, if there are a plurality of face images in the captured image displayed through, the subject that the photographer does not detect the face image of the subject to be focused on and the subject is not intended to be recorded There are cases where a face image of a person is detected, or face images of all the subject persons in a captured image displayed through. As a result, when the photographer performs a half-press operation on the shutter key to instruct the digital camera to perform the focusing process, there is a possibility that the face image area of the person not intended by the photographer is determined as the area to be focused. .

  In this case, the photographer must repeat the half-press operation of the shutter key and the release thereof until the face image of the person to be designated as the area to be focused is detected. There was a possibility that the problem of taking time would arise.

  The present invention has been made in view of such conventional problems, and an object of the present invention is to provide an imaging apparatus and a face area detection program that enable recording with detection of a specific image with detection accuracy desired by a photographer. And

In order to solve the above-mentioned problem, in the invention according to claim 1, an image pickup means and a determination means for determining whether or not a specific image to be focused is included in images sequentially picked up by the image pickup means. When the determination unit determines that the specific image is included, the counting unit that counts the number of the specific images, the first instruction unit that instructs the focusing process, and the first instruction unit If the determination unit determines that the specific image is included, the in-focus processing is performed with the image area where the specific image is present as the in-focus position. And a second instruction means for instructing recording of the image after the focusing process by the focusing means, and recording the image on the recording means when it is determined that the recording is instructed by the second instruction means. On the other hand, recording is not instructed Determines that the instructions by the serial first instruction means is released, and, as the number counted by the counting means when the predetermined value or more, reducing the accuracy of detection of the specific image to be focused the alloy by the determination means Control means for controlling.

According to a second aspect of the present invention, in the first aspect of the invention, when the number counted by the counting unit is less than the predetermined value, the control unit performs the focusing by the determining unit. Control is performed to increase the accuracy of detection of a specific image to be performed.

According to a third aspect of the invention, in the first or second aspect of the invention, the specific image is information defined by a specific amount , and a plurality of specific images are stored for each accuracy. An information storage unit; and a reading unit that reads information corresponding to a predetermined accuracy from the specific image information storage unit, and the determination unit is sequentially imaged based on the information read by the reading unit. It is determined whether or not a specific image to be focused is included in the image to be focused.

According to a fourth aspect of the invention, in the invention according to any one of the first to third aspects, when the captured image is an image of a person, the specific image is a face image of the person. It is characterized by that.

Further, in the invention according to claim 5, the computer having the apparatus having the imaging unit determines whether or not a specific image to be focused is included in the images sequentially captured by the imaging unit. Means, when the determination means determines that the specific image is included, the counting means for counting the number of the specific images, the first instruction means for instructing the focusing process, and the first instruction means. When the focusing process is instructed, when the determining unit determines that the specific image is included, the focusing unit performs the focusing process with the image area where the specific image is present as the in-focus position. After the focusing process by the focusing means, a second instruction means for instructing recording of the image, and if it is determined that the recording is instructed by the second instruction means, the image is recorded in the recording means, Recording directed If it is determined that the instruction by the first instruction unit has been canceled and the number counted by the counting unit is equal to or greater than a predetermined value , the accuracy of detection of the specific image to be focused by the determination unit It is made to function as a control means for controlling to lower the pressure .

According to the present invention, a specific image can be detected and recorded with detection accuracy desired by a photographer .

1 is a block diagram of a digital camera according to an embodiment of the present invention. It is a conceptual diagram which shows the structure of the dictionary set for faces. It is explanatory drawing which shows a rotation angle. (A) is explanatory drawing which shows a pixel area range, (b) is a figure which shows the large area, medium area, and small area which are made into a pixel area range. It is a figure which shows the setting content of a detection accuracy setting part. It is a display transition diagram in a 1st embodiment. It is a flowchart which shows the process sequence in 1st Embodiment. It is a flowchart which shows the process sequence in 2nd Embodiment. It is a display transition diagram in a 3rd embodiment. It is a flowchart which shows the process sequence in 3rd Embodiment.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
(First embodiment)
FIG. 1 is a block diagram showing a schematic configuration of a digital camera common to the respective embodiments of the present invention. This digital camera has functions of AE (automatic exposure), AF (autofocus), and face area detection, and mainly includes the following parts.

  That is, the digital camera has a CCD 2 as an imaging unit that photoelectrically converts an optical image of a subject imaged by the optical system 1 and outputs it as an imaging signal. The CCD 2 is driven by a drive signal generated by a CCD drive circuit 4 based on a timing signal generated by a timing generator (TG) 3. The output signal of the CCD 2 is converted into a digital signal by the A / D converter 5 and then sent to the image processing unit 6.

  The optical system 1 includes a focus lens and a lens motor for moving the focus lens on the optical axis, and the lens motor is driven by a drive signal generated by the lens drive circuit 8 in accordance with a command from the control unit 7. Thus, the position of the focus lens on the optical axis is controlled.

  The image processing unit 6 performs RGB interpolation processing for generating R, G, and B color component data (RGB data) for each pixel with respect to the imaging signal after A / D conversion, and luminance signal (Y) from the RGB data. And YUV conversion processing for generating YUV data consisting of color difference signals (U, V) for each pixel, and digital signal processing for improving image quality such as auto white balance and edge enhancement. The YUV data converted by the image processing unit 6 is sequentially stored in the SDRAM 9.

  In addition, the digital camera has a display unit 10 including a liquid crystal monitor and a driving circuit thereof. When the recording mode for shooting is set, the YUV data stored in the SDRAM 9 is converted into a video signal in the image processing unit 6 every time one frame is accumulated in the SDRAM 9, and then the display unit 10 (liquid crystal monitor). ) Is displayed on the screen as a through image. When shooting in the recording mode, the image data temporarily stored in the SDRAM 9 is compressed and encoded by the JPEG conversion unit 11 using the JPEG method, and then configured by, for example, various memory cards via the external memory I / F 12. It is recorded in the external memory 13 as a still image file.

  The still image file recorded in the external memory 13 is appropriately read in accordance with the user's selection operation in the reproduction mode, is decompressed by the JPEG conversion unit 11 and developed as SDRAM data on the SDRAM 9, and then is statically displayed on the display unit 10. Displayed as an image.

  The key input block 14 includes a plurality of keys such as a shutter key 141, a mode selection key (not shown), a power key, and a MENU key, and the operation state of each key is scanned by the control unit 7 as needed. Note that the shutter key 141 has a so-called half shutter function capable of two-step operation, that is, a half-pressed position for a user to make a shooting preview and a full-pressed position for instructing an actual shooting operation. .

The control unit 7 is composed of a CPU and its peripheral circuits, etc., and controls each part of the digital camera, AE control based on luminance information included in the imaging signal, AF control by a contrast detection method, and a face area described later Perform detection control. Various programs for causing the control unit 7 to perform these controls and various data necessary for each control are stored in the flash memory 16 which is a nonvolatile memory capable of rewriting stored data. The flash memory 16 also stores setting information regarding various functions of the digital camera, including a face detection mode setting table 161. The face detection mode setting table 161 is a table that stores face image detection accuracy Lv.1 to Lv.3 as flag information.
Lv.1-Lv.3 is
Lv.1 = The detection accuracy is rough (low), but the detection speed is fast.
Lv.2 = Detection accuracy Almost intermediate level Lv.3 = The detection accuracy is fine (high), but the detection speed is slow.
There is a relationship.

  The face recognition processing unit 15 performs face image detection processing for detecting a face image of a subject image using a face feature dictionary described later in the image data for one frame generated by the image processing unit 6. More specifically, a frame memory for temporarily storing the image data for one frame, an image processing circuit for performing various image processing necessary for the face image detection operation, and a dictionary set 151, A detection number setting unit 152, a rotation angle setting unit 153, a range setting unit 154, and a detection accuracy setting unit 155 are provided.

  FIG. 2 is a configuration diagram showing the dictionary set 151. The dictionary set 151 stores a plurality (N) of feature dictionaries A to N. Each of the feature dictionaries A to N is characterized by, for example, the shape and size of a plurality of parts (eyes, nose, mouth) of the face image, the outline of the face image, the position of each part when viewed from a substantially front direction, and the like. A plurality of pieces of information are registered for each age group. The face recognition processing unit 15 sequentially reads out an image area (a pixel area range 21 described later) of YUV data obtained as a result of imaging, and performs edge extraction and the like on the image area (pixel area range 21). The feature amount is calculated, the degree of coincidence between the calculated feature amount and the feature amount registered in the feature dictionary A to N in the dictionary set 151 is evaluated, and the evaluation value is output to the control unit 7. When these evaluation values are input, the control unit 7 determines that a face image exists when the image area having a predetermined area has the highest evaluation value, that is, substantially matches.

  The detection number setting unit 152 sets the upper limit of the number of detected face images in image data for one frame, and the upper limit number set in the present embodiment is “1”, “3”, “8”. Any of the above.

  The rotation angle setting unit 153 sets an allowable rotation angle of the face, and the allowable rotation angle set in the present embodiment is “± α”, “± β (>) as shown in FIG. α) ”or“ ± γ (> β) ”.

  The range setting unit 154 sets an area for detecting a face image in an image for one frame, and the detection areas in the present embodiment are “large area”, “medium area”, and “small area”. There are three types. That is, as shown in FIG. 4 (a), a rectangular detection area 21 corresponding to the size of the face image to be detected is set on the imaging surface 20 of the CCD 2, and the detection area 21 is sequentially moved to set the face. Try to detect the image. As for the detection area 21, as shown in FIG. 4B, the largest area (−x3, −y3) − (x3, y3) is the large area 210, and the next largest area (−x2, −y2). − (X2, y2) is a middle region 211, and the smallest region (−x1, −y1) − (x1, y1) is a small region 212.

  The detection accuracy setting unit 155 stores the setting contents of the dictionary set 151, the detection number setting unit 152, and the range setting unit 154 as flag information corresponding to the detection accuracy Lv.1 to Lv.3.

  Therefore, as shown in FIG. 5, the detection accuracy setting unit 155 indicates that the detection accuracy is rough (low) but the detection speed is fast = Lv. The rotation angle is “± α”, the upper limit of the number of detections is “1”, the detection area is the large area 210 (−x3, −y3) − (x3, y3), and the detection accuracy level flag = 1. Is remembered. Further, the detection accuracy corresponds to the intermediate level = Lv.2, the reference feature dictionary is “A” and “B”, the allowable rotation angle is “± β (> α)”, and the upper limit of the number of detections Is “3”, the detection areas are large areas 210 (−x3, −y3) − (x3, y3) and middle areas 211 (−x2, −y2) − (x2, y2). Flag = 1 is stored. Furthermore, the detection accuracy is fine (high), but the detection speed is slow = Lv.3, “A”, “B”, “CN” are included in the feature dictionary to be referenced, and “ ± γ (> β) ”, the upper limit of the number of detections is“ 8 ”, the detection area is the large area 210 (−x3, −y3) − (x3, y3), the middle area 211 (−x2, −y2) )-(X2, y2) and small area 212 (-x1, -y1)-(x1, y1), detection accuracy level flag = 1 is stored.

  Next, the operation of the digital camera having the above configuration will be described. When the power switch is turned on and the shooting mode is set, the control unit 7 displays the subject image 101 (through image) on the display unit 10 and stores it in the flash memory 16 as shown in FIG. The processing is executed as shown in the flowchart of FIG.

  First, the face detection mode setting table 161 is referred to (step S101). At this time, in the face detection mode setting table 161, when the flag = 1 is set in Lv.1 as an initial value, for example, “Lv.1” is displayed as the detection accuracy level icon 105 as shown in FIG. Is displayed on the display unit 10.

  Next, the feature dictionary (step S102), the allowable rotation angle (step S103), and the detection area (step S104) are read according to the state of the flag in the referenced face detection mode setting table 161.

  That is, if the detection accuracy is Lv.1 in the face detection mode setting table 161, the feature dictionary referred to by Lv.1 in the detection accuracy setting unit 155 shown in FIG. Since the allowable rotation angle is “± α” (step S102), the allowable rotation angle ± α is read (step S103). The detection area is “large area 210 (−x3, −y3) − (x3, y3)”. Therefore, the large area 210 (−x3, −y3) − (x3, y3) is read out as the detection area 21 (step S104).

  In this way, data used in face image detection processing (step S105) described later can be facilitated in advance by a simple process of reading data from the detection accuracy setting unit 155 with reference to the face detection mode setting table 161. .

  Next, face image detection processing is executed (step S105). In this detection process, when the detection accuracy is Lv.1, the detection area 21 is read out in step S104 on the imaging surface 20 of the CCD 2 that outputs image data for one frame that is stored in the SDRAM 9 and displayed through. Large area 210 (−x3, −y3) − (x3, y3). Then, by sequentially moving the large area 210, the imaging surface 20 is scanned (scanned) with the large area 210 as a unit to try to detect a face image. At this time, it is determined whether or not the allowable rotation angle has the feature amount stored in the feature dictionary A read in step S102 as ± α read in step S103.

  Subsequently, it is determined whether or not a half-press operation of the shutter key 141 is detected (step S106). If a half-press operation of the shutter key 141 is detected, it is determined whether or not a face image is detected (step S106). Step S107). That is, whether or not a region having a feature value stored in the feature dictionary A within the allowable rotation angle ± α, that is, a face image is detected in any large region 210 on the imaging surface 20 of the scanned CCD 2. Judging.

  If the face image is not detected as a result of the determination in step S107, that is, the human face image 103 shown in FIG. 6A is much smaller than the large area 210, and the allowable rotation angle ± If it is not detected as a region having the feature amount stored in the feature dictionary A within α, the process proceeds from step S107 to step S109, the in-focus region is determined according to the image contrast evaluation value, and the in-focus region is determined. Processing, AE processing, and WB adjustment processing are executed.

  Subsequently, it is determined whether a full-press operation or an operation cancellation is detected from the shutter key 141 (step S110). If release of the operation of the shutter key 141 is detected, the detection accuracy is increased by one step (step S112), and the processing from step S11 is repeated.

  Therefore, if the current detection accuracy is Lv.1, as described above, the process in step S112 is reset to Lv.1 = 0 in the face detection mode setting table 161, and Lv.2 = 1. The detection accuracy is set to Lv.2 by maintaining Lv.3 = 0 as it is.

  When the detection accuracy of the face detection mode setting table 161 is set to Lv.2 in this way, when the processing from step S101 is executed again, as shown in FIG. 105 displays “Lv.2” on the display unit 10. Further, since the feature dictionaries referred to by Lv.2 in the detection accuracy setting unit 155 shown in FIG. 5 are “A” and “B”, the feature dictionaries A and B are read (step S102), and the allowable rotation angle is “± Since β (> α) ”, the allowable rotation angle ± β (> α) is read (step S103), and the detection areas are“ large area 210 (−x3, −y3) − (x3, y3) ”and“ medium. Area 211 (−x2, −y2) − (x2, y2) ”, the detection area 21 includes a large area 210 (−x3, −y3) − (x3, y3) and a middle area 211 (−x2, −y2). )-(X2, y2) is read (step S104). Subsequently, the processing after step S105 described above is executed.

  At this time, in the detection accuracy Lv.2, the detection areas are the large area 210 and the middle area 211. Therefore, in the face image detection process in step S107, the detection area 21 is set to the large area 210 on the imaging surface 20 of the CCD 2. And the middle region 211. Then, by sequentially moving the large area 210 and the middle area 211, the imaging surface 20 is scanned (scanned) in units of the large area 210 and the middle area 211, and a face image is obtained for each of the large area 210 and the middle area 211. Try to detect. At this time, it is determined whether or not the feature amount stored in the feature dictionaries A and B has an allowable rotation angle ± β.

  If the face image is not detected as a result of the determination in step S107, that is, the human face image 103 shown in FIG. 6B is still smaller than the middle region 211 and within the allowable rotation angle ± β. If it is not detected as a region having the feature amount stored in the feature dictionaries A and B, the process proceeds from step S107 to step S109. At this time, if it is detected that the shutter key 141 is fully pressed, the process proceeds from step S109 to step S110 and step S112.

  If the detection accuracy at the time of executing the process of step S112 is Lv.2 as compared with the process at the previous step S112 as described above, the process of increasing the detection accuracy at the current step S112 by one step, In the face detection mode setting table 161, Lv.1 = 0 is maintained as it is, is reset to Lv.2 = 0, is set to Lv.3 = 1, and the detection accuracy is set to Lv.3.

  When the detection accuracy of the face detection mode setting table 161 is set to Lv.3 as described above, when the processing from step S101 is executed again, as shown in FIG. 105 displays “Lv.3” on the display unit 10. Further, since the feature dictionaries referred to by Lv.3 in the detection accuracy setting unit 155 shown in FIG. 5 are “A”, “B”, and “C to N”, the feature dictionaries A to N are read (step S102) and allowed. Since the rotation angle is “± γ (> β)”, the allowable rotation angle ± γ (> β) is read (step S103), and the detection region is “large region 210 (−x3, −y3) − (x3, y3). ) ”,“ Medium region 211 (−x2, −y2) — (x2, y2) ”, and“ small region 212 (−x1, −y1) — (x1, y1) ”, so that the detection pair region 21 is large. An area 210 (−x3, −y3) − (x3, y3), a middle area 211 (−x2, −y2) − (x2, y2), and a small area 212 (−x1, −y1) − (x1, y1) Reading (step S104), and subsequently the processing after step S105 described above is executed.

  At this time, in the detection accuracy Lv.3, the detection areas are the large area 210, the middle area 211, and the small area 212. Therefore, in the detection process of step S105, the detection area 21 is set to the large area on the imaging surface 20 of the CCD 2. 210, a middle region 211, and a small region 212. Then, by sequentially moving the large area 210, the medium area 211, and the small area 212, the imaging surface 20 is scanned (scanned) in units of the large area 210, the medium area 211, and the small area 212. Attempts to detect a face image for each of the regions 211 and 212. Further, at this time, it is determined whether or not the feature amount stored in the feature dictionary A to N as the allowable rotation angle ± γ.

  Further, the determination in step S107 is performed when the face image is detected, that is, the human face image 103 shown in FIG. 6C is within the small region 212, and the feature dictionary A, within the allowable rotation angle ± β. If it is detected as a region having the feature quantity stored in B, the answer is YES. Accordingly, the process proceeds from step S107 to step S108. In step S108, as shown in FIG. 6C, the face frame 104 surrounding the area including the detected face image is superimposed and displayed, and focusing processing, AE processing, and WB adjustment are performed on this area. Execute the process.

  That is, the face image data in the face frame 104 is used as a contrast evaluation value acquisition region (AF evaluation region), and AF control is performed to drive the lens driving circuit 8 so as to focus on the face image. Execute exposure measurement and exposure adjustment processing with the detected face image data, perform AE control based on the luminance information included in the imaging signal in this area, and execute white balance adjustment processing with the detected face area Adjust the white balance so that the face image can be recorded properly.

  In step S110 subsequent to step S108, when the full press operation of the shutter key 141 is detected, the process proceeds from step S110 to step S111 to execute compression encoding / recording processing. That is, the image data for one frame at the time point when the full pressing operation of the shutter key 141 is detected is taken into the SDRAM 9, and the image data temporarily stored in the SDRAM 9 is compression-encoded by the JPEG method in the JPEG conversion unit 11, and subsequently this The compressed and encoded image data is written as a still image file in the external memory 13 via the external memory I / F 12. Thereafter, the processing from step S106 is executed.

  Therefore, when a face image is detected, the focus processing is performed on the face image by the processing in step S108 described above, and exposure adjustment and white balance adjustment are performed on the image portion. Therefore, when a face image is detected, an image in which the face image of the subject person is in focus and exposure and white balance are appropriate can be recorded as a still image file.

  Further, if the half-press of the shutter key 141 is released, the detection accuracy is changed, so that the troublesomeness that the photographer interrupts the preparation for shooting recording and changes the setting to eliminate erroneous detection, The specific image can be detected and recorded with the desired detection accuracy.

(Second Embodiment)
In the second embodiment of the present invention, as shown in FIG. 1, the control unit 7 is provided with a counter 71 and a collation unit 72. The counter 71 counts the number of times the shutter key operation is released after the shutter key 141 is pressed halfway. The collation unit 72 collates the count value of the counter 71 with a predetermined number, and is used when determining whether or not the count value exceeds the predetermined number.

  FIG. 8 is a flowchart showing a processing procedure in the second embodiment. In this flowchart, the processes of steps S201 to S211 and S214 are the same as all the steps in the flowchart of FIG. 7 described above, that is, steps S201 to S212.

  If the release of the operation of the shutter key 141 is detected as a result of the determination in step S210, the count value of the counter 71 is increased (step S212). Subsequently, it is determined whether or not the counted value of the increased counter 71 exceeds a predetermined number (step S213). If not, the processing from step S205 is repeated.

  At this time, since the face frame 104 (see FIG. 6C) is not displayed on the display unit 10 even when the shutter key 141 is half-pressed, the photographer determines that no face image has been detected. Then, the operation of the shutter key 141 is canceled, the direction and angle of view of the digital camera are adjusted, and the shutter key 141 is half-pressed so as to detect a desired face image again.

  If the face frame 104 is displayed on the display unit 10 and a face image is detected as a result of the operation by the photographer, the photographer fully presses the shutter key 141, so that the determination in step S210 is YES. It becomes. Therefore, the process proceeds from step S210 to step S211, and the compression encoding / recording process described above is executed.

  However, if the photographer adjusts the orientation and angle of view of the digital camera multiple times and no face image is detected even if the shutter key 141 is pressed halfway each time, the operation of the shutter key 141 is canceled each time. The count value is increased (step S212). As a result, the determination in step S213 is YES, and the determination in step S213 is YES as described above because the current face image detection accuracy is too low. Therefore, when the determination in step S213 is YES as described above, the detection accuracy is increased by one step (step S214), and the processing from step S201 is repeated.

  Therefore, according to the present embodiment, since the detection accuracy is changed according to the number of times the shutter key 141 is released halfway, the image is taken with an appropriate detection accuracy according to the shooting operation (will) of the photographer. can do.

(Third embodiment)
In the third embodiment of the present invention, the counter 71 provided in the control unit 7 shown in FIG. 1 counts the number of face images detected by the face image detection process. The collation unit 72 collates the number counted by the counter 71 with the upper limit (“1”, “3”, “8”) of the detection number of the detection accuracy setting unit 155 shown in FIG. This is used when determining whether or not the counted number is equal to or greater than the upper limit.

  Next, the operation of the digital camera according to the third embodiment will be described. When the power switch is turned on and the shooting mode is set, the control unit 7 displays the subject image on the display unit 10 as shown in FIG. 9A, and the control program stored in the flash memory 16 is displayed. Based on this, processing is executed as shown in the flowchart of FIG.

  First, the face detection mode setting table 161 is referred to (step S101). At this time, in the face detection mode setting table 161, for example, if flag = 1 is set in Lv.2 as an initial value, “Lv.2” is displayed as the detection accuracy level icon 105 as shown in FIG. Is displayed on the display unit 10.

  Next, according to the state of the flag in the referenced face detection mode setting table 161, the feature dictionary (step S302), the allowable rotation angle (step S303), the upper limit of the number of detections (step S304), and the detection area (step S305) are displayed. Read each one.

  That is, if the detection accuracy is Lv.2 in the face detection mode setting table 161, the feature dictionary referred to by Lv.2 in the detection accuracy setting unit 155 shown in FIG. The feature dictionaries A and B are read (step S302). Since the allowable rotation angle is “± β (> α)”, the allowable rotation angle ± β (> α) is read (step S303). Since the upper limit of the detection number is “3”, the upper limit “3” of the detection number is read (step S304), and the detection area is “large area 210 (−x3, −y3) − (x3, y3)” and Since it is “middle area 211 (−x2, −y2) − (x2, y2)”, the detection area 21 is a large area 210 (−x3, −y3) − (x3, y3) and a middle area 211 (−x2, -Y2)-(x2, y2) is read (step S305).

  Next, face image detection processing is executed (step S306). In this face image detection processing, when the detection accuracy is Lv.2, the detection area 21 is set on the imaging surface 20 of the CCD 2 that outputs image data for one frame stored in the SDRAM 9 and displayed as a through image. A large area 210 (−x3, −y3) − (x3, y3) and a middle area 211 (−x2, −y2) − (x2, y2) are assumed. Then, by sequentially moving the large area 210 and the middle area 211, the imaging surface 20 is scanned (scanned) in units of the large area 210 and the middle area 211, and a face image is obtained for each of the large area 210 and the middle area 211. Try to detect. At this time, it is determined whether or not the feature amount stored in the feature dictionaries A and B has the allowable rotation angle ± β.

  Subsequently, it is determined whether or not a half-press operation of the shutter key 141 is detected (step S307). If a half-press operation of the shutter key 141 is detected, it is determined whether or not a face image is detected (step S307). Step S308). That is, in any of the large area 210 or the middle area 211 on the imaging surface 20 of the scanned CCD 2, the area having the feature amount stored in the feature dictionary A or B within the allowable rotation angle ± β, that is, the face image It is determined whether or not is detected.

  If a face image is detected as a result of the determination in step S308, the process proceeds to step S9. That is, as shown in FIG. 9A, when the face image 111 of the person A, the face image 112 of the person B smaller than the face image 111 of the person A, and the face image 112 of the person C exist in the subject image 101. In other words, the face image 111 of the person A is detected by scanning with the large area 210, and the face image 112 of the person B and the face image 113 of person C are detected by the scanning with the middle area 211. Accordingly, the determination in step S308 is YES, and the process proceeds from step S308 to step S309.

  In step S309, as shown in FIG. 9A, the face frame 104 surrounding the detected face image is superimposed and displayed for each of the face images 111, 112, and 113, and the face image is focused. , AE processing and WB adjustment processing are executed. That is, the image data in the face image 111 detected in the large area 210 is used as a contrast evaluation value acquisition area (AF evaluation area), and AF control is performed so that the face image is focused, and the lens driving circuit 8 is Drives and performs exposure measurement and exposure adjustment processing on the detected face image, performs AE control based on luminance information included in the imaging signal in the face image, and further executes white balance adjustment processing on the detected face area Then, the white balance is adjusted so that the face area is appropriate.

  If the face image is not detected as a result of the determination in step S308, the in-focus area is determined according to the image contrast evaluation value, and the in-focus process, the AE process, and the WB adjustment process are executed. (Step SS310).

  Subsequently, it is determined whether a full-press operation or an operation cancellation is detected from the shutter key 141 (step SS311). If the release of the operation of the shutter key 141 is detected, it is determined whether or not a face image has been detected (step S313). Therefore, as in this example shown in FIG. 9A, if a face image is detected, the determination in step S313 is YES, the process proceeds from step S313 to step S314, and the number of detected face images. Is counted by the counter 71 (step S314). At this time, in the present example shown in FIG. 9A, since three face images are detected, the counter 71 counts the number “3”.

  Next, it is determined whether or not the detected number is the same as the upper limit of the detected number set with the detection accuracy (step S315). Here, in the present example, the number detected as described above is “3”, and the upper limit of the detection number of Lv.2 in the detection accuracy setting unit 155 of FIG. Therefore, the determination in step S315 is YES, the detection accuracy is lowered by one step (step S316), and the processing from step S301 is repeated.

  Therefore, if the detection accuracy in the current face detection mode is Lv.2 as described above, the detection accuracy Lv.3 = 0 is maintained as it is in the face detection mode setting table 161 by the processing in step S316. , Lv.2 = 0 and Lv.1 = 1 are set, and the detection accuracy Lv.1 is set.

  When the detection accuracy Lv.1 = 1 is set in the face detection mode setting table 161 and the detection accuracy Lv.1 is set as described above, when the processing from step S301 is executed again, FIG. In the illustrated detection accuracy setting unit 155, the feature dictionary referred to by Lv.1 is “A”, so the feature dictionary A is read (step S302). Further, since the allowable rotation angle is “± α”, the allowable rotation angle ± α is read (step S303), and since the upper limit of the detected number is “1”, the upper limit “1” of the detected number is read (step S304). ). Since the detection area is “large area 210 (−x3, −y3) − (x3, y3)”, the large area 210 (−x3, −y3) − (x3, y3) is read as the detection area 21 ( Step S305) Subsequently, the processing after step S306 described above is executed.

  At this time, since the detection area is only the large area 210 in the detection accuracy Lv.1, the detection area 21 is set as the large area 210 on the imaging surface 20 of the CCD 2 in the face image detection process in step S306. Then, by sequentially moving the large area 210, the imaging surface 20 is scanned (scanned) with the large area 210 as a unit, and detection of a face image is attempted for each large area 210. At this time, it is determined whether the feature amount stored in the feature dictionary A has an allowable rotation angle ± α.

  Subsequently, it is determined whether or not a half-press operation of the shutter key 141 is detected (step S307). If a half-press operation of the shutter key 141 is detected, it is determined whether or not a face image is detected (step S307). Step S308). That is, it is determined whether or not an area having a feature amount stored in the feature dictionary A within an allowable rotation angle ± α is detected in the large area 210 on the imaging surface 20 of the scanned CCD 2.

  If a face image is detected as a result of the determination in step S308, the process proceeds to step S9. That is, as shown in FIG. 9B, the face image 111 of the person A, the face image 112 of the person B smaller than the face image 111 of the person A, and the face image 112 of the person C exist in the subject image 101. The face image 112 of the person B and the face image 11 of the person C are detected by the scan by the middle region 211 in Lv.2 described above, but are detected by the scan by the large region 210 in this Lv.1. Not. On the other hand, the face image 111 of the person A is detected by scanning with the large area 210 in this Lv.1. Accordingly, the determination in step S308 is YES, and the process proceeds from step S308 to step S309.

  In step S309, as shown in FIG. 9B, the face frame 104 surrounding the detected face image is superimposed on the face image 111 of the person A and the face image is focused, AE Processing and WB adjustment processing are executed. Subsequently, it is determined whether a full-press operation or an operation release has been detected from the shutter key 141 (step SS311). If an operation release of the shutter key 141 is detected, the process proceeds to step S313 described above.

  If a full press operation of the shutter key 141 is detected, the process proceeds from step S311 to step S312 to execute compression encoding / recording processing. That is, the image data for one frame at the time point when the full pressing operation of the shutter key 141 is detected is taken into the SDRAM 9, and the image data temporarily stored in the SDRAM 9 is compression-encoded by the JPEG method in the JPEG conversion unit 11, and subsequently this The compressed and encoded image data is written as a still image file in the external memory 13 via the external memory I / F 12. Thereafter, the processing from step S306 is executed.

  Therefore, as shown in FIG. 9B, when the face image 111 of the person A is detected, the face image 111 is focused and the face image 111 is processed by the process in step S309 described above. The exposure adjustment and white balance adjustment are made at. Therefore, when a face image is detected, an image in which the face of the person A is focused and whose exposure and white balance are appropriate can be recorded as a still image file.

  On the other hand, if it is determined in step S313 that no face image has been detected, and if it is determined that the number counted in step S315 is less than the preset upper limit of the number of detections, step S317 is entered. Then, the detection accuracy is increased by one step, and then the processing from step S301 is repeated.

  Therefore, assuming that the current detection accuracy is Lv.2, as described above, the processing in step S317 resets Lv.2 = 0 and sets Lv.3 = 1 in the face detection mode setting table 161. Thus, Lv.1 = 0 is maintained as it is, so that the detection accuracy is set to Lv.3.

  When the detection accuracy of the face detection mode setting table 161 is set to Lv.3 as described above, “Lv.3” is displayed as the detection accuracy level icon 105 when the processing from step S101 is performed again. 10 is displayed. Further, since the feature dictionaries referred to by Lv.3 in the detection accuracy setting unit 155 shown in FIG. 5 are “A”, “B”, and “C to N”, the feature dictionaries A to N are read (step S302) and allowed. Since the rotation angle is “± γ (> β)”, the allowable rotation angle ± γ (> β) is read (step S303), and since the upper limit of the number of detections is “8”, the upper limit of the detection number is “8”. (Step S304), the detection areas are “large area 210 (−x3, −y3) − (x3, y3)”, “medium area 211 (−x2, −y2) − (x2, y2)” and “small”. Area 212 (−x1, −y1) − (x1, y1) ”, the detection area 21 is the large area 210 (−x3, −y3) − (x3, y3), the middle area 211 (−x2, −y2). )-(X2, y2) and the small area 212 (-x1, -y1)-(x1, y1) are read out (Step S304) Subsequently, the processing after Step S306 described above is executed.

  At this time, in the detection accuracy Lv.3, the pixel area ranges to be detected are the large area 210, the medium area 211, and the small area 212. Therefore, in the face image detection process in step S306, the imaging surface 20 of the CCD 2 is used. , The pixel area range 21 to be detected is a large area 210, a medium area 211, and a small area 212. Then, by sequentially moving the large area 210, the medium area 211, and the small area 212, the imaging surface 20 is scanned (scanned) in units of the large area 210, the medium area 211, and the small area 212. Attempts to detect a face image for each of the regions 211 and 212. Further, at this time, it is determined whether or not the feature amount stored in the feature dictionary A to N as the allowable rotation angle ± γ.

  Subsequently, it is determined whether or not a half-press operation of the shutter key 141 is detected (step S307). If a half-press of the shutter key 141 is detected, it is determined whether or not a face image is detected (step S307). S308). That is, in any one of the large area 210, the middle area 211, or the small area 212 on the imaging surface 20 of the scanned CCD 2, an area having a feature amount stored in the feature dictionary A or B within an allowable rotation angle of ± β. It is determined whether or not it has been detected.

  If the face image is detected as a result of the determination in step S308, that is, the face image of the person is within the small area 212 and is stored in the feature dictionary A to N within the allowable rotation angle ± γ. If it is detected as a region having, the process proceeds from step S308 to step S309. In step S309, the face frame 104 surrounding the detected face image is superimposed and displayed, and focusing processing, AE processing, and WB adjustment processing are executed on the face image.

  In step S311 following step S309, when the full pressing operation of the shutter key 141 is detected, the process proceeds from step S311 to step S312 to execute compression encoding / recording processing.

  Therefore, in the present embodiment, when the counted number of face images is the same as the upper limit of the detection number set by the detection accuracy, the detection accuracy in the detection mode is lowered by one step (step S316), and the upper limit is set. If it is less, the detection accuracy in the detection mode is increased by one step (step S317), so that it is possible to photograph with appropriate detection accuracy according to the photographing operation (intention) of the photographer.

  In each embodiment, the detection accuracy is provided in three stages of Lv.1 to Lv.1 to 3. However, the detection accuracy may be smaller or more stages. In the present embodiment, not only the number of feature dictionaries to be used but also the allowable rotation angle is varied according to the detection accuracy. However, only the number of feature dictionaries to be used may be varied without performing control related to the allowable rotation angle.

  In the embodiment, the specific image to be focused on is a person's face image. However, the specific image is not limited to a person's face image, and may be a partial image of an object. Furthermore, in the embodiment, the case where the present invention is applied to a digital camera has been described in detail. However, the present invention is not limited to this, and the present invention is not limited to this as long as the apparatus has an imaging unit or can execute a program including an imaging process. It is applicable without being done.

1 Optical system 2 CCD
4 CCD drive circuit 6 Image processing unit 7 Control unit 10 Display unit 11 JPEG conversion unit 15 Face recognition processing unit 16 Flash memory 20 Imaging surface 21 Pixel area range 71 Counter 72 Verification unit 101 Subject image 103 Face image 104 Face frame 105 Detection accuracy Level icon 141 Shutter key 151 Dictionary set 152 Detection number setting unit 153 Rotation angle setting unit 154 Range setting unit 155 Detection accuracy setting unit 161 Face detection mode setting table 210 Large area 211 Medium area 212 Small area

Claims (5)

Imaging means;
Determination means for determining whether or not a specific image to be focused is included in images sequentially captured by the imaging means;
A counting unit that counts the number of the specific images when the determination unit determines that the specific image is included;
First instruction means for instructing focusing processing;
When the focusing process is instructed by the first instructing unit, when the determining unit determines that the specific image is included, the image area where the specific image exists is set as the in-focus position. A focusing means for performing a focusing process;
A second instruction means for instructing recording of the image after the focusing process by the focusing means;
If it is determined that the recording is instructed by the second instruction means, the image is recorded on the recording means, while it is determined that the instruction by the first instruction means is released without instructing recording, and the counting means Control means for controlling to reduce the accuracy of detection of the specific image to be focused by the determination means when the number counted by the above is a predetermined value or more;
An imaging apparatus comprising: 2. The control unit according to claim 1, wherein when the number counted by the counting unit is less than the predetermined value, the control unit controls the accuracy of detecting the specific image to be focused by the determining unit. The imaging device described. The specific image is information defined by a specific amount, and specific image information storage means for storing a plurality of this information for each accuracy,
Reading means for reading out information corresponding to a predetermined accuracy from the specific image information storage means;
Further comprising
3. The determination unit according to claim 1, wherein the determination unit determines whether or not a specific image to be focused is included in the sequentially captured image based on information read by the reading unit. Imaging device.   The imaging apparatus according to claim 1, wherein when the captured image is an image of a person, the specific image is a face image of the person. A computer included in an apparatus including an imaging unit,
Determining means for determining whether or not a specific image to be focused is included in images sequentially captured by the imaging means;
Counting means for counting the number of the specific images when the determination means determines that the specific images are included;
First instruction means for instructing focusing processing;
When the focusing process is instructed by the first instructing unit, when the determining unit determines that the specific image is included, the image area where the specific image exists is set as the in-focus position. A focusing means for performing a focusing process,
A second instruction means for instructing recording of the image after the focusing process by the focusing means;
If it is determined that the recording is instructed by the second instruction means, the image is recorded on the recording means, while it is determined that the instruction by the first instruction means is released without instructing recording, and the counting means Control means for controlling to reduce the accuracy of detection of the specific image to be focused by the judging means when the number counted by the above is a predetermined value or more;
A specific image detection program which is made to function as a computer program.

Images