JP5665512B2 - Imaging apparatus and main subject detection method - Google Patents

Description

  The present invention relates to an imaging device and a main subject detection method, and more particularly to an imaging device and a main subject detection method for detecting a main subject using a result of multipoint distance measurement.

  As one of methods for detecting the main subject in the screen, a method is known in which the main subject is detected using multipoint distance measurement in the screen and the result (for example, Patent Documents 1 and 2, etc.). ). This method, for example, divides the screen into a plurality of areas, measures the distance for each area, and uses the obtained object distance information and position information to estimate the main subject. Specify the position to be performed.

JP 2010-66378 A JP 11-25263 A

  However, since the detection accuracy of the main subject by multi-point distance measurement is limited to the sampling interval level of the area to be measured, there is a drawback that it is difficult to accurately detect the subject position.

  The present invention has been made in view of such circumstances, and an object thereof is to provide an imaging apparatus and a main subject detection method that can detect the position of a main subject with high accuracy.

In order to achieve the above object, the invention according to claim 1 divides an imaging optical system, an imaging unit that captures an image through the imaging optical system, and an imaging area of the imaging unit into a plurality of ranging areas. and a multi-point distance measuring means for measuring a subject distance in each ranging area based on the measurement result of the multi-point distance measuring device, a distance measuring area as the main subject candidate area main subject is estimated to belong Main subject candidate area detection means to detect, wherein the subject distance of each distance measurement area is weighted with a greater weight for the distance measurement area closer to the center of the screen, and the subject distance of each distance measurement area is a greater weight for a shorter distance in and adds the weighted values, and to quantify the main subject likeness of each ranging area, a main object candidate area detection means for selecting a ranging area the number is maximized main object candidate area, before Detailed search range setting means for setting a specific range centered on the center of the main subject candidate area as a detailed search range, and an image captured by the imaging means is analyzed within the detailed search range to determine the position of the main subject. A main subject position detecting means for detecting the focus, a focus area setting means for setting a focus area at the position of the main subject detected by the main subject position detecting means, and a fixed range based on the in-focus position of the main subject candidate area In the focus area, the position at which the contrast of the image obtained from the imaging means is maximized is detected by setting a detection range in, and changing the focal position of the imaging optical system in the detection range. Ranging means for measuring the subject distance of the subject, the subject distance measured by the ranging means and the main subject Obtains the difference between the subject distance of the area, and further comprising a, a main object position correction means for correcting the center of the main subject candidate area the position of the main object when the difference exceeds the threshold value An imaging device is provided.

According to the present invention, first, the position of the main subject is roughly detected by multi-point distance measurement, and then the image analysis is performed based on the position of the main subject roughly detected to detect the position of the main subject. Is done. As a result, even when the main subject exists outside the center, the position of the main subject can be accurately detected. Further, the position of the main subject can be efficiently detected by first determining the position of the main subject roughly and then detecting in detail.
In addition, according to the present invention, the range for detecting the main subject in detail by image analysis based on the position of the main subject (position of the main subject candidate area) roughly obtained from the result of multipoint ranging (reference) Detailed search range) is set. As a result, the range of image analysis is limited, and the main subject can be detected efficiently.
Further, according to the present invention, a predetermined range centered on the position of the main subject (position of the main subject candidate area) roughly obtained from the result of multipoint ranging is set as the detailed search range. By limiting the detailed search range around the position of the main subject obtained roughly, the main subject can be detected efficiently.
Further, according to the present invention, when the detailed position of the main subject is detected by image analysis, a focus area is set at the detected position of the main subject, and so-called contrast-type distance measurement is performed. At this time, a detection range (a range in which the focus of the imaging optical system is changed) is set based on the in-focus position of the main subject candidate area, and a position where the contrast of the image is maximized is detected in the detection range. Then, the difference between the subject distance obtained from the distance measurement result and the subject distance of the main subject candidate area is obtained. As a result, when the obtained difference is equal to or smaller than a preset threshold value, the position of the main subject obtained by image analysis as it is is set as the position of the main subject. On the other hand, if the obtained value exceeds the threshold value, the center of the main subject candidate area is set as the position of the main subject. As a result, even if the position of the main subject obtained by image analysis deviates from the position of the main subject roughly obtained (position of the main subject candidate area), the position of the main subject is greatly deviated. Can be prevented.
According to a second aspect of the present invention, in order to achieve the object, the main subject position correcting means determines the position of the main subject even when the distance measuring means cannot detect the position where the contrast is maximum. The imaging apparatus according to claim 1, wherein the imaging apparatus is corrected to the center of the candidate area.
According to the present invention, the center of the main subject candidate area can be obtained even when the position where the contrast becomes maximum when the detailed position of the main subject obtained by image analysis is set as the focus area and the distance is measured by the contrast method is not detected. Is the position of the main subject. As a result, even if the position of the main subject obtained by image analysis deviates from the position of the main subject roughly obtained (position of the main subject candidate area), the position of the main subject is greatly deviated. Can be prevented.
In order to achieve the above object, the invention according to claim 3 divides an imaging optical system, an imaging unit that captures an image through the imaging optical system, and an imaging area of the imaging unit into a plurality of ranging areas. And a multi-point distance measuring means for measuring a subject distance in each distance measuring area, and a distance measuring area estimated to belong to the main subject is detected as a main subject candidate area based on the measurement result of the multi-point distance measuring means. The main subject candidate area detection means for detecting the subject distance of each distance measurement area with a greater weight for the distance measurement area closer to the center of the screen and the subject distance of each distance measurement area with a greater weight for the shorter distance. The main subject candidate area detecting means for adding the weighted value to the numerical value to determine the main subject-likeness of each distance measurement area and selecting the distance measurement area having the maximum value as the main subject candidate area; Detailed search range setting means for setting a specific range centered on the center of the main subject candidate area as a detailed search range, and an image captured by the imaging means is analyzed within the detailed search range to determine the position of the main subject. A main subject position detecting means for detecting the focus, a focus area setting means for setting a focus area at the position of the main subject detected by the main subject position detecting means, and a predetermined range before and after the focus position of the main subject candidate area. A distance measuring unit that detects a position where the contrast of an image obtained from the imaging unit is maximized in the focus area by changing a focal position of the imaging optical system, and measures a subject distance in the focus area; If the position where the contrast is maximum is not detected as a result of the measurement by the distance measuring means, To provide an imaging apparatus characterized by comprising a, a main object position correction means for correcting the center of the main subject candidate area the position of the main subject.
According to the present invention, first, the position of the main subject is roughly detected by multi-point distance measurement, and then the image analysis is performed based on the position of the main subject roughly detected to detect the position of the main subject. Is done. As a result, even when the main subject exists outside the center, the position of the main subject can be accurately detected. Further, the position of the main subject can be efficiently detected by first determining the position of the main subject roughly and then detecting in detail.
In addition, according to the present invention, the range for detecting the main subject in detail by image analysis based on the position of the main subject (position of the main subject candidate area) roughly obtained from the result of multipoint ranging (reference) Detailed search range) is set. As a result, the range of image analysis is limited, and the main subject can be detected efficiently.
Further, according to the present invention, a predetermined range centered on the position of the main subject (position of the main subject candidate area) roughly obtained from the result of multipoint ranging is set as the detailed search range. By limiting the detailed search range around the position of the main subject obtained roughly, the main subject can be detected efficiently.
Further, according to the present invention, when the detailed position of the main subject is detected by image analysis, a focus area is set at the detected position of the main subject, and so-called contrast-type distance measurement is performed. At this time, a detection range (a range in which the focus of the imaging optical system is changed) is set based on the in-focus position of the main subject candidate area, and a position where the contrast of the image is maximized is detected in the detection range. If the position where the contrast is maximized is not detected by this detection, the center of the main subject candidate area is set as the position of the main subject. As a result, even if the position of the main subject obtained by image analysis deviates from the position of the main subject roughly obtained (position of the main subject candidate area), the position of the main subject is greatly deviated. Can be prevented.

  Note that the subject distance information does not necessarily need to be the distance information itself, and can be obtained as long as the information indicating the positional relationship of the subject existing in the screen in the optical axis direction, that is, the amount of information corresponding to the subject distance can be acquired. Good. Therefore, for example, when the distance is measured by the contrast method, the position information of the focus lens when the contrast becomes maximum (when the focus evaluation value becomes maximum (peak)) or the contrast becomes maximum The amount of movement of the focus lens from the reference point (number of pulses when driving a pulse) can be used as subject distance information (amount information corresponding to subject distance). Can be estimated.

In the invention according to claim 2, in order to achieve the object , the main subject candidate area detecting means is configured so that the subject distance in each distance measuring area or the amount corresponding to the subject distance and the position in the screen are respectively 2. The imaging apparatus according to claim 1, wherein the main subject likelihood of each ranging area is digitized by weighting and adding, and the ranging area having the maximum numeric value is selected as the main subject candidate area. provide.
In order to achieve the above object, the main subject candidate area detecting means has two main subjects: “the closer to the center of the screen, the more likely the main subject” and “the closer the distance, the more likely the main subject”. Based on the hypothesis, the object distance of each distance measurement area or the amount corresponding to the object distance and the position in the screen are weighted and added, respectively, to quantify the main object-likeness of each distance measurement area, 3. The imaging apparatus according to claim 2, wherein a distance measuring area having a maximum numerical value is selected as a main subject candidate area.
According to a fourth aspect of the present invention, in order to achieve the above object, the main subject position detecting means detects a position where the main subject is most likely as the position of the main subject within the detailed search range. An imaging apparatus according to any one of 3 is provided.
In the invention according to claim 5, in order to achieve the object, the main subject position detecting means searches the detailed search range from the center toward the outside, and the evaluation value of the main subjectness exceeds a threshold value. The imaging device according to any one of claims 1 to 3, wherein a position at a certain time point is detected as a position of a main subject.

The invention according to claim 6, in order to achieve the above object, the detailed search range setting means, when comparing the object distance of each ranging area obtained by the multi-point distance measuring means, the primary subject difference of the subject distance Metropolitan candidate areas by excluding or more ranging areas threshold, the imaging apparatus according to any one of claims 1 to 5, characterized in that setting the fine search range provide.

  When detecting the position of the main subject in detail within the detailed search range, for example, if the position of the main subject is detected using the number of edge components as an index, the position of the background object is determined by the background edge component. May be recognized. Therefore, by using the result of multipoint distance measurement, the subject distance of the main subject candidate area and the area with a far subject distance are excluded from the detailed search range. Thereby, the position of the main subject can be detected more accurately.

In the invention according to claim 7, in order to achieve the object, the multipoint ranging means changes the focal position of the imaging optical system, and the image obtained from the imaging means for each ranging area. by detecting the position at which the contrast becomes the maximum, the measured subject distance in each ranging area, the detailed search range setting means, a distance measurement area whose position is detected two or more of the contrast becomes the maximum The imaging apparatus according to claim 1, wherein the detailed search range is set by excluding the imaging apparatus.

  When multipoint distance measurement is performed using a so-called contrast method, two or more positions where the contrast becomes maximum may be detected. For example, when the position of the main subject is detected using the number of edge components as an index in an area in which two or more positions where the contrast is maximized is detected, the background edge component causes the background object to be detected. The position may be recognized as the main subject. Therefore, an area in which two or more positions where the contrast becomes maximum is detected as a result of multipoint distance measurement is excluded from the detailed search range. Thereby, the position of the main subject can be detected more accurately.

The invention according to claim 8, in order to achieve the above object, the detailed search range setting means may be set by scaling the magnitude of the detailed search range according to the object distance of the main subject candidate areas An imaging device according to any one of claims 1 to 5 is provided.

  According to the present invention, the size of the detailed search range is set to be enlarged or reduced according to the subject distance of the main subject candidate area. In general, it is considered that the main subject is shown larger on the screen as its position is closer to the imaging device. Therefore, the detail search range is set to be larger as the subject distance of the main subject candidate area is closer. That is, the detailed search range is set according to the size of the main subject shown in the screen. As a result, an accurate and lean detailed search can be performed.

According to a ninth aspect of the invention, in order to achieve the object, the imaging optical system includes a variable magnification optical system, and the detailed search range setting means is configured to perform the detailed search according to a focal length of the imaging optical system. It is set by scaling a range of a size to provide an imaging apparatus according to any one of claims 1 to 5, characterized in.

  According to the present invention, the size of the detailed search range is set to be enlarged or reduced in accordance with the focal length (zoom magnification) of the main subject candidate area. In general, it is considered that the main subject is shown larger on the screen as the zoom is set to the tele side (telephoto side). Accordingly, the size of the detailed search range is set to be larger as the focal length is longer (zoom is on the telephoto side (telephoto side)). That is, the detailed search range is set according to the size of the main subject shown in the screen. As a result, an accurate and lean detailed search can be performed.

According to a tenth aspect of the present invention, in order to achieve the object, the main subject position detecting means uses at least one of information on the number of edge components or the size of a lump of the same color system as an image characteristic. Then, the main subject is detected, and the imaging apparatus according to any one of claims 1 to 9 is provided.

  According to the present invention, the main subject is detected using information on at least one of the number of edge components and the size of the same color system as the image characteristics. That is, the number of edge components, the size of the same color system, and the like are used as indices for evaluating the main subjectness.

The invention according to claim 11, in order to achieve the object, dividing the imaging region of the imaging means for capturing an image into a plurality of distance measurement areas through the imaging optical system, the object distance in each ranging area A step of measuring multiple points, and a step of detecting, as a main subject candidate area, a range finding area estimated to belong to a main subject based on the result of the multipoint ranging, Add the value weighted with greater weight to the distance measurement area closer to the center of the screen and the value weighted with greater weight to the subject distance in each distance measurement area. Quantify the subject likeness, select the distance measurement area with the maximum value as the main subject candidate area, and set a certain range centered on the center of the main subject candidate area as the detailed search range That the method analyzes the image captured in the detailed search range by the imaging means, and detecting a position of the main subject, and setting a focus area on the detected position of the main subject, the By setting a detection range within a certain range with reference to the in-focus position of the main subject candidate area, and changing the focal position of the imaging optical system within the detection range, the image obtained from the imaging means in the focus area The step of detecting the position where the contrast is maximized and measuring the subject distance of the focus area and the difference between the measured subject distance and the subject distance of the main subject candidate area are obtained, and the difference exceeds the threshold value. and correcting the position of the main object in the center of the main subject candidate area when mainly characterized in that it consists Providing an object detection method.
In order to achieve the above object, the invention according to claim 12 divides the imaging area of the imaging means for imaging an image via an imaging optical system into a plurality of ranging areas, and measures the subject distance in each ranging area. A step of performing multipoint ranging, and a step of detecting, as a main subject candidate area, a ranging area estimated to belong to a main subject based on a result of the multipoint ranging, wherein the subject of each ranging area Add the value weighted with a greater weight to the distance measurement area closer to the center of the screen and the value weighted with the greater weight the subject distance of each distance measurement area to the main subject of each distance measurement area. Quantifying the likelihood, selecting the ranging area with the maximum numerical value as the main subject candidate area, and setting a certain range centered on the center of the main subject candidate area as the detailed search range Analyzing the image captured by the imaging means within the detailed search range to detect the position of the main subject; setting the focus area at the detected position of the main subject; By changing the focal position of the imaging optical system within a predetermined range before and after the in-focus position of the main subject candidate area, the position where the contrast of the image obtained from the imaging means is maximized is detected in the focus area, Measuring the subject distance of the focus area, and correcting the position of the main subject to the center of the main subject candidate area when the position where the contrast is maximized is not detected as a result of the measurement. A main subject detection method is provided.

  According to the present invention, the position of the main subject can be detected with high accuracy even when the main subject exists outside the center.

1 is an external view showing an example of a digital camera to which the present invention is applied. 1 is an external view showing an example of a digital camera to which the present invention is applied. Block diagram showing functional configuration of digital camera Block diagram showing functional configuration of AF processing unit Figure showing an example of dividing the ranging area Conceptual diagram of continuous AF Conceptual diagram of main subject detection processing by the first method The flowchart which shows the procedure of the process which detects the main subject by the 1st method. The flowchart which shows the procedure of the process which detects a main subject by the 2nd method. Explanatory drawing when detecting the main subject using the number of edge components as an index The flowchart which shows the procedure of the process which detects a main subject by the 3rd method. Explanatory drawing when a main subject is detected by the third method Explanatory drawing when focus evaluation value has double peak for distance information The flowchart which shows the procedure of the process which detects the main subject by the 4th method. Explanatory drawing when a main subject is detected by the fourth method The flowchart which shows the procedure of the process which detects the main subject by the 5th method. Explanatory drawing when a main subject is detected by the fifth method Explanatory drawing when a main subject is detected by the sixth method The flowchart which shows the procedure of the process which detects a main subject by the 6th method. Explanatory drawing when a main subject is detected by the seventh method The flowchart which shows the procedure of the process which detects a main subject by the 7th method. The flowchart which shows the procedure of the process which detects the main subject by the 8th method. Explanatory drawing when a main subject is detected by the eighth method State transition diagram of continuous AF performed before actual shooting Flowchart showing the procedure of continuous AF processing performed before the main photographing Illustration of monitor display example when main subject is detected Flow chart showing the procedure for setting the focus area for the main subject detected during live view Flow chart showing the procedure of processing up to the actual shooting Illustration of automatic tracking function Flow chart showing the procedure of processing for setting the position of the main subject detected during live view as the target of automatic tracking

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings.

≪Overall structure≫
1 and 2 are external views showing an example of a digital camera to which the present invention is applied.

  As shown in the figure, the digital camera 10 is a so-called compact type digital camera.

  On the front surface of the camera body 12 formed in a rectangular box shape with a small thickness in the front-rear direction, a photographing lens 14, a flash 16, and the like are arranged. Also, a release button 18, a zoom lever 20, a power button 22 and the like are arranged on the top surface, and a monitor 24, a menu / OK button 26, a cross button 28, a Back button 30, a playback button 32 and the like are arranged on the back surface. Has been placed.

  The photographic lens 14 is a retractable zoom lens, and is fed out from the camera body 12 when the digital camera 10 is turned on.

  The flash 16 is composed of, for example, a xenon tube, and emits light as necessary.

  The release button 18 is constituted by a two-stage press button that can perform so-called “half press” and “full press”. When the release button 18 is half-pressed, a shooting preparation instruction (S1) is input to the digital camera 10, and when it is fully pressed, a main shooting instruction (S2) is input.

  The zoom lever 20 is provided so as to be swingable left and right around the release button 18. When the zoom lever 20 is operated in one direction, a zoom instruction in the tele direction is input to the digital camera 10, and when it is operated in the other direction, a zoom instruction in the wide direction is input.

  The power button 22 is a push button. When the power button 22 is pressed while the digital camera 10 is turned off, the digital camera 10 is turned on. In addition, when the power button 22 is pressed for a predetermined time while the power of the digital camera 10 is on, the power of the digital camera 10 is turned off.

  The monitor 24 is composed of a liquid crystal display, for example. Since the digital camera 10 of the present embodiment does not include an optical viewfinder, the monitor 24 is used as an electronic viewfinder (live view monitor). That is, at the time of photographing, an image captured by the image sensor is displayed on the monitor 24 in real time. The photographer looks at an image (through image) displayed in real time on the monitor 24 and confirms the focus state, composition, and the like. The monitor 24 is used for reproducing a captured image (photographed image), and is used as a setting screen for various settings of the digital camera 10.

  The menu / OK button 26 functions as an operation button for instructing to call a menu screen and also functions as a button for instructing confirmation.

  The cross button 28 is provided so that it can be pressed in four directions, up, down, left, and right, and a function corresponding to the state of the digital camera 10 at that time is assigned. For example, in the shooting mode, a macro function switching is assigned to the right button, and a strobe mode switching function is assigned to the left button. In the playback mode, the right button is assigned a frame advance function, and the left button is assigned a frame reverse function.

  The Back button 30 functions as a button for instructing cancellation of input operation information and the like, and the playback button 32 functions as a button for instructing switching to the playback mode.

  FIG. 3 is a block diagram illustrating a functional configuration of the digital camera 10.

  As shown in the figure, the digital camera 10 includes a photographing lens 14, a lens driving unit 50, a flash 16, a flash control unit 52, an image sensor 54, a CCD control unit 56, an analog signal processing unit 58, a timing generator 60, an image input. Controller 62, AF processing unit 64, AE / AWB processing unit 66, image processing unit 68, face detection unit 70, compression / decompression processing unit 72, frame memory 74, media control unit 76, recording medium 78, display control unit 80, Monitor 24, operation unit (release button 18, zoom lever 20, power button 22, menu / OK button 26, cross button 28, back button 30, play button 32, etc.) 82, operation system control unit 84, main subject candidate area detection Unit 86, main subject detection unit 88, CPU 90, EEPROM 92, and the like. .

  The photographing lens 14 includes a focus lens 110, a zoom lens 112, a diaphragm 114, and a shutter 116. The focus lens 110 and the zoom lens 112 are provided to be movable back and forth along the optical axis. The focus adjustment (focusing) is performed by moving the focus lens 110, and the zooming adjustment (zooming) is performed by moving the zoom lens 112.

  The diaphragm 114 is composed of, for example, a luminous diaphragm, and adjusts the amount of light incident on the image sensor 54.

  The shutter 116 is configured by a mechanical shutter, and blocks light from entering the image sensor 54.

  The lens driving unit 50 includes a focus lens driving unit 120 that drives the focus lens 110, a zoom lens driving unit 122 that drives the zoom lens 112, an aperture driving unit 124 that drives the diaphragm 114, and a shutter driving unit 126 that drives the shutter 116. Consists of.

  The focus lens driving unit 120 includes a motor and a motor driver, and step-drives the focus lens 110 in accordance with a command from the CPU 90 to move the focus lens 110 back and forth along the optical axis.

  The zoom lens driving unit 122 includes a motor and a motor driver. The zoom lens driving unit 122 drives the zoom lens 112 stepwise in accordance with a command from the CPU 90 and moves the zoom lens 112 back and forth along the optical axis.

  The aperture driving unit 124 includes a motor and a motor driver, and step-drives the aperture 114 in accordance with a command from the CPU 90 to expand or contract the aperture diameter.

  The shutter driving unit 126 includes a motor and a motor driver, and opens and closes the shutter 116 according to a command from the CPU 90.

  The flash controller 52 controls the light emission operation of the flash 16 in accordance with a command from the CPU 90.

  The imaging element 54 is disposed behind the photographing lens 14 and captures an image captured by the photographing lens 14. The image sensor 54 is constituted by, for example, a color CCD having a predetermined color filter array (hereinafter referred to as a CCD). The CCD 54 outputs the charges accumulated in each pixel as serial analog image data line by line in synchronization with the vertical transfer clock signal and horizontal transfer clock signal supplied from the CCD controller 56.

  The CCD controller 56 controls the drive of the CCD 54 in accordance with a command from the CPU 90.

  The analog signal processing unit 58 takes in analog image data output from the CCD 54, performs necessary analog signal processing (correlated double sampling processing, gain adjustment, etc.), and then performs digital image data (RGB for each pixel). CCD-RAW data having density values) and output.

  The timing generator 60 generates a predetermined timing signal in response to a command from the CPU 90 and outputs it to the shutter driving unit 126, the CCD control unit 56, and the analog signal processing unit 58. Processing of each unit is synchronized by a timing signal output from the timing generator 60.

  The image input controller 62 takes in digital image data (CCD-RAW data) output from the analog signal processing unit 58 and writes it into the frame memory 74.

  The frame memory 74 is used as a storage area for image data (CCD-RAW data) captured from the CCD 54, and is also used as a work area when performing various processes.

  The AF processing unit 64 detects a focal position based on image data (CCD-RAW data) captured by the CCD 54. The AF processing unit 64 will be described in detail later.

  The AE / AWB processing unit 66 measures subject brightness based on image data (CCD-RAW data) captured by the CCD 54, and determines an exposure value (aperture value, shutter speed, ISO sensitivity, etc.). Further, based on the image captured by the CCD 54, the light source type is detected and the white balance correction value is determined.

  The image processing unit 68 performs necessary digital signal processing on the image captured by the CCD 54 (CCD-RAW data) to generate YCrCb image data composed of luminance data and color difference data. The generated YCrCb image data is stored in the frame memory 74.

  The face detection unit 70 analyzes the YCrCb image data generated by the image processing unit 68 and detects a human face area.

  The compression / decompression processing unit 72 performs predetermined compression processing (for example, compression processing by the JPEG method) on the YCrCb image data generated by the image processing unit 68 to generate compressed image data. The compressed image data is subjected to a predetermined expansion process to generate non-compressed image data (YCrCb image data).

  The media control unit 76 reads and writes data from and to the recording medium 78 in response to a command from the CPU 90.

  The recording medium 78 is composed of, for example, a memory card, and is detachably attached to a card slot provided in the camera body 12. The recording medium 78 can also be built in the camera body 12.

  The display control unit 80 controls display on the monitor 24 in response to a command from the CPU 90. As described above, in the digital camera 10 of the present embodiment, a through image is displayed on the monitor 24 at the time of shooting. The display control unit 80 takes in the YCrCb image data taken from the CCD 54 and subjected to necessary signal processing, converts it into a display signal format, and outputs it to the monitor 24. As a result, a through image is displayed on the monitor 24. At the time of reproduction, the image data read from the recording medium 78 and subjected to a predetermined expansion process is taken in, converted into a display signal format, and output to the monitor 24. As a result, the captured image is displayed on the monitor 24.

  The operation system control unit 84 is an operation signal corresponding to the operation content of the operation unit (release button 18, zoom lever 20, power button 22, menu / OK button 26, cross button 28, back button 30, play button 32, etc.). Is output to the CPU 90.

  The main subject candidate area detection unit 86 detects, as a main subject candidate area, a distance measurement area estimated to belong to the main subject based on the result of multipoint ranging. Specifically, based on the two hypotheses: “The closer to the center of the screen, the more likely it is to be the main subject” and “The closer the distance is, the more likely the subject is to be the main subject”. , Each is weighted and added, and the likelihood of the main subject is digitized. Then, the area having the maximum value is recognized as the area to which the main subject belongs.

  The main subject detection unit 88 sets a detailed search range based on the position of the ranging area detected as the main subject candidate area, analyzes an image obtained from the CCD 54 in the detailed search range, and detects a main subject from the image. To detect. Specifically, the main subject is detected by image analysis using the number of edge components and the size of the same color system as an index for evaluating the likelihood of the main subject.

  The procedure of main subject detection processing by the main subject detection unit 88 will be described in detail later.

  The CPU 90 is a control unit that performs overall control of the operation of the digital camera 10 and controls each unit of the digital camera 10 according to a predetermined control program. Further, predetermined calculation processing is executed according to a predetermined calculation processing program.

  The EEPROM 92 stores various programs executed by the CPU 90, data necessary for control, and the like.

≪Basic operation≫
Next, basic shooting and playback operations of the digital camera 10 will be described.

  When the power button 22 is pressed while the power is off, the digital camera 10 is activated in the shooting mode. Further, when the playback button 32 is pressed for a certain period of time while the power is off, the playback mode is activated.

  When the digital camera 10 activated in the shooting mode is pressed in the shooting mode, the digital camera 10 shifts to the playback mode. If the release button 18 is pressed halfway during the playback mode, the mode is changed to the shooting mode.

  When the shooting mode is set, a through image display process is started. The through image is displayed by continuously capturing images at a constant frame rate. The continuously captured images are continuously processed and output to the monitor 24, whereby the image captured by the CCD 54 is displayed on the monitor 24 in real time. The photographer looks at the through image displayed in real time on the monitor 24 and confirms the focus state, composition, and the like.

  The main shooting (shooting for recording an image on the recording medium 78) is performed by first pressing the release button 18 halfway and then pressing the release button 18 fully.

  When the release button 18 is pressed halfway, preparation for shooting is performed, and AE, AF, and AWB processes are performed. That is, the subject brightness is measured, the exposure value (aperture value, shutter speed, ISO sensitivity, etc.) is determined, the light source type is detected, and the white balance correction value is determined. In addition, focus adjustment is performed so as to focus on the main subject.

  After the preparation for shooting, when the release button 18 is fully pressed, actual shooting is executed. That is, recording for recording is performed with the exposure value determined by the AE process. The obtained image data is subjected to required signal processing and recorded on the recording medium 78.

  The image data recorded on the recording medium 78 can be reproduced and displayed on the monitor 24 by setting the mode of the digital camera 10 to the reproduction mode.

  When the reproduction mode is set, the image data recorded last on the recording medium 78 is read out, subjected to necessary signal processing, and reproduced and displayed on the monitor 24. Thereafter, when an operation for frame advance / frame return is performed on the operation unit 82, captured images are reproduced and displayed in order according to the operations.

≪Configuration of AF processing section≫
The AF processing unit 64 performs necessary signal processing on the image data (CCD-RAW data) captured from the image input controller 62 to calculate a focus evaluation value.

  FIG. 4 is a block diagram illustrating a functional configuration of the AF processing unit. As shown in the figure, the AF processing unit 64 includes a high pass filter (HPF) 64A, a gate 64B, and an adder 64C.

  The high pass filter 64A extracts high frequency components included in the image data. The higher the image sharpness (contrast), the more high frequency components of the image are contained in the luminance signal. Therefore, by integrating this high frequency component, it is possible to quantify the level of average image sharpness in that range. The numerical value of the sharpness of this image is called a focus evaluation value.

  The image data from which the high frequency component has been extracted by the high pass filter 64A is input to the gate 64B. The gate 64B divides the image data into vertical and horizontal n × m and outputs it to the adder 64C. In this example, as shown in FIG. 5, it is divided into 5 × 7 vertical and horizontal and output to the adder 64C.

  The adder 64C integrates high-frequency components of the image data for each divided area, and calculates a focus evaluation value for each area.

  Information on the calculated focus evaluation value for each area is output to the CPU 90. The CPU 90 executes focus adjustment and main subject recognition processing based on the focus evaluation value information for each area.

≪Focus adjustment≫
First, the focus adjustment operation will be described.

  There are one-shot AF and continuous AF as focus adjustment methods. In the one-shot AF, a search operation is performed only once so as to focus on a subject in a specific focus area, a focus position is detected, and focusing is performed. On the other hand, the continuous AF repeatedly performs focus adjustment so as to keep focusing on a subject in a specific focus area.

<One-shot AF>
First, one-shot AF will be described.

  As described above, the one-shot AF detects a focus position by performing a search operation only once so that a subject in a specific focus area is focused.

  When a focusing command (for example, half-pressing the release button) is input, the CPU 90 moves the focus lens 110 from the closest end to the infinity end. Images are sequentially captured while the focus lens 110 moves from the closest end to the infinite end, and the AF processing unit 64 sequentially calculates focus evaluation values.

  The CPU 90 acquires information on focus evaluation values sequentially calculated by the AF processing unit 64.

  Here, since the information on the focus evaluation value acquired from the AF processing unit 64 is information on the focus evaluation value for each area, the CPU 90 is set in advance from the obtained information on the focus evaluation value for each area. Information on the focus evaluation value of the area corresponding to the focus area is extracted. And the position where the focus evaluation value of the extracted focus area becomes maximum (peak) is detected.

  As described above, since the focus evaluation value is a numerical value of the sharpness (contrast) of the image, the position where the value is maximized is the position where the subject image is most sharply captured and is in focus. It can be regarded as a position. Therefore, the CPU 90 detects a position where the focus evaluation value is maximized in the range from the closest end to the infinity end (focus evaluation value search range), and sets this as the focus position.

  The CPU 90 moves the focus lens 110 to the position detected as the in-focus position, and ends the focus adjustment process.

  As described above, in the one-shot AF, a position where the focus evaluation value is maximized is detected by one search, and the focus lens 110 is moved to the detected position to perform focusing.

<Continuous AF>
In continuous AF, focus adjustment is repeatedly performed so that a subject in a specific focus area continues to be focused.

  When the focusing command is input, the CPU 90 sets the focus lens 110 to the in-focus position by hill climbing control. That is, as shown in FIG. 6, the focus lens 110 is slightly moved back and forth (wobbling) to check the increase / decrease direction of the focus evaluation value in a specific focus area, and the focus lens 110 reaches a position where the focus evaluation value becomes maximum. Move gradually. The focus lens 110 is continuously wobbled after the position at which the focus evaluation value is maximized, the increase / decrease of the focus evaluation value is checked, the position at which the focus evaluation value is maximized is detected, and the focus is set at that position. The lens 110 is moved.

  As described above, in the continuous AF, the focus state is constantly checked even after the focus position is detected, so that the focus state is maintained.

≪Detection of main subject≫
<First method>
FIG. 7 is a conceptual diagram of main subject detection processing according to the first method.

  In this example, first, multipoint distance measurement is performed, and using the measurement result, a distance measurement area that is estimated to belong to the main subject is specified as a main subject candidate area ((a) in the figure). Next, the center position of the main subject candidate area is set to the main subject rough position ((a) in the figure). Next, a predetermined range centered on the rough position of the main subject is set as a detailed search area ((b) in the figure). Next, image analysis is performed within the detailed search range to detect a main subject ((c) in the figure).

  As described above, in the present embodiment, first, the rough position of the main subject is detected from the result of multipoint ranging, and the position of the main subject is determined in detail by image analysis based on the detected rough position. To detect.

  Thereby, it is possible to efficiently detect the accurate position of the main subject.

  FIG. 8 is a flowchart showing a procedure of processing for detecting a main subject by this method (first method).

  First, multipoint ranging is performed (step S10). Then, based on the result of the multipoint distance measurement, it is determined whether or not there is a distance measurement area where the position where the focus evaluation value is maximized (peak) is detected (step S11).

  When there is no distance measurement area where the position where the focus evaluation value is maximized is detected (including a case where a peak greater than a preset threshold is not detected), the main subject candidate detection process cannot be performed. Therefore, the process ends.

  On the other hand, if there is a ranging area where the position where the focus evaluation value is maximized is detected, a main subject candidate area detection process is performed based on the detection result of the multipoint ranging (step S12). This processing is performed by the main subject candidate area detection unit 86.

  The main subject candidate area detection unit 86 acquires the result of multipoint distance measurement, performs predetermined calculation processing, and determines the main subject candidate area. Specifically, based on the two hypotheses: “The closer to the center of the screen, the more likely it is to be the main subject” and “The closer the distance is, the more likely the subject is to be the main subject”. , Each is weighted and added, and the likelihood of the main subject is digitized. Then, the distance measuring area having the maximum value is selected as the main subject candidate area.

  When the main subject candidate area is detected, the information is output to the main subject detection unit 88. The main subject detection unit 88 sets the center of the obtained main subject candidate area as the main subject rough position (step S13).

  Then, a detailed search range is set within a certain range around the rough position of the main subject (step S14). This detailed search range is a range for detecting the main subject in detail by image analysis. In the example shown in FIG. 7B, a fixed rectangular area around the main subject rough position is set as the detailed search range. ing.

  When the detailed search range is set, the main subject detection unit 88 analyzes the image obtained from the CCD 54 within the set detailed search range, and detects the main subject (step S15). At this time, the main subject is detected from the image using the number of edge components, the size of the same color cluster, and the like as an index for evaluating the likelihood of the main subject.

  Then, the detected position of the main subject is determined as the main subject detail position (step S16).

  In this way, in this method, the rough position of the main subject is determined from the result of multipoint ranging, and the main subject is detected in detail by image analysis based on the determined rough position of the main subject. Thereby, the position of the main subject can be accurately detected.

  The method of finally determining the position of the main subject within the detailed search range is roughly divided into the following two methods.

(I) Maximum value priority: A method in which the main subject is located in the main subject detail search range where the main subject is the largest. (Ii) Center priority: Search in the main subject detail search range from the center toward the outside. The method of setting the position of the main subject to the position at which the evaluation value of the main subjectness exceeds the reference value (threshold) so that the designer can select whether to use the maximum priority or the center priority. It is preferable to do.

  In this example, the multipoint distance measurement is performed by a so-called contrast method, but the distance measurement method is not limited to this. In addition, for example, it is possible to adopt a configuration in which distance measurement is performed by a phase difference method or an infrared method.

  In addition, the subject distance information obtained by distance measurement does not necessarily need to be the distance information itself, but is information that shows the positional relationship of the subject existing in the screen in the optical axis direction, that is, an amount corresponding to the subject distance. I just need it. Therefore, for example, when the distance is measured by the contrast method as in this example, the position information of the focus lens when the contrast is maximized (when the focus evaluation value is maximized (peak)), and the contrast is The amount of movement of the focus lens from the reference point at the time of local maximum (the number of pulses during pulse driving) or the like can be used as subject distance information (amount information corresponding to the subject distance).

<Second method>
FIG. 9 is a flowchart showing a procedure of processing for detecting a main subject by the second method.

  In this method, first, a face is detected from an image obtained from the CCD 54 (step S1). Then, the presence / absence of a face on the screen is determined (step S2), and the main subject detection process described in the first method is performed only when there is no face on the screen (steps S10 to S16). That is, when a face is detected, the detected face is set as a main subject (step S3), and the detected face is set as a main subject detailed position.

  In general, when a person is photographed on the screen, the person is considered to be a main subject, so the face is set as the main subject with priority. Thereby, the main subject can be detected without waste.

<Third method>
When the main subject is detected in detail by image analysis, if the main subject is determined using the number of edge components as an index, the position of the background object may be recognized as the main subject by the background edge component. Yes (see FIG. 10).

  Therefore, a distance map based on the subject distance of the main subject candidate area is created using the result of multipoint ranging. When setting the detailed search range, the area close to the subject distance of the main subject candidate area is set as the target of the detailed search range, but the area far away is excluded from the target and the detailed search range is limited.

  Thereby, when detecting a main subject by image analysis, it can detect more correctly. Further, since the detailed search range is limited, detection can be performed more efficiently.

  FIG. 11 is a flowchart showing a procedure of processing for detecting a main subject by this method (third method).

  First, multipoint ranging is performed (step S20). Then, based on the result of the multipoint distance measurement, it is determined whether or not there is a distance measurement area where the position where the focus evaluation value is maximized (peak) is detected (step S21).

  If there is no distance measurement area where the position where the focus evaluation value is maximized is detected, the main subject candidate detection process cannot be performed, and the process ends.

  On the other hand, if there is a ranging area where the position where the focus evaluation value is maximized is detected, the main subject candidate area detection process is performed based on the detection result of the multipoint ranging (step S22). This processing is performed by the main subject candidate area detection unit 86.

  When the main subject candidate area is detected, the information is output to the main subject detection unit 88. The main subject detection unit 88 sets the center of the obtained main subject candidate area as the main subject rough position (step S23) (see FIG. 12A).

  Next, based on the result of the multipoint distance measurement, the main subject detection unit 88 uses a distance map based on the subject distance of the main subject candidate area (each distance measurement based on the subject distance of the main subject candidate area as a reference (0)). An area distance map) is created (step S24).

  In FIG. 12, the shade of the color represents the distance from the subject in the main subject candidate area (the color becomes darker as the focus evaluation value reaches the peak at the infinity side, and the color moves away from the closest side. It is shown to be thin).

  When the distance map is created, the detailed search range is determined by the main subject detection unit 88 based on the distance map and the rough position of the main subject (step S25).

  First, the main subject detection unit 88 sets a temporary detailed search range (temporary detailed search range) within a certain range around the main subject rough position (see FIG. 12B). Then, the detailed search range is set by excluding an area having a subject distance far away from the subject distance of the main subject candidate area in the temporary detailed search range. That is, the detailed search range is set by excluding a region where the difference (absolute value) from the subject distance of the main subject candidate area exceeds a preset threshold (see FIG. 12C).

  When the detailed search range is set, the main subject detection unit 88 analyzes the image obtained from the CCD 54 within the set detailed search range, and detects the main subject (step S26). At this time, the main subject is detected from the image using the number of edge components, the size of the same color cluster, and the like as an index for evaluating the likelihood of the main subject.

  Then, the detected position of the main subject is determined as the main subject detail position (step S27) (see FIG. 12D).

  Thus, in this method, when the detailed search range is set, the detailed search range is set by excluding an area having a subject distance far away from the subject distance of the main subject candidate area. As a result, it is possible to prevent the background object from being recognized as the main subject due to the edge component of the background, and to detect the main subject more accurately. Further, by limiting the search range, it is possible to detect the main subject more efficiently.

  In this method as well, it is possible to detect a main subject by this method only when a face is detected first and no face is detected (see the second method).

  In this example, the multipoint distance measurement is executed by the so-called contrast method, but the distance measurement method is not limited to this. In addition, for example, it is possible to adopt a configuration in which distance measurement is performed by a phase difference method or an infrared method.

<Fourth method>
As described above, distance information of each ranging area can be obtained by performing contrast-based multipoint ranging. At this time, when the focus evaluation value has a double peak with respect to the distance information (see FIG. 13), there is a possibility that the side effect is caused by the side effect of the detailed search.

  Therefore, in this method, when the detailed search range is set, the area where the double peak is present and its vicinity are excluded from the target of the detailed search range, and the main subject is detected in detail by image analysis. Reduce side effects.

  FIG. 14 is a flowchart showing a procedure of processing for detecting a main subject by this method (fourth method).

  First, multipoint ranging is performed (step S30). Then, based on the result of the multipoint distance measurement, it is determined whether or not there is a distance measurement area where the position where the focus evaluation value is maximized (peak) is detected (step S31).

  If there is no distance measurement area where the position where the focus evaluation value is maximized is detected, the main subject candidate detection process cannot be performed, and the process ends.

  On the other hand, if there is a ranging area where the position where the focus evaluation value is maximized is detected, the main subject candidate area detection process is performed based on the detection result of the multipoint ranging (step S32). This processing is performed by the main subject candidate area detection unit 86.

  When the main subject candidate area is detected, the information is output to the main subject detection unit 88. The main subject detection unit 88 sets the center of the obtained main subject candidate area as the main subject rough position (step S33) (see FIG. 15A).

  Next, based on the result of the multipoint distance measurement, the main subject detection unit 88 detects a distance measurement area (double peak distance measurement area) in which the focus evaluation value reaches a maximum (peak) at two or more positions. (Step S34).

  In FIG. 15, the double peak ranging area is filled in.

  Next, based on the detected information on the double peak ranging area and the information on the main subject rough position, the main subject detection unit 88 determines the detailed search range (step S35).

  First, the main subject detection unit 88 sets a temporary detailed search range (temporary detailed search range) within a certain range around the main subject rough position (see FIG. 15B). Then, the detailed search range is set by excluding the double peak ranging area within the temporary detailed search range (see FIG. 15C).

  When the detailed search range is set, the main subject detection unit 88 analyzes the image obtained from the CCD 54 within the set detailed search range, and detects the main subject (step S36). At this time, the main subject is detected from the image using the number of edge components, the size of the same color cluster, and the like as an index for evaluating the likelihood of the main subject.

  Then, the detected position of the main subject is determined as the main subject detail position (step S37) (see FIG. 15D).

  Thus, in this method, when setting the detailed search range, the detailed search range is set by excluding the double peak region. As a result, it is possible to prevent the background object from being recognized as the main subject due to the edge component of the background, and to detect the main subject more accurately. Further, by limiting the search range, it is possible to detect the main subject more efficiently.

  In this method as well, it is possible to detect a main subject by this method only when a face is detected first and no face is detected (see the second method).

<Fifth method>
In this method, an area having a subject distance far away from the subject distance of the main subject candidate area is excluded from the detailed search target, and an area having a double peak is excluded from the detailed search target.

  FIG. 16 is a flowchart showing a processing procedure for detecting a main subject by this method (fifth method).

  First, multipoint ranging is performed (step S40). Then, based on the result of the multipoint distance measurement, it is determined whether or not there is a distance measurement area where the position where the focus evaluation value is maximized (peak) is detected (step S41).

  If there is no distance measurement area where the position where the focus evaluation value is maximized is detected, the main subject candidate detection process cannot be performed, and the process ends.

  On the other hand, if there is a ranging area where the position where the focus evaluation value is maximized is detected, the main subject candidate area detection process is performed based on the detection result of the multipoint ranging (step S42). This processing is performed by the main subject candidate area detection unit 86.

  When the main subject candidate area is detected, the information is output to the main subject detection unit 88. The main subject detection unit 88 sets the center of the obtained main subject candidate area as the main subject rough position (step S43) (see FIG. 17A).

  Next, based on the result of the multipoint distance measurement, the main subject detection unit 88 uses a distance map based on the subject distance of the main subject candidate area (each distance measurement based on the subject distance of the main subject candidate area as a reference (0)). An area distance map) is created (step S44).

  Next, based on the result of the multipoint distance measurement, the main subject detection unit 88 detects a distance measurement area (double peak distance measurement area) in which the focus evaluation value reaches a maximum (peak) at two or more positions. (Step S45).

  Next, based on the distance map information, the double peak ranging area information, and the main subject rough position information, the main subject detection unit 88 determines a detailed search range (step S46).

  First, the main subject detection unit 88 sets a temporary detailed search range (temporary detailed search range) within a certain range around the main subject rough position (see FIG. 17B). Then, within the temporary detailed search range, an area having a subject distance that is far away from the subject distance of the main subject candidate area (a detailed search range is set. That is, the difference (absolute value) from the subject distance of the main subject candidate area is , An area exceeding a preset threshold value) is excluded, and a double peak ranging area is excluded to set a detailed search range (see FIG. 17C).

  In FIG. 17C, the area where the double peak is detected is represented by a diagonal lattice, and the area far from the main subject rough position is represented by a lattice.

  When the detailed search range is set, the main subject detection unit 88 analyzes the image obtained from the CCD 54 within the set detailed search range, and detects the main subject (step S46). At this time, the main subject is detected from the image using the number of edge components, the size of the same color cluster, and the like as an index for evaluating the likelihood of the main subject.

  Then, the detected position of the main subject is determined as the main subject detail position (step S47) (see FIG. 17D).

  In this way, in this method, when setting the detailed search range, the area far from the subject distance of the main subject candidate area and the double peak area are excluded, the detailed search range is set, and the detailed search is performed. To do. As a result, it is possible to prevent the background object from being recognized as the main subject due to the edge component of the background, and to detect the main subject more accurately. Further, by limiting the search range, it is possible to detect the main subject more efficiently.

  In this method as well, it is possible to detect a main subject by this method only when a face is detected first and no face is detected (see the second method).

<Sixth method>
When a main subject is detected by image analysis, an accurate and lean detection can be performed by setting a detailed search range according to the size of the main subject in the screen.

  In general, it is considered that the main subject is larger in the screen as it is located at a shorter distance (the peak position of the focus evaluation value is closer).

  Therefore, as shown in FIG. 18, the size of the detailed search range is adjusted according to the subject distance of the main subject candidate area. That is, the closer the subject distance of the main subject candidate area is, the larger the detailed search range is set ((a) in the same figure), and the smaller detailed search range is set (in the same figure (b)).

  Here, as the method of adjusting the main subject detailed search range in the screen, the following two methods can be considered.

  (i) Increase the main subject detail search range without changing the overall size of the image.

  (ii) Change the size of the entire screen and fix the main subject detail search range.

  That is, the relative size is adjusted. Even if any method is adopted, the same effect can be obtained.

  FIG. 19 is a flowchart showing a processing procedure for detecting a main subject by this method (sixth method).

  First, multipoint ranging is performed (step S50). Then, based on the result of the multipoint distance measurement, it is determined whether or not there is a distance measurement area where the position where the focus evaluation value is maximized (peak) is detected (step S51).

  If there is no distance measurement area where the position where the focus evaluation value is maximized is detected, the main subject candidate detection process cannot be performed, and the process ends.

  On the other hand, when there is a distance measurement area where the position where the focus evaluation value is maximized is detected, the main subject candidate area detection process is performed based on the detection result of the multipoint distance measurement (step S52). This processing is performed by the main subject candidate area detection unit 86.

  When the main subject candidate area is detected, the information is output to the main subject detection unit 88. The main subject detection unit 88 sets the center of the obtained main subject candidate area as the main subject rough position (step S53).

  Next, the detailed search range is determined by the main subject detection unit 88 based on the subject distance information of the main subject candidate area and the main subject rough position information (step S54). First, the main subject detection unit 88 determines the size of the detailed search range based on the subject distance of the main subject candidate area. At this time, as described above, the size of the detailed search range is set so that the subject distance of the main subject candidate area increases as the subject distance decreases. Then, the detailed search range of a predetermined shape is set around the main subject rough position by adjusting to the determined size. In the example shown in FIG. 18, the detailed search range is set to a rectangular shape.

  When the detailed search range is set, the main subject detection unit 88 analyzes the image obtained from the CCD 54 within the set detailed search range, and detects the main subject (step S55). At this time, the main subject is detected from the image using the number of edge components, the size of the same color cluster, and the like as an index for evaluating the likelihood of the main subject.

  Then, the detected position of the main subject is determined as the main subject detailed position (step S56).

  As described above, in this method, the size of the detailed search range is set to be enlarged or reduced according to the subject distance of the main subject candidate area. Accordingly, the detailed search range can be appropriately set according to the size of the main subject displayed on the screen, and the main subject can be accurately detected without waste.

  In this method as well, it is possible to detect a main subject by this method only when a face is detected first and no face is detected (see the second method).

  Further, an area far from the subject distance of the main subject candidate area and / or a double peak area may be excluded to set a detailed search range and perform a detailed search (third to third). See method 5).

<Seventh method>
As described in the sixth method above, when performing a detailed search, the main subject can be accurately detected without waste by adjusting the size of the detailed search range according to the size of the main subject displayed on the screen. can do.

  When shooting with zooming, it is generally considered that the larger the zoom is set to the telephoto side (telephoto side), the larger the main subject appears on the screen.

  Therefore, as shown in FIG. 20, the main subject detail search range is adjusted according to the zoom position (focal length). That is, the detailed search range is set to be larger as the zoom position is at the telephoto side (telephoto side) ((a) in the figure) (the detailed search range is smaller as the zoom position is at the wide side (wide angle side)) It sets so that it may become (the figure (b)).

  Here, as the method of adjusting the main subject detailed search range in the screen, the following two methods can be considered.

  (i) Increase the main subject detail search range without changing the overall size of the image.

  (ii) Change the size of the entire screen and fix the main subject detail search range.

  That is, the relative size is adjusted. Even if any method is adopted, the same effect can be obtained.

  FIG. 21 is a flowchart showing a procedure of processing for detecting a main subject by this method (seventh method).

  First, multipoint ranging is performed (step S60). Then, based on the result of the multipoint distance measurement, it is determined whether or not there is a distance measurement area where the position where the focus evaluation value is maximized (peak) is detected (step S61).

  If there is no distance measurement area where the position where the focus evaluation value is maximized is detected, the main subject candidate detection process cannot be performed, and the process ends.

  On the other hand, when there is a distance measurement area where the position where the focus evaluation value is maximized is detected, the main subject candidate area detection process is performed based on the detection result of the multipoint distance measurement (step S62). This processing is performed by the main subject candidate area detection unit 86.

  When the main subject candidate area is detected, the information is output to the main subject detection unit 88. The main subject detection unit 88 sets the center of the obtained main subject candidate area as the main subject rough position (step S63).

  Next, based on the zoom position information and the main subject rough position information, the main subject detector 88 determines a detailed search range (step S64). First, the main subject detection unit 88 acquires information on the current zoom position, and determines the size of the detailed search range according to the acquired zoom position. At this time, as described above, the size of the detailed search range is set so that the zoom position becomes larger as it is on the telephoto side. Then, the detailed search range of a predetermined shape is set around the main subject rough position by adjusting to the determined size. In the example shown in FIG. 20, the detailed search range is set to a rectangular shape.

  When the detailed search range is set, the main subject detection unit 88 analyzes the image obtained from the CCD 54 within the set detailed search range, and detects the main subject (step S65). At this time, the main subject is detected from the image using the number of edge components, the size of the same color cluster, and the like as an index for evaluating the likelihood of the main subject.

  Then, the detected position of the main subject is determined as the main subject detailed position (step S66).

  Thus, in this method, the size of the detailed search range is set to be enlarged or reduced according to the zoom position. Accordingly, the detailed search range can be appropriately set according to the size of the main subject displayed on the screen, and the main subject can be accurately detected without waste.

  In this method as well, it is possible to detect a main subject by this method only when a face is detected first and no face is detected (see the second method).

  Further, an area far from the subject distance of the main subject candidate area and / or a double peak area may be excluded to set a detailed search range and perform a detailed search (third to third). See method 5).

<Eighth method>
In this method, when the position of the main subject (main subject detailed position) detected by the image analysis is deviated from the main subject rough position, the position of the main subject is prevented from being greatly removed.

  FIG. 22 is a flowchart showing a procedure of processing for detecting a main subject by this method (eighth method).

  First, multipoint ranging is performed (step S70). Then, based on the result of the multipoint distance measurement, it is determined whether or not there is a distance measurement area where the position where the focus evaluation value is maximized (peak) is detected (step S71).

  If there is no distance measurement area where the position where the focus evaluation value is maximized is detected, the main subject candidate detection process cannot be performed, and the process ends.

  On the other hand, when there is a distance measurement area where the position where the focus evaluation value is maximized is detected, a main subject candidate area detection process is performed based on the detection result of the multipoint distance measurement (step S72). This processing is performed by the main subject candidate area detection unit 86.

  When the main subject candidate area is detected, the information is output to the main subject detection unit 88. The main subject detection unit 88 sets the center of the obtained main subject candidate area as the main subject rough position (step S73).

  Next, the detailed search range is determined by the main subject detection unit 88 (step S74). When the detailed search range is set, the main subject detection unit 88 analyzes the image obtained from the CCD 54 within the set detailed search range, and detects the main subject (step S75). Then, the detected position of the main subject is determined as the main subject temporary detailed position (step S76).

  Next, the focus area is set to a predetermined size centered on the main subject temporary detailed position. Then, the focus lens 110 is moved within a predetermined range, and the focus evaluation value of the set focus area is acquired (step S77).

  Here, the range in which the focus lens 110 is moved (focus evaluation value search range) is a position where the focus evaluation value of the main subject candidate area is maximized as a result of multipoint distance measurement (the focus position of the main subject candidate area). As a reference, a predetermined range is set before and after that (see FIG. 23).

  Next, it is determined whether or not the acquired focus evaluation value has a maximum point (peak) (step S78). As a result of this determination, if it is determined that there is no maximum point (peak) as shown in FIG. 23A, the main subject rough position is determined as the main subject detailed position (step S81).

  On the other hand, if there is a local maximum point, it is determined whether or not the position of the local maximum point is within an allowable range (step S79).

  Here, the allowable range is set to a predetermined range before and after the position where the focus evaluation value of the main subject candidate area becomes the maximum as a result of the multipoint distance measurement (the focus position of the main subject candidate area) (see FIG. (See FIG. 23).

  Whether or not the position of the maximum point is within the allowable range is determined based on the position where the acquired focus evaluation value is maximum and the position where the focus evaluation value of the main subject candidate area is maximum (the focus position of the main subject candidate area). ) Is determined (absolute value), and it is determined whether or not the difference is within a preset threshold value.

  As a result of this determination, as shown in FIG. 23B, if it is determined that the difference exceeds the threshold, the main subject rough position is determined as the main subject detail position (step S81).

  On the other hand, if it is determined that the difference is within the threshold, the main subject temporary detailed position is set to the main subject detailed position (step S80).

  In this way, the main subject obtained by multi-point ranging is measured again in a fixed search range (focus evaluation value search range) based on the position of the main subject (main subject temporary detailed position) detected by image analysis. By comparing the deviation (in the optical axis direction) of the candidate area with the position of the main subject, it is possible to prevent the position of the main subject from being greatly changed.

≪Pre-shooting process with main subject detection process≫
Next, a process before the main photographing using the main subject detection technique will be described.

  By using the main subject detection technique described above, the position of the main subject on the screen can be detected.

  Therefore, the main subject is detected using the through image in the live view, and various processes are performed using the detection result.

<First Embodiment>
In the present embodiment, main subject detection processing is performed during live view, a focus area is set for the detected main subject, and continuous AF is executed.

  FIG. 24 is a state transition diagram of continuous AF performed before the actual photographing.

  As shown in the figure, when a face is detected in a through image, the process proceeds to face continuous AF, a focus area is set for the detected face, and face continuous AF is performed.

  On the other hand, if no face is detected, the process proceeds to the central continuous AF. Then, when focusing is performed with the central continuous AF (or close to focusing) and the shooting scene is changed, the multi-point distance measurement is started.

  As a result of the multipoint distance measurement, when there is no distance measurement area where a peak is detected, the process proceeds to the central continuous AF. In this case, when a face is detected, the process proceeds to face continuous AF.

  On the other hand, when there is a ranging area where a peak is detected as a result of multipoint ranging, main subject detection processing is performed based on the multipoint ranging result. Then, a focus area is set for the detected main subject, and continuous AF (main subject continuous) is performed. Thereafter, when a face is detected, a focus area is set for the detected face and face continuous AF is performed. If the scene changes, the process proceeds to the central continuous AF.

  FIG. 25 is a flowchart showing a procedure of continuous AF processing performed before the actual photographing.

  First, face detection processing is performed (step S111), and it is determined whether a face is detected (step S112).

  If a face is detected, a focus area is set at the detected face position, and face continuous AF is performed (step S113). The face continuous AF is performed until no face is detected.

  On the other hand, when a face is not detected, a focus area is set at the center of the screen, and central continuous AF is performed (step S120). Thereafter, it is determined whether or not the central continuous AF is in focus (or whether or not the focus is close to the focus) (step S121).

  If the continuous AF is not in focus, the process returns to step S111, and face detection is performed again.

  On the other hand, if the continuous AF is in focus (or close to the focus), it is determined whether or not the shooting scene has changed (step S122).

  In this case, the scene change is detected by comparing a pre-registered image with the currently captured image to determine whether or not the scene has changed.

  The image to be compared is an image when the main subject is detected in this example. Therefore, multipoint ranging is executed for the first time without determining whether or not there is a scene change.

  If there is no scene change, the process returns to step S111, and face detection is performed again.

  On the other hand, if the scene changes, the central continuous AF is stopped and multipoint distance measurement is performed (step S123). Then, based on the result of the multipoint distance measurement, it is determined whether or not there is a distance measurement area where the position where the focus evaluation value is maximized (peak) is detected (step S124).

  As a result of this determination, if it is determined that there is no ranging area in which the peak is detected, the process returns to step S111, and face detection is performed again.

   On the other hand, when it is determined that there is a ranging area where the peak is detected, an image is captured. The captured image is registered as a reference image for comparison (step S125). This process is performed, for example, by recording a reference image for comparison in the frame memory 74.

  After registration of the reference image for comparison, main subject detection processing is performed using the main subject detection technique described above (step S126). Then, a focus area is set for the detected main subject, and main subject continuous AF is performed (step S127).

  Thereafter, it is determined whether or not the shooting scene has changed (step S128). If there is no change in the shooting scene, the main subject continuous AF is continued. On the other hand, if there is a change in the shooting scene, the process returns to step S111, and face detection is performed again.

  As described above, in the present embodiment, when a face is detected in the through image, the focus area is set to the face and the face continuous AF is performed. When the face is not detected, the main subject detection process is performed. To perform main subject continuous AF. Thereby, even if there is no face on the screen, a through image in which the main subject is in focus can be displayed.

  When performing the main subject detection process, first, the central continuous AF is performed, and the multipoint distance measurement is performed only when the central continuous AF is in focus and the shooting scene changes. Do. Thereby, the frequency | count of execution of multipoint ranging can be suppressed, and the view angle fluctuation | variation of the through image displayed on the monitor 24 can be suppressed. Thereby, it is possible to make the screen easy to see even during the through image display.

  When the main subject is detected, a frame indicating the position of the focus area set for the main subject is displayed on the screen of the monitor 24 as shown in FIG. As shown in FIG. 5B, similarly, when a face is detected, a frame indicating the position of the focus area set for the face is displayed on the screen of the monitor 24.

  In this example, when the main subject detection process is performed, the multipoint ranging is executed only when the central continuous AF is in focus and the shooting scene is changed. A configuration that performs multipoint ranging only when continuous AF is in focus, a configuration that performs multipoint ranging only when the shooting scene changes, and a fixed time from the previous multipoint ranging It is also possible to adopt a configuration in which multipoint ranging is executed only when it has elapsed. Alternatively, these may be combined as appropriate, and the multipoint ranging may be executed only when all the conditions are satisfied.

  Further, in this example, when determining whether or not the shooting scene has changed, an image when the main subject is detected is registered as a reference image, and whether or not the shooting scene has changed in comparison with the reference image is determined. Although it is determined, it is also possible to adopt a configuration in which a change in the shooting scene is detected at a constant interval (a constant frame interval) (for example, an image before N frames is registered as a reference image, Detect scene changes.)

  Further, the comparison information does not necessarily have to be the image data itself, and a feature point for comparison may be extracted and the feature point information may be registered as reference information. In this case, scene variation is also performed by extracting feature points from the current image and comparing the extracted feature points.

  In this example, the main subject detection process is performed only when face detection is performed and no face is detected. However, face detection is not necessarily performed. Further, even when a face is detected, main subject detection processing by multipoint distance measurement may be executed.

<Second Embodiment>
In the present embodiment, main subject detection processing is performed during live view, and the detected main subject is set as a focus area when the release button 18 is half-pressed (when focusing is instructed (S1)).

  FIG. 27 is a flowchart illustrating a processing procedure for setting a focus area for a main subject detected during live view.

  First, face detection is performed (step S131), and it is determined whether a face is detected (step S132).

  If a face is detected, a focus area is set at the detected face position, and face continuous AF is performed (step S133). The face continuous AF is performed until no face is detected.

  On the other hand, if a face is not detected, a focus area is set at the center of the screen, and central continuous AF is performed (step S140). Thereafter, it is determined whether or not the central continuous AF is in focus (or whether or not the state is close to the focus) (step S141).

  If the continuous AF is not in focus, the process returns to step S131, and face detection is performed again.

  On the other hand, if the continuous AF is in focus (or close to the focus), it is determined whether or not the shooting scene has changed (step S142).

  In this case, the scene change is detected by comparing a pre-registered image with the currently captured image to determine whether or not the scene has changed. The image to be compared is an image when the main subject is detected. Therefore, multipoint ranging is executed for the first time without determining whether or not there is a scene change.

  If there is no scene change, the process returns to step S131, and face detection is performed again.

  On the other hand, when the scene changes, the central continuous AF is stopped and multipoint distance measurement is performed (step S143). Then, based on the result of the multipoint distance measurement, it is determined whether or not there is a distance measurement area where the position where the focus evaluation value is maximized (peak) is detected (step S144).

  As a result of this determination, if it is determined that there is no ranging area where a peak is detected, the process returns to step S131 and face detection is performed again.

   On the other hand, when it is determined that there is a ranging area where the peak is detected, an image is captured. Then, the captured image is registered as a reference image for comparison (step S145). This process is performed, for example, by recording a reference image for comparison in the frame memory 74.

  After registration of the reference image for comparison, the main subject is detected using the technique for detecting the main subject (step S146). Then, the focus area at S1 is set at the position of the detected main subject (step S147).

  As described above, in the present embodiment, the main subject is detected while the through image is displayed, and the focus area at S1 is set for the detected main subject.

  FIG. 28 is a flowchart illustrating a processing procedure up to the actual photographing.

  When the photographing mode is set, processing before the main photographing is executed (step S151). That is, the through image is displayed on the monitor 24 and the main subject detection process is performed.

  The CPU 90 determines whether or not the release button 18 is half-pressed based on the input from the operation system control unit 84 (step S152). When the release button 18 is half-pressed, an S1 ON signal is input to the CPU 90 as a shooting preparation instruction signal (S1 OFF when the half-press is released). The CPU 90 determines whether or not the release button 18 is half-pressed based on whether or not the S1 ON signal is input.

  When the release button 18 is half-pressed, the CPU 90 executes shooting preparation processing.

  First, AE processing for main photographing is executed, and exposure conditions for main photographing are determined (step S153). That is, an AE image is taken in, the brightness of the subject is detected, and exposure conditions (shutter speed, aperture value, ISO sensitivity, etc.) that achieve proper exposure are determined.

  Next, AF processing is executed (step S154). That is, the focus lens 110 is moved from the closest end to the infinity end, and a position where the focus evaluation value in a preset focus area is maximized is detected. Then, the focus lens is moved to the detected position.

  At this time, when a main subject is detected, a focus area is set for the detected main subject, and focusing is performed.

  Thereafter, the CPU 90 determines whether or not the release button 18 has been fully pressed based on an input from the operation system control unit 84 (step S155). When the release button 18 is fully pressed, an S2 ON signal is input to the CPU 90 as a main shooting instruction signal (S2 OFF when the full press is released). The CPU 90 determines whether or not the release button 18 is fully pressed based on whether or not the S2ON signal is input.

  If it is determined that the release button 18 has not been fully pressed, the CPU 90 determines whether or not the half-press of the release button 18 has been released based on the input from the operation system control unit 84 (step S158). . If it is determined that the half-press of the release button 18 is released, the process returns to step S151 and the pre-shooting process is executed again.

  On the other hand, when determining that the release button 18 is fully pressed, the CPU 90 executes a process of actual photographing (step S156). That is, the CCD 54 is exposed under the exposure conditions obtained by the AE process, and a recording image is captured. Then, the obtained image is recorded on the recording medium 78 (step S157).

  In this way, the main subject is detected before the main shooting, and the main shooting is performed when the main subject is present in a region other than the center by focusing on the detected main subject. However, it is possible to take an image in which the main subject is in focus.

  In the above example, the main subject detection process is performed before the main photographing, and the focus area at S1 is set for the detected main subject. However, when a face is detected, the detected face is set to S1. It may be configured to set the current focus area.

  In the above example, the focus lens is moved from the closest end to the infinity end during AF processing after S1, and the position where the focus evaluation value is maximized is detected, but the main subject is detected. In this case, the AF processing may be performed by limiting the focus evaluation value search range to a predetermined range before and after the position of the focus lens when the main subject is detected. Thereby, focusing can be performed more quickly.

<Third Embodiment>
As one form of continuous AF, an automatic tracking function is known. As shown in FIG. 29, in the automatic tracking function, when a subject to be focused is designated, the focus area is automatically moved in accordance with the movement of the subject, and the designated subject is continuously focused.

  In the present embodiment, main subject detection processing is performed during live view, and the detected main subject is automatically set as an automatic tracking target.

  FIG. 30 is a flowchart illustrating a processing procedure for setting the position of the main subject detected during live view as a target for automatic tracking.

  First, face detection is performed (step S171), and it is determined whether a face is detected (step S172).

  If a face is detected, a focus area is set at the position of the detected face, and face continuous AF is performed (step S173). The face continuous AF is performed until no face is detected.

  On the other hand, when a face is not detected, a focus area is set at the center of the screen, and central continuous AF is performed (step S180). Thereafter, it is determined whether or not the central continuous AF is in focus (or whether or not the focus is close to the focus) (step S181).

  If the central continuous AF is not in focus, the process returns to step S171 and face detection is performed again.

  On the other hand, when the continuous AF is in focus (or when it is close to the focus), it is determined whether or not the shooting scene has changed (step S182).

  In this case, the scene change is detected by comparing a pre-registered image with the currently captured image to determine whether or not the scene has changed. The image to be compared is an image when the main subject is detected. Therefore, multipoint ranging is executed for the first time without determining whether or not there is a scene change.

  If there is no scene change, the process returns to step S171, and face detection is performed again.

  On the other hand, when the scene changes, the central continuous AF is stopped and multipoint distance measurement is performed (step S183). Then, based on the result of the multipoint distance measurement, it is determined whether or not there is a distance measurement area where the position where the focus evaluation value is maximized (peak) is detected (step S184).

  As a result of this determination, if it is determined that there is no ranging area in which the peak is detected, the process returns to step S171 and face detection is performed again.

   On the other hand, when it is determined that there is a ranging area where the peak is detected, an image is captured. Then, the captured image is registered as a reference image for comparison (step S185). This process is performed, for example, by recording a reference image for comparison in the frame memory 74.

  After registration of the reference image for comparison, main subject detection is performed using the above-described main subject detection processing technique (step S186). Then, the detected position of the main subject is set to the initial position for automatic tracking (step S187).

  As described above, in the present embodiment, the main subject is detected during the live view, and the detected position of the main subject is automatically set as the initial position for automatic tracking. This eliminates the need for the photographer to specify the initial position when performing automatic tracking, thereby improving the usability of the digital camera.

  DESCRIPTION OF SYMBOLS 10 ... Digital camera, 12 ... Camera body, 14 ... Shooting lens, 16 ... Flash, 18 ... Release button, 20 ... Zoom lever, 22 ... Power button, 24 ... Monitor, 26 ... Menu / OK button, 28 ... Cross button, DESCRIPTION OF SYMBOLS 30 ... Back button, 32 ... Playback button, 50 ... Lens drive part, 52 ... Flash control part, 54 ... Imaging device (CCD), 56 ... CCD control part, 58 ... Analog signal processing part, 60 ... Timing generator, 62 ... Image input controller, 64 ... AF processing unit, 64A ... high pass filter (HPF), 64B ... gate, 64C ... adder, 66 ... AE / AWB processing unit, 68 ... image processing unit, 70 ... face detection unit, 72 ... compression / Decompression processing unit, 74 ... Frame memory, 76 ... Media control unit, 78 ... Recording media, 80 ... Display control unit, 8 DESCRIPTION OF SYMBOLS ... Operation part, 84 ... Operation system control part, 86 ... Main subject candidate area detection part, 88 ... Main subject detection part, 90 ... CPU, 92 ... EEPROM, 110 ... Focus lens, 112 ... Zoom lens, 114 ... Aperture, 116 ... Shutter, 120 ... Focus lens drive unit, 122 ... Zoom lens drive unit, 124 ... Aperture drive unit, 126 ... Shutter drive unit

Claims (12)

An imaging optical system;
Imaging means for imaging an image via the imaging optical system;
Dividing the imaging region of the imaging unit into a plurality of distance measurement areas, and multi-point distance measuring means for measuring a subject distance in each ranging area,
A main subject candidate area detecting unit that detects, as a main subject candidate area, a ranging area estimated to belong to a main subject based on a measurement result of the multipoint ranging unit, and calculates a subject distance of each ranging area. Add the value weighted with a greater weight to the distance measurement area closer to the center of the screen and the value weighted with a greater weight to the subject distance of each distance measurement area to obtain the characteristics of the main subject in each distance measurement area. A main subject candidate area detecting means for digitizing and selecting a distance measuring area having a maximum value as a main subject candidate area;
Detailed search range setting means for setting a specific range centered on the center of the main subject candidate area as a detailed search range;
A main subject position detecting means for analyzing an image captured by the imaging means within the detailed search range and detecting a position of a main subject;
Focus area setting means for setting a focus area at the position of the main subject detected by the main subject position detecting means;
An image obtained from the imaging means in the focus area by setting a detection range within a certain range with reference to the focus position of the main subject candidate area and changing the focal position of the imaging optical system within the detection range. Ranging means for detecting the position where the contrast of the maximum is detected and measuring the subject distance of the focus area;
The difference between the subject distance measured by the distance measuring means and the subject distance of the main subject candidate area is obtained, and when the difference exceeds a threshold, the position of the main subject is corrected to the center of the main subject candidate area. Main subject position correcting means for
An imaging apparatus comprising: The main subject position correcting means, to claim 1, characterized in that to correct the position of the main subject even when the contrast can not be detected by the position with the maximum in the distance measuring means in the center of the main subject candidate areas The imaging device described. An imaging optical system;
Imaging means for imaging an image via the imaging optical system;
Dividing the imaging region of the imaging unit into a plurality of distance measurement areas, and multi-point distance measuring means for measuring a subject distance in each ranging area,
A main subject candidate area detecting unit that detects, as a main subject candidate area, a ranging area estimated to belong to a main subject based on a measurement result of the multipoint ranging unit, and calculates a subject distance of each ranging area. Add the value weighted with a greater weight to the distance measurement area closer to the center of the screen and the value weighted with a greater weight to the subject distance of each distance measurement area to obtain the characteristics of the main subject in each distance measurement area. A main subject candidate area detecting means for digitizing and selecting a distance measuring area having a maximum value as a main subject candidate area;
Detailed search range setting means for setting a specific range centered on the center of the main subject candidate area as a detailed search range;
A main subject position detecting means for analyzing an image captured by the imaging means within the detailed search range and detecting a position of a main subject;
Focus area setting means for setting a focus area at the position of the main subject detected by the main subject position detecting means;
By changing the focal position of the imaging optical system within a predetermined range before and after the in-focus position of the main subject candidate area, a position where the contrast of the image obtained from the imaging unit is maximized in the focus area is detected. Ranging means for measuring the subject distance of the focus area;
As a result of the measurement by the distance measuring means, when the position where the contrast is maximum is not detected, main subject position correcting means for correcting the position of the main subject to the center of the main subject candidate area;
An imaging apparatus comprising:   The imaging apparatus according to any one of claims 1 to 3, wherein the main subject position detection unit detects a position where the main subject is most likely to be a main subject position within the detailed search range.   The main subject position detecting means searches the detailed search range from the center to the outside, and detects the position at the time when the evaluation value of the main subjectness exceeds a threshold as the position of the main subject. 4. The imaging device according to any one of 3. The detailed search range setting means, wherein when comparing the object distance of each ranging area obtained by a multi-point distance measuring means, said distance measuring the difference of the subject distance Prefecture is not less than the threshold value of the main object candidate areas 6. The imaging apparatus according to claim 1, wherein the detailed search range is set excluding an area. The multipoint distance measuring means changes the focal position of the imaging optical system, and detects the position where the contrast of the image obtained from the imaging means is maximized for each distance measurement area, thereby each of the distance measuring means. the subject distance is measured in the area,
7. The detailed search range setting means sets the detailed search range by excluding a ranging area where two or more positions where the contrast becomes maximum is detected. The imaging device according to item. The detailed search range setting means, according to any one of claims 1 to 5, characterized in that set by scaling the magnitude of the detailed search range according to the object distance of the main subject candidate areas Imaging device. The imaging optical system includes a variable magnification optical system,
The imaging according to any one of claims 1 to 5, wherein the detailed search range setting means sets the size of the detailed search range in accordance with a focal length of the imaging optical system. apparatus. 2. The main subject position detection unit detects a main subject by using at least one of information on the number of edge components or the size of a cluster of the same color system as an image characteristic. The imaging device according to any one of 1 to 9 . A step of the imaging region of the imaging means into a plurality of distance measurement areas is ranging multipoint measuring the subject distance in each ranging area for capturing an image through an imaging optical system,
A step of detecting, as a main subject candidate area, a range finding area to which the main subject belongs based on the result of the multipoint ranging, and measuring the subject distance of each range finding area close to the center of the screen The value weighted with a larger weight for each area and the value weighted with a larger weight for the subject distance of each distance measurement area are added together to quantify the characteristics of the main subject in each distance measurement area. Selecting a target ranging area as a main subject candidate area,
Setting a specific range centered on the center of the main subject candidate area as a detailed search range;
Analyzing an image captured by the imaging means within the detailed search range, and detecting a position of a main subject;
Setting a focus area at the position of the detected main subject;
An image obtained from the imaging means in the focus area by setting a detection range within a certain range with reference to the focus position of the main subject candidate area and changing the focal position of the imaging optical system within the detection range. Detecting the position where the contrast of the maximum is detected and measuring the subject distance of the focus area;
Obtaining a difference between the measured subject distance and the subject distance of the main subject candidate area, and correcting the position of the main subject to the center of the main subject candidate area when the difference exceeds a threshold;
A main subject detection method characterized by comprising: A step of the imaging region of the imaging means into a plurality of distance measurement areas is ranging multipoint measuring the subject distance in each ranging area for capturing an image through an imaging optical system,
A step of detecting, as a main subject candidate area, a range finding area to which the main subject belongs based on the result of the multipoint ranging, and measuring the subject distance of each range finding area close to the center of the screen The value weighted with a larger weight for each area and the value weighted with a larger weight for the subject distance of each distance measurement area are added together to quantify the characteristics of the main subject in each distance measurement area. Selecting a target ranging area as a main subject candidate area,
Setting a specific range centered on the center of the main subject candidate area as a detailed search range;
Analyzing an image captured by the imaging means within the detailed search range, and detecting a position of a main subject;
Setting a focus area at the position of the detected main subject;
By changing the focal position of the imaging optical system within a predetermined range before and after the in-focus position of the main subject candidate area, a position where the contrast of the image obtained from the imaging unit is maximized in the focus area is detected. Measuring the subject distance of the focus area;
As a result of measurement, if the position where the contrast is maximum is not detected, correcting the position of the main subject to the center of the main subject candidate area;
A main subject detection method characterized by comprising:

Images