JP5027580B2 - Imaging apparatus, method, and program - Google Patents

Description

  The present invention relates to a photographing apparatus such as a digital camera that acquires an image by photographing.

  When shooting an object such as a person in a shooting device such as a digital camera or a digital video camera, settings such as focus and exposure appropriate for shooting are performed, and various image processing is performed on the object in the shot image. Therefore, it is expected to detect an object from a photographed image (subject image).

In such an image photographed by the photographing apparatus, the photographing conditions are not always constant. Therefore, in Patent Document 1, a position where the correlation value between the target image and the face template image is equal to or larger than a set threshold is set as a face region. There has been proposed a method for improving the face detection performance by changing the correlation threshold value in accordance with the brightness of the target image.
JP-A-2005-004287

  However, the method proposed in Patent Document 1 that changes the correlation value setting threshold according to the brightness cannot sufficiently cope with changes in various photographing conditions, and its detection performance is insufficient. There is a case.

  In view of the circumstances described above, an object of the present invention is to provide an imaging apparatus that can further improve the detection performance against changes in various imaging conditions.

 An imaging apparatus according to the present invention includes an object detection unit that performs an object detection process for detecting an object from a captured image, a scene determination unit that determines an image scene, and a target according to a determination result of the scene determination unit. And a parameter control means for controlling a parameter used for the object detection process.

 The parameter means an auxiliary variable used for the object detection processing means. For example, a gain value used in the object detection processing means, a detection determination threshold value, a degree of noise reduction amount in noise reduction, a degree of contour correction, a numerical value including one or more of gamma curve values, etc. There may be.

 In the above apparatus, the object detection means detects the object from the image using a plurality of parameters having different values including the changed parameter changed by the parameter control means, and integrates the detection results to obtain the object. It may be an object detection result.

 The parameter control means may change the parameter when no object is detected by the object detection means using the parameter.

 Further, it may be provided with a scene discriminating unit that discriminates an image scene, and the parameter control unit controls the parameter according to the discrimination result of the scene discriminating unit.

 Further, the scene determination means may be determined by one or more determination processes of backlight determination, exposure determination, lightness determination, and noise determination.

 Note that an image scene means an image situation that affects detection of an object such as backlight, night view, or underwater.

  The parameter control means may change the parameter control method according to the timing at which the object detection means captures an image for performing the object detection processing or the timing for performing the object processing. The object detection process is not limited to the detection main body, and may include various pre-processes performed on the target image before the detection main body is performed.

 The imaging method of the present invention performs an object detection process for detecting an object from a captured image, discriminates an image scene, and controls parameters used for the object detection process according to the discrimination result. It is characterized by.

  In addition, you may provide as a program for making a computer perform the imaging | photography method by this invention.

  According to the imaging apparatus of the present invention, the apparatus includes an object detection unit that performs an object detection process for detecting an object from a captured image and a parameter control unit that controls a parameter used in the object detection process. Therefore, even if the shooting situation changes, the parameter can be changed to a parameter suitable for detection processing from an image shot under the shooting situation, so that the object detection performance can be improved.

  In the above apparatus, the object detection means detects the object from the image using a plurality of parameters having different values including the changed parameter changed by the parameter control means, and integrates the detection results to detect the object. When the detection result is obtained, the detection accuracy of the object can be further improved.

  In the above apparatus, when the parameter control means is to change the parameter when the object is not detected by the object detection means, by performing the detection process using the changed parameter, The detection rate of the object can be improved.

 In the above apparatus, in the case where a scene discrimination means for discriminating an image scene is provided, and the parameter control means controls a parameter in accordance with the discrimination result of the scene discrimination means, the image based on the discrimination result of the scene Since the parameters suitable for the detection process can be determined efficiently, the detection efficiency can be improved.

 Further, when the parameter control means changes the parameter control method according to the timing at which the object detection means captures an image for performing the object detection processing or the timing at which the object processing is performed, an image is captured. On the basis of the timing information, it is possible to efficiently determine a parameter suitable for detection processing on the image.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a perspective view centering on the back surface of a digital camera 1 which is an embodiment of a photographing apparatus of the present invention. As shown in FIG. 1, an operation mode switch 11, a menu / OK button 12, a zoom / up / down lever 13, a left / right button 14, Back (back) are provided on the back of the digital camera 1 as an interface for operation by a photographer. ) Button 15 and display switching button 16, as well as a finder 17 and a liquid crystal monitor 18. A release button 19 is provided on the upper surface. FIG. 2 is a perspective view centering on the front surface of the digital camera 1 according to the first embodiment. As shown in FIG. 2, a lens 20, a lens cover 21, a power switch 22, a finder window 23, a flash 24, a self-timer lamp 25, and AF auxiliary light 26 are provided on the front surface, and a media slot 27 is provided on the side surface. It has been.

  The operation mode switch 11 is a slide switch for switching operation modes of a still image shooting mode, a moving image shooting mode, and a playback mode.

  When the menu / OK button 12 is pressed, various menus for setting the shooting mode, flash emission mode, number of recording pixels, sensitivity, and the like are displayed on the liquid crystal monitor 18 or the menu displayed on the liquid crystal monitor 18 is displayed. This is a button for confirming selection / setting based on the. The menu / OK button 12 can be used to set the shooting mode to an auto mode for automatically setting exposure and white balance and a manual mode for manually setting.

  The zoom / up / down arrow lever 13 is used to adjust the telephoto / wide angle during shooting by tilting the lever in the vertical direction, and to move the cursor in the menu screen displayed on the liquid crystal monitor 18 up and down during various settings. The lever.

  The left and right arrow buttons 14 are buttons for moving the cursor in the menu screen displayed on the liquid crystal monitor 18 to the left and right at various settings.

  The Back button 15 is a button for stopping various setting operations when pressed and displaying the previous screen on the monitor 18.

  The display switching button 16 is a button for switching ON / OFF of the display of the liquid crystal monitor 18, various guide displays, ON / OFF of character display, and the like when pressed.

  The viewfinder 17 is used for a user to look into the subject for composition and focus when photographing a subject. The subject image seen from the viewfinder 17 is projected through the viewfinder window 23.

  The contents set by the operation of each button and lever described above can be confirmed by the display on the liquid crystal monitor 18, the position of the lamp in the finder, the slide lever, and the like.

  The liquid crystal monitor 18 displays a through image for confirming a subject at the time of shooting, thereby functioning as an electronic viewfinder, displaying a still image and a moving image after shooting, and displaying various setting menus.

  The release button 19 is a two-stage push switch. Various shooting preparation processes such as automatic exposure adjustment and automatic focus adjustment are performed when the user presses the release button 19 halfway. When the release button 19 is fully pressed in this state, shooting is performed based on the determined exposure and focus position, and the image displayed on the liquid crystal monitor 18 is recorded as a captured image.

  The photographic lens 20 is for forming a subject image on a predetermined image plane (CCD etc. in the digital camera 1), and includes a focus lens, a zoom lens, and the like. The lens cover 21 covers the surface of the photographic lens 20 when the power of the digital camera 1 is turned off or in the playback mode, and protects the photographic lens 20 from dirt and dust. The power switch 22 is a switch for switching on / off the power of the digital camera 1. The strobe light 24 is used to instantaneously irradiate the subject with light necessary for photographing while the shutter button 19 is pressed and the shutter inside the digital camera 1 is opened. The self-timer lamp 25 is for informing the subject of the opening / closing timing of the shutter when shooting with the self-timer.

  FIG. 3 is a schematic block diagram showing the functional configuration of the digital camera 1. The digital camera 1 shown in FIG. 3 converts image data acquired by shooting into an Exif format image file and records it on an external recording medium 70 that can be attached to and detached from the main body.

  The operation system of this digital camera includes the above-described operation mode switch 11, menu / OK button 12, zoom / up / down arrow lever 13, left / right arrow button 14, Back button 15, display switch button 16, release button 19, The power switch 22 and an operation system control unit 74 which is an interface part for transmitting the operation contents of these switches to the CPU 75 are provided.

  The optical system includes a focus lens 20a and a zoom lens 20b. Each of these lenses is step-driven by a focus lens driving unit 51 and a zoom lens driving unit 52 including a motor and a motor driver, and is configured to be movable in the optical axis direction. The focus lens driving unit 51 step-drives the focus lens 20 a based on the focus driving amount data output from the AF processing unit 62. The zoom lens driving unit 52 controls step driving of the zoom lens 20 b based on the operation amount data of the zoom / up / down lever 13.

  The diaphragm 54 is driven by a diaphragm driving unit 55 including a motor and a motor driver. The aperture drive unit 55 adjusts the aperture diameter of the aperture 54 based on aperture value data output from the AE (automatic exposure) / AWB (auto white balance) processing unit 63.

  The shutter 56 is a mechanical shutter and is driven by a shutter drive unit 57 including a motor and a motor driver. The shutter drive unit 57 controls the opening / closing of the shutter 56 in accordance with a signal generated by pressing the release button 19 and the shutter speed data output from the AE / AWB processing unit 63.

  A CCD 58, which is an image sensor, is provided behind the optical system. The CCD 58 has a photoelectric surface in which a large number of light receiving elements are two-dimensionally arranged, and subject light that has passed through the optical system forms an image on the photoelectric surface and is subjected to photoelectric conversion. In front of the photocathode, there is a microlens array (not shown) for condensing light on each pixel, and a color filter array (not shown) in which filters of R, G, B colors are regularly arranged. Has been placed. The CCD 58 outputs the charges accumulated for each pixel as a serial analog photographing signal line by line in synchronization with the vertical transfer clock and horizontal transfer clock supplied from the CCD controller 59. The time for accumulating charges in each pixel, that is, the exposure time is determined by an electronic shutter drive signal given from the CCD controller 59. The gain of the CCD 58 is adjusted by the CCD control unit 59 so that an analog imaging signal having a predetermined size can be obtained.

  The image signal output from the CCD 58 is input to the analog signal processing unit 60. The analog signal processing unit 60 includes a correlated double sampling circuit (CDS) that removes noise from the analog signal, an auto gain controller (AGC) that adjusts the gain of the analog signal, and an A / A that converts the analog signal into a digital signal. And a D converter (ADC). The image data converted into the digital signal is CCD-RAW data having R, G, and B density values for each pixel.

  The timing generator 72 generates a timing signal. The timing signal is input to the shutter drive unit 57, the CCD control unit 59, and the analog signal processing unit 60, and the operation of the shutter button 19 and the opening / closing of the shutter 56 are performed. The charge capturing of the CCD 58 and the processing of the analog signal processing 60 are synchronized.

  The flash control unit 73 causes the flash 24 to emit light during shooting. Specifically, when the flash emission mode is set to flash on, and when the flash emission mode is the auto mode and the pre-image is not at a predetermined brightness, the flash 24 is turned on, and the flash 24 is turned on at the time of shooting. Make it emit light. On the other hand, when the flash emission mode is set to flash off, the flash 24 is prohibited from emitting light during shooting.

  Further, the flash control unit 73 determines the light emission amount of the flash 24 based on the focal position obtained by the AF processing unit 62 so that the light emission amount increases as the focal position increases. This light emission amount is defined as a first light emission amount. In the case where shooting is performed with the flash 24 emitting light, if the face is included in the pre-image acquired by the pre-shooting, the light emission amount of the flash 24 is set to the second light emission amount smaller than the first light emission amount. Determine the amount of luminescence. Here, the amount of light emission may be reduced to such an extent that no whiteout occurs on the face, and the degree may be obtained experimentally.

  The image input controller 61 writes the CCD-RAW data input from the analog signal processing unit 60 in the frame memory 66. The frame memory 66 is a working memory used when various digital image processing (signal processing) described later is performed on image data. For example, an SDRAM that performs data transfer in synchronization with a bus clock signal having a fixed period. (Synchronous Dynamic Random Access Memory) is used.

  The display control unit 71 displays the image data stored in the frame memory 66 on the liquid crystal monitor 18 as a through image, or displays the image data stored in the external recording medium 70 on the liquid crystal monitor 18 in the reproduction mode. Is for. The through image is taken by the CCD 58 at a predetermined interval while the shooting mode is selected.

  The AF processing unit 62 and the AE / AWB processing unit 63 determine shooting conditions based on the pre-image. This pre-image is an image represented by image data stored in the frame memory 66 as a result of the CPU 75 having detected a half-press signal generated by half-pressing the release button 19 causing the CCD 58 to perform pre-photographing. It is.

  The AF processing unit 62 detects the focal position based on the pre-image and outputs focus drive amount data (AF processing). As a focus position detection method, for example, a passive method that detects a focus position using a feature that the contrast of an image is high in a focused state can be considered.

  The AE / AWB processing unit 63 measures subject brightness based on the pre-image, determines an aperture value, shutter speed, etc. based on the measured subject brightness, and determines aperture value data and shutter speed data as exposure setting values. At the same time (AE processing), white balance at the time of shooting is automatically adjusted (AWB). The exposure and white balance can be set manually by the user of the digital camera 1 when the shooting mode is set to the manual mode. Even when the exposure and white balance are set automatically, the user can manually adjust the exposure and white balance by giving an instruction from the operation system such as the menu / OK button 12.

  The image processing unit 64 performs image quality correction processing such as gamma correction, sharpness correction, and contrast correction on the image data of the main image, CCD-RAW data as Y data that is a luminance signal, Cb data that is a blue color difference signal, and YC processing is performed for conversion to YC data composed of Cr data which is a red color difference signal. The main image is an image based on image data that is captured from the CCD 58 in the main photographing executed when the release button 19 is fully pressed and stored in the frame memory 66 via the analog signal processing unit 60 and the image input controller 61. It is. The upper limit of the number of pixels of the main image is determined by the number of pixels of the CCD 58. For example, the number of recorded pixels can be changed by setting such as fine and normal. On the other hand, the number of images of the through image and the pre-image is smaller than that of the main image.

  The compression / decompression processing unit 65 performs compression processing on the image data of the main image that has been corrected and converted by the image processing unit 64 in a compression format such as JPEG, and generates an image file. A tag storing incidental information such as shooting date and time is added to the image file based on the Exif format or the like. In the reproduction mode, the compression / decompression processing unit 65 reads the compressed image file from the external recording medium 70 and performs the decompression process. The decompressed image data is output to the liquid crystal monitor 18.

  The media control unit 67 accesses the external recording medium 70 and controls writing and reading of the image file.

  The internal memory 68 stores various constants set in the digital camera 1, programs executed by the CPU 75, and the like.

  The object detection unit 80 performs an object detection process for detecting an object from a photographed image and outputs the detection result. This object detection process includes a detection main body and various pre-processes performed on the target image data before performing the detection main body. Specifically, as shown in FIG. The output CCD-RAW data is input, and the input image data is subjected to digital signal processing such as scratch correction, offset correction, gain correction, contrast correction, gamma correction, resizing, etc. as preprocessing, and signal processing The detection result of the detection main body for detecting the position where the correlation value with the face template image is equal to or larger than the set threshold value is output from the obtained image data.

  In addition, the object detection unit 80 detects an area having a facial feature included in the face (for example, having skin color, having eyes, or having a face shape) as the face area.

  For example, a technique disclosed in Japanese Patent Application Laid-Open No. 2006-202276 (hereinafter referred to as Reference Document 1) can be used. In the method of Reference 1, the face tracking is performed by updating the weights sequentially when re-sampling a known method or learning data such as motion vector and feature point detection. It is conceivable that a machine learning method based on Adaboost, which is a method of creating an integrated learning machine by adding weights to learning machines and adding them, is used. For example, the average frame model is fitted into the actual face image, and the position of each landmark on the average frame model is moved so as to match the position of the corresponding landmark detected from the face. When constructing a deformed face frame model, the machine learning method shows that the brightness profile at points on the sample images that are known to be a given landmark and that it is not the landmark Using the classifier obtained by learning the luminance profiles at points on the plurality of sample images and the identification conditions for each classifier (such as a face detection determination threshold value), This is a method for detecting the position of a point indicating. Moreover, it is also possible to use the method of Unexamined-Japanese-Patent No. 2004-334836 (henceforth the reference document 2). The technique of Reference 2 cuts out image data of a certain size from image data, compares each cut-out image data with the matching data of feature part image data, and the feature part image exists in the processing target image. This is a technique using a feature portion extraction method of image data for detecting whether or not to perform. In addition to the human face area, an animal face or the like may be detected as a specific subject as in the technique disclosed in Japanese Patent Application Laid-Open No. 2007-11970 (hereinafter referred to as Reference Document 3).

  In any of the above references 1-3, it is determined whether or not the face is based on a face detection determination threshold. The face detection determination threshold indicates a threshold of face likelihood (face score). The face detection determination threshold value also includes a setting threshold value based on the correlation value with the face template image described above.

In the object detection unit 80, for example, the sub-window scans the entire image of the main image or the through image at a plurality of pixel intervals to generate a plurality of partial images. From the generated partial images, the front face detection unit and the side face detection unit may detect a partial image that is a front face or a side face.

  Further, when the detection mode switching means is provided, when the mode is switched to any one of the detection rate mode, the false detection suppression mode, the calculation speed mode, the front face detection mode, and the side face detection mode, the parameter control unit The detection determination threshold value for determining the detection performance of the face detection unit may be changed, and face detection may be performed with emphasis on detection performance in each detection mode in the side face detection unit and the front face detection unit.

  Further, the object detection unit 80 may use a technique disclosed in Japanese Patent Application Laid-Open No. 2007-94633 (hereinafter referred to as Reference Document 4).

  The object detection unit 80 can detect a motion vector from a through image or image data. The object detection unit 80 detects detection performance that determines the detection performance so that the captured face-through image or image data can perform the profile detection means based on the presence of a predetermined number of objects having a large amount of motion vector. You may control to change a threshold value.

  In addition, the object detection unit 80 can exhibit the profile detection means more effectively than the front face detection means based on the fact that the photographed through image or image data includes a predetermined number of objects with a large amount of motion vectors. In this manner, the detection determination threshold value for determining the detection performance may be changed.

  Further, as shown in FIG. 11, CCD-RAW data output from the analog signal processing unit is input, and for example, scratch correction, offset correction, gain correction, contrast correction, gamma correction, and resizing are performed on the input image data. Alternatively, digital signal processing such as noise reduction may be performed as preprocessing, and a detection result of a detection main body that detects a position that is equal to or greater than a face detection determination threshold value may be output from the signal processed image data.

Furthermore, as shown in FIG. 12, CCD-RAW data output from the analog signal processing unit is input, and for example, scratch correction, offset correction, gain correction, contrast correction, gamma correction, A detection main body that performs digital signal processing such as resizing (may be noise reduction processing after resizing processing) as preprocessing, and detects a position that is equal to or greater than a face detection determination threshold value from the signal processed image data. Output the detection result.

  Furthermore, as shown in FIG. 13, the CCD-RAW data output from the analog signal processing unit is input, and for example, scratch correction, offset correction, gain correction, contrast correction, gamma correction, Digital signal processing such as resizing, noise reduction, contour correction, etc. is performed as preprocessing, and the detection result of the detection body that detects a position that is equal to or greater than the face detection threshold value is output from the signal processed image data. May be. Here, various parameters used for the object detection processing such as gain in gain correction and a threshold value for setting a correlation value in the detection body are controlled by a parameter control unit 84 described later. Further, the parameters used for the object detection process are controlled separately from the parameters used for image quality correction applied to the display or recording image. Thus, for example, in order to improve the face detection performance for an image with a high degree of backlight, the displayed or recorded image is white even if the gain value used for gain correction in the face detection process is sufficiently large. There is nothing to skip.

  Also, as shown in FIG. 12, when setting the gain in the gain correction, it is controlled by the parameter control unit 84 in accordance with the white balance correction gain value obtained from the AE / AWB processing unit 63. Also good.

  The scene discriminating unit 82 discriminates a scene of the captured image, for example, backlight, night view, underwater or the like. The backlight determination process, which is one of the scene determination processes, will be described below with reference to FIGS. First, a rectangular area having a predetermined size, for example, a horizontal width and a vertical width, each of which is obtained by dividing the horizontal width and the vertical width of the entire image by 16 is defined as a unit photometry area. Scan two unit metering areas adjacent vertically, obliquely (NE, NW) onto a predetermined backlight judgment area in the image, and set the difference between the metering values (brightness) in the two unit metering areas When it is equal to or greater than the threshold value, it is determined that the light is backlit, and the maximum value of the difference between the photometric values is output as the degree of backlighting of the scene. Note that position information and the like of a portion where the difference in photometric values is maximized may be acquired and output at the same time.

  The scene determination unit 82 is not limited to the backlight determination process described above, and may be a brightness determination process, a determination process based on an EV value (exposure determination process), or a determination process based on a noise amount. These processes generally determine whether a dark image or a noisy image is a dark image because the accuracy of object detection decreases.

  For example, the brightness determination process determines whether the image is a dark image by determining the brightness of the pixel value from the through image or the main image.

  In the determination process based on the EV value, it is determined whether the image is dark based on the EV value that is an exposure value determined by the aperture value and the exposure time.

  In the determination process based on the noise amount, the noise amount specific to the CCD 58 serving as an image sensor is detected from the CCD-RAW to determine the noise amount.

  The parameter control unit 84 controls parameters used for the object detection process by the object detection unit 80 according to the scene determination result output from the scene determination unit 82. For example, when the scene determination unit 82 determines that the light is backlit, as shown in FIG. 7, a gain (detection gain) value in gain correction corresponding to the backlight level output from the scene determination unit 82 is calculated ( Step ST1) is determined within a range of preset upper and lower limits (steps ST2 to ST5).

  Further, not only the backlight level but also the EV value, brightness, and noise reduction amount may be determined. Gain (detection gain) in gain correction according to the EV value, brightness, and noise amount level May be calculated.

  The parameter may be a numerical value including at least one of a noise reduction amount of noise reduction used for the object detection processing unit, a contour correction degree of contour correction, and a gamma curve value of gamma correction. .

  For example, the parameter may be controlled so as to change the gamma curve of the gamma correction according to the backlight level (for example, when it can be determined that the backlight is back).

  Further, the parameter may be controlled in accordance with the degree of backlight (for example, when it can be determined that the light is back), so that the degree of contour correction for contour correction is increased. Further, the parameter may be controlled to increase the degree of noise reduction according to the degree of lightness (for example, when it can be determined that the image is dark).

  It should be noted that the degree of contour correction is set by numerical values or characters as to whether contour correction is performed (ON), contour correction is not performed (OFF), and how much contour correction is performed.

  Note that the degree of noise reduction is set by numerical values or characters to indicate whether noise reduction is performed (ON), noise reduction is not performed (OFF), and how much noise reduction is performed.

  The parameter control unit 84 can control the parameters more efficiently by referring to the signal generated by pressing the release button 19 and changing the control method at each stage of photographing, that is, at each processing timing. I have to. Specifically, when the release button 19 is pressed halfway, only when a face is detected using a backlight gain in the through image immediately before the pre-photographing, it is used for the face detection process in the through image. The gain value is set as the gain used for the face detection process for the pre-image, and the threshold value of the correlation value in the processing of the detection main body is set to a preset backlight threshold value.

  When setting the EV value threshold based on the EV value, only when the release button 19 is pressed halfway and a face is detected using the EV value gain in the through image immediately before the pre-shooting, The gain value used for the face detection process in the through image is set as the gain used for the face detection process for the pre-image, and the detection determination threshold value in the process of the detection main body is a preset EV. Set to the value threshold.

  When setting the lightness threshold based on lightness, when the release button 19 is half-pressed, the through image is used only when a face is detected using the lightness gain in the through image immediately before the pre-shooting. The gain value used for the face detection processing in the image is set as the gain used for the face detection processing for the pre-image, and the detection determination threshold value in the processing of the detection main body is a threshold value for brightness set in advance. Set to a value.

  When setting the threshold for noise based on the amount of noise, when the release button 19 is half-pressed, only when the face is detected using the noise gain in the through image immediately before the pre-shooting, the through The gain value used for the face detection process on the image is set as the gain used for the face detection process on the pre-image, and the detection judgment threshold value in the detection main body process is set for a preset noise. Set to the threshold.

  In addition, when the release button 19 is fully pressed, a face is detected using a backlight gain in a through image immediately before the pre-shooting, and when the auto-exposure mode and the flash emission prohibition mode are set during the main shooting. As long as the gain value used for the face detection process for the through image is set as the gain used for the face detection process for the main image, the detection judgment threshold (correlation The threshold value etc.) is set to a preset backlight threshold value.

  An EV value threshold value, a lightness threshold value, and a noise threshold value may be set, or a plurality of these threshold values may be used.

  The CPU 75 controls each part of the main body of the digital camera 1 according to signals from the operation system such as the operation mode switch 11 and various processing units such as the AF processing unit 62.

  The data bus 76 includes an image input controller 61, various processing units 62 to 65, a frame memory 66, a media control unit 67, an internal memory 68, a display control unit 71, an object detection unit 80, a scene determination unit 82, and a parameter control unit 84. , And the CPU 75 for exchanging digital image data and the like.

  Next, processing performed at the time of shooting in the digital camera 1 having the above configuration will be described. First, the CPU 75 determines the operation mode of the digital camera 1, and when it is in the playback mode, the CPU 75 performs playback processing for displaying the image data recorded on the external recording medium 70 on the liquid crystal monitor 18, and the operation mode is the shooting mode. In this case, a through image is displayed on the liquid crystal monitor 18 and monitoring of whether or not the release button 19 is half-pressed is started, and a countermeasure detection process for the through image is performed.

  FIG. 8 is a flowchart showing the parameter control process used by the parameter control unit 84 for the object (face) detection process for the through image. First, it is determined whether or not the backlight non-detected relief mode is on (step ST11). If the backlight non-detected relief mode is on, the value of noface_count indicating the number of times the face is not detected continuously is “0”. Is set (step ST13), the face detection result is acquired (step ST14), and it is determined whether or not a face is detected (step ST15). If a face is detected, the value of noface_count is set to “0” (step ST16), and the process returns to step ST14 to continue the face detection process. If no face is detected, the detection process is performed. It is determined whether or not the current gain used sometimes is a follow light gain (step ST17), and if not, the process proceeds to step ST26 described later. When the gain is for direct light, the value of noface_count indicating the number of times that a face is not continuously detected is greater than or equal to the value of th_noface_count_std indicating the normal gain continuous use threshold value in a state where a preset face is not detected It is determined whether or not (step ST18). If step ST18 is negative, the value of noface_count is incremented (step ST19), and the process returns to step ST14. If step ST18 is affirmed, a backlight determination process is performed by the scene determination unit 82 (step ST20), and it is determined whether the backlight level output from the scene determination unit 82 is equal to or greater than a set threshold value. (Step ST21). If step ST21 is negative, the process returns to step ST14. When it is determined that the backlight level is equal to or greater than the set threshold value, a backlight gain value corresponding to the backlight level output from the scene determination unit 82 is calculated (step ST22), and the calculated gain value is calculated. Is set as a gain value for backlight (step ST23). Further, the threshold value of the correlation value in the processing of the detection main body is set to be a preset backlight threshold value (step ST24). The value of noface_count is set to “0” (step ST16), and the process returns to step ST14.

  On the other hand, in step ST17, it is determined whether or not the current gain used in the detection process is a follow light gain. If the current gain is not a follow light gain, the number of times that a face is not continuously detected is indicated. It is determined whether or not the value of noface_count is equal to or greater than a value of th_noface_count_backlight indicating a threshold value for continuous use of backlight gain in a state where a preset face is not detected (step ST26). If step ST26 is negative, the value of noface_count is incremented (step ST27), and the process returns to step ST14. When step ST26 is affirmed, the value of the follow light gain is set as the gain value (step ST28), and the correlation value threshold in the processing of the detection body is set to the follow light threshold. (Step ST29), the value of noface_count is set to “0” (step ST30), and the process returns to step ST14.

  Subsequent to the object detection process using the parameters determined by the parameter control process described above, when the release button 19 is half-pressed, pre-photographing is performed to obtain a pre-image, and the object detection process for the pre-image is performed. Is done.

  FIG. 9 is a flowchart showing parameter control processing used by the parameter control unit 84 for object detection processing for a pre-image. First, it is determined whether or not a face has been detected in the through image immediately before the pre-photographing (step ST31). When a face is detected, it is determined whether or not the gain value used for the detection process in the through image immediately before the pre-photographing is a follow light gain (step ST32). The gain value is set so as to be the gain value used for the face detection process in the through image (step ST33), and the correlation value threshold value in the processing of the detection body is set to a preset backlight. Set to the threshold value for use (step ST34) and return.

  When step ST31 is affirmed and when step ST32 is affirmed, the process proceeds to step ST35, where the gain value is set to be a gain for direct light (step ST35). The threshold value of the correlation value at is set so as to be a threshold value for forward light set in advance (step ST36), and the process returns.

  Following the object detection process using the parameters determined by the above parameter control process, the AE / AWB processing unit 63 and the AF processing unit 62 perform AE / AWB and AF processing based on the pre-image, respectively, The flash controller 73 determines the light emission amount of the flash 24. Subsequently, the CPU 75 determines whether or not the half-press of the release button 19 is released. If the release button 19 is released, the CPU 75 displays a through image on the liquid crystal monitor 18 and starts monitoring whether or not the release button 19 is half-pressed. At the same time, the process returns to the countermeasure detection process for the through image. On the other hand, if the half-press of the release button 19 is not released, it is further determined whether or not the release button 19 is fully pressed. If not, the half-press of the release button 19 is released. Return to the determination. Further, when the release button 19 is fully pressed, the main image is shot (main shooting), and the main image acquired by the main shooting is set using the parameters controlled by the parameter control unit 84 as follows. On the other hand, an object detection process is performed.

  FIG. 10 is a flowchart showing the parameter control process used by the parameter control unit 84 for the object detection process for the main image. First, it is determined whether or not the backlight compatible face detection processing mode at the time of actual photographing is on (step ST41). When the backlight compatible face detection processing mode is on, it is determined whether or not the manual exposure mode is set (step ST42). If step ST42 is negative, it is further determined whether or not the flash has emitted (step ST43). If step ST43 is negative, it is determined whether or not a face is detected in the latest result of the through image (step ST44). When a face is detected, it is determined whether or not the gain value used in the detection process with the latest result of the through image is a gain for direct light (step ST45). The gain value is set so as to be the gain value used in the face detection process in the through image (step ST46), and the correlation value threshold value in the detection body process is set in advance. The threshold value for backlight is set (step ST47), and the process returns.

  In addition, when step ST41 is denied, when step ST42 is affirmed, when step ST43 is affirmed, when step ST44 is affirmed, and when step ST45 is affirmed, the process proceeds to step ST48. The value is set so that the gain value is the gain for direct light (step ST48), and the correlation value threshold value in the processing of the detection main body is a preset threshold value for direct light. (Step ST49) and return.

  After the object detection process using the parameters determined by the above parameter control process, image processing is performed on the main image acquired by the main photographing, and the processed main image is displayed on the liquid crystal monitor 18, Recording on the external recording medium 70. Then, it is determined whether or not the power is turned off. If the power is not turned off, monitoring of whether or not the release button 19 has been half-pressed is started, and a countermeasure detection process for the through image is performed. When the power is turned off, the process is terminated.

  As described above, in the embodiment of the present invention, the object detection processing is performed by controlling the parameters used in the object detection processing in accordance with the determination result obtained by determining the scene of the image. In this way, by controlling the parameters to parameters suitable for the detection process in the image based on the scene discrimination result, even if the shooting situation changes, the detection process from the image shot under the shooting situation Therefore, the object detection performance can be improved.

 In addition, by changing the parameter control method according to the timing at which the object detection means captures the image for which the object detection processing is performed, it is suitable for detection processing on the image based on the timing information at which the image was captured. Parameters can be determined efficiently.

  In the above-described embodiment, face detection has been described as an example of the predetermined object detection process. However, the present invention is not limited to this, and a subject other than the face may be used.

  The digital camera according to the embodiment of the present invention has been described above, but the scope of application of the present invention is not limited to this. For example, other electronic devices having an electronic imaging function, such as a mobile phone with a camera and a PDA with a camera It is applicable to.

The perspective view centering on the back surface of the digital camera which is one Embodiment of the imaging device of this invention The perspective view centering on the front of the digital camera which is one Embodiment of the imaging device of this invention 1 is a block diagram showing the configuration of a digital camera that is an embodiment of a photographing apparatus of the present invention (part 1); The block diagram which shows the target object detection process by the target object detection part 80 The figure for demonstrating the backlight determination process of the scene discrimination | determination part 82 The figure for demonstrating the backlight determination process of the scene discrimination | determination part 82 The flowchart which shows the process which calculates the value of the gain according to the backlight level The flowchart which shows the parameter control process used for the target object detection process with respect to a through image by the parameter control part 84 The flowchart which shows the parameter control process used for the target object detection process with respect to the pre image by the parameter control part 84 The flowchart which shows the parameter control process used for the target object detection process with respect to this image by the parameter control part 84 Block diagram showing the configuration of a digital camera which is an embodiment of the photographing apparatus of the present invention (No. 2) Block diagram showing the configuration of a digital camera which is an embodiment of the photographing apparatus of the present invention (part 3) Block diagram showing the configuration of a digital camera which is an embodiment of the photographing apparatus of the present invention (part 3)

Explanation of symbols

1 Digital camera 80 Object detection unit
82 Scene discrimination part 84 Parameter control part

Claims (6)

A face detection unit, a scene determination unit, and a parameter control unit;
For the through image acquired before the release button is pressed, the scene determination unit determines whether the through image is a backlight scene, and the parameter control unit uses the scene determination unit to detect the backlight scene. Is determined as a parameter used for processing for detecting a face from the through image, and when the scene determination unit determines that the scene is not a backlight scene, A parameter for forward light is determined as a parameter used for processing for detecting a face from the through image, and the face detection unit includes a plurality of parameters having different values including the determined parameter determined by the parameter control unit. To detect a face from the through image, and integrate the detection results into a face detection result,
For the main image acquired when the release button is fully pressed, the parameter control means determines whether or not the flash has been emitted when acquiring the main image, and the flash has not been emitted in the determination. If it is determined, it is further determined whether or not a face is detected in the face detection processing performed on the through image, and it is determined that a face is detected in the determination, and the through image If the parameter used in the face detection process for the image is a backlight parameter, the backlight parameter is determined as a parameter used for the process of detecting a face from the main image, and the flash is emitted If it is determined that the flash is emitted in the determination of whether or not the face is detected in the determination of whether or not the face is detected If it is determined that an off determines the parameters for forward light as a parameter to be used in the process of detecting a face from the main image, the face detection unit, after decision determined by the parameter control means An imaging apparatus, wherein a face is detected from the main image using a plurality of parameters including different parameters, and the detection results are integrated into a face detection result . Said parameter control means, said when the face is not detected by the parameter by the face detection means, image taking apparatus according to claim 1, characterized in that for changing the parameters. The parameters, imaging apparatus according to claim 1 or 2, wherein the a detection determination threshold used to gain or the face detecting unit in the gain correction to be used for the face detection means. The parameter is the degree of contour correction in the contour correction to be used in the face detecting means, said face degree of noise reduction in the noise reduction to be used for the detection unit, the gamma curve value in the gamma correction used in the face detection means The imaging device according to claim 1, comprising at least one of the following. With respect to the through image acquired before the release button is pressed, it is determined whether or not the through image is a backlight scene. If it is determined in the determination that the through image is a backlight scene, a parameter for backlight is determined. Is determined as a parameter used for processing to detect a face from the through image, and if it is determined in the determination that the scene is not a backlight scene, a parameter for forward light is used for processing to detect a face from the through image. A procedure for determining a face to be used, detecting a face from the through image using a plurality of parameters having different values including the determined parameter after determination, and integrating the detection results into a face detection result When,
For the main image acquired when the release button is fully pressed, it is determined whether or not a flash has been emitted when the main image is acquired, and if it is determined in the determination that the flash has not emitted, It is further determined whether or not a face is detected in the face detection process performed on the through image, and when it is determined that a face is detected in the determination, and in the face detection process on the through image When the used parameter is a parameter for backlighting, the parameter for backlighting is determined as a parameter used for processing for detecting a face from the main image, and the flash is used in determining whether or not the flash is emitted. When it is determined that the light is emitted, and when it is determined that the face has not been detected in the determination of whether or not the face has been detected. The parameters for forward light determined as parameters used in the processing of detecting a face from the main image, a face from the present image using the parameter of the plurality of different values including the parameter after the decision has been 該決 constant And a procedure for detecting and integrating the detection results to obtain a face detection result . Computer
A program for functioning as a face detection means, a scene determination means, and a parameter control means,
For the through image acquired before the release button is pressed, the scene determination unit determines whether the through image is a backlight scene, and the parameter control unit uses the scene determination unit to detect the backlight scene. Is determined as a parameter used for processing for detecting a face from the through image, and when the scene determination unit determines that the scene is not a backlight scene, A parameter for forward light is determined as a parameter used for processing for detecting a face from the through image, and the face detection unit includes a plurality of parameters having different values including the determined parameter determined by the parameter control unit. To detect a face from the through image, and integrate the detection results into a face detection result,
For the main image acquired when the release button is fully pressed, the parameter control means determines whether or not the flash has been emitted when acquiring the main image, and the flash has not been emitted in the determination. If it is determined, it is further determined whether or not a face is detected in the face detection processing performed on the through image, and it is determined that a face is detected in the determination, and the through image If the parameter used in the face detection process for the image is a backlight parameter, the backlight parameter is determined as a parameter used for the process of detecting a face from the main image, and the flash is emitted If it is determined that the flash is emitted in the determination of whether or not the face is detected in the determination of whether or not the face is detected If it is determined that an off determines the parameters for forward light as a parameter to be used in the process of detecting a face from the main image, the face detection unit, after decision determined by the parameter control means A program which detects a face from the main image using a plurality of parameters including different parameters and integrates the detection results into a face detection result .

Images