JP2010114752A - Device and method of imaging and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To achieve smooth follow up in shooting a subject while panning a camera, by distinguishing a figure and suppressing correction of blurring of images with respect to a main subject in the case where the camera having a function to correct blurring of images due to hand movement by vector detection performs correction of blurring. <P>SOLUTION: The imaging device divides an image into several block areas, detects a motion vector for each of the divided block areas, detects a face of a person from the image. When a face is detected by face detection, it selects specified block areas which are positioned in the face and its surroundings from the divided several block areas, performs predetermined weighting with respect to the frequency of the motion vectors in the selected block areas, distributes the motion vector in each block area on a frequency distribution while setting the size of the motion vectors as classes, on the basis of the weighted frequency, determines a representative motion vector on the basis of the level of frequency and corrects the image on the basis of the representative motion vector. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an imaging apparatus, an imaging method, and a program that include a shake correction apparatus that detects a motion vector and corrects camera shake.

  In an imaging apparatus such as a video camera that captures a moving image, there is a problem that the image is blurred due to camera shake, particularly when the lens is zoomed to the telephoto side. In order to prevent such image blur due to camera shake, a technique has been proposed in which a motion vector of an image is detected from an image signal taken in the past and the image blur is corrected based on the motion vector. In addition, there is a technique for detecting a part of a human face in the screen in order to focus on the face of the person who wants to take the picture during shooting.

  Among techniques for detecting motion vectors and correcting image blurring, a correlation method based on correlation calculation, a block matching method, and the like have been known as methods for detecting a motion vector of an image.

  In the block matching method, an input image signal is divided into a plurality of block areas (for example, 8 pixels × 8 lines) having an appropriate size. Then, the difference between the previous field (or frame) and a certain range of pixels is calculated in units of blocks, and the block of the previous field (or frame) that minimizes the sum of the absolute values of the differences is searched. And the relative shift | offset | difference between screens has shown the motion vector of the block.

  For matching operations, Morio Onoe et al., Information Processing Vol. 17, no. 7, p. 634-640 July 1976.

  An example of a conventional motion vector detection method using the block matching method will be described with reference to the drawings. FIG. 11 is a schematic block diagram of an apparatus for preventing shake by a conventional motion vector detection method.

  First, an image signal (field or frame) as a motion vector detection target is applied to the image memory 9 and the filter 2 for extracting a spatial frequency component. The image memory 9 temporarily stores image signals. The filter 2 extracts a spatial frequency component useful for motion vector detection from the image signal. That is, the low spatial frequency component and the high spatial frequency component of the image signal are removed.

  The image signal that has passed through the filter 2 is added to the binarization circuit 3. The binarization circuit 3 binarizes the image signal with reference to the zero level. Specifically, the sign bit of the output signal is output. The binarized image signal is applied to the correlation calculation circuit 4 and the memory 5 as one field period delay means. Further, the image signal of the previous field is added from the memory 5 to the correlation calculation circuit 4.

  In accordance with the block matching method, the correlation calculation circuit 4 divides the image area into block areas of an appropriate size as described above, performs correlation calculation between the current field and the previous field in units of blocks, and moves the correlation value of the result Add to the vector detection circuit 6. The motion vector detection circuit 6 detects a block-based motion vector from the calculated correlation value. Specifically, the block of the previous field having the smallest correlation value is searched, and the relative shift is used as the motion vector. The block-based motion vector is added to the motion vector determination circuit 7. The motion vector determination circuit 7 determines the entire motion vector from the motion vector in units of blocks. Specifically, a motion vector in units of blocks is set to a vertical and horizontal vector, and a histogram is created with a magnitude frequency for each of the vertical and horizontal motion vectors.

  10A is a captured image, FIG. 10B is a motion vector for each detected block, and FIG. 10C is a vertical and horizontal histogram of the detected motion vector. In this way, a histogram is created, and the place where the frequency reaches a peak is used as the entire motion vector.

  The motion vector determination circuit 7 adds the determined motion vector of the entire image to the memory read control circuit 8. The memory readout control circuit 8 controls the readout position of the image memory 9 so that the motion of the image is canceled in accordance with the overall motion vector, and an image signal in which the blur is corrected is output from the image memory 9.

  However, in the above method, when shooting while chasing a moving person, the frequency with respect to the background vector increases if the ratio of the person to the screen is small. As a result, the read position of the image memory changes so that the background remains at the original position in the screen, which changes the position of the area in the screen that should not move, so the screen was pulled to the background It moves like this.

  As a method for solving this problem, Patent Document 1 discloses the following method. That is, by using an identification method in which the smaller absolute value of the motion vector is a motion vector to be corrected and the larger one is a moving object, a motion that is pulled by the background is avoided.

  As a panning detection method using motion vectors, Patent Document 2 discloses a method for determining whether or not the size of a motion vector in each detection region is substantially constant and continues for a predetermined time.

Next, as a technique related to face detection, a method of performing focus detection by setting a focus detection area including a face area recognized in Patent Document 3 is disclosed.
Japanese Patent No. 2956056 Japanese Patent No. 3039669 JP 2006-227080 A

  The method disclosed in Patent Literature 1 is effective when the shake of the imaging device is relatively small, or when the image generated by the imaging device is smaller than the speed of the moving object. For example, this may be the case when the focal length of the imaging optical system that is mounted on the imaging device or is provided in the imaging device is relatively short.

  However, the magnitude of camera shake varies depending on the shooting environment, the skill level of the photographer, and the like. In addition, the image generated by the shake of the image pickup apparatus increases in proportion to the focal length of the image pickup optical system. Further, when the person moves slowly, the background motion vector itself becomes smaller. For this reason, if the motion vector generated by the movement of the imaging device that occurs when the person is chased (panning operation) and the motion vector due to camera shake of a moving subject such as a person are identified only by the magnitude of the motion vector, an error occurs. Becomes larger. As a result, the movement of the subject becomes very unsightly because the movement of the actual subject and the background cannot be distinguished.

  Further, Patent Document 2 describes that the image stabilization function is stopped because the camera itself is moving at the time of panning detection. However, when the subject is chased by the camera, blur correction for the subject is performed. It will not be broken.

  Therefore, the present invention has been made in view of the above-described problems, and when shooting a subject while chasing it with a camera, a person is identified, blur correction on the main subject is suppressed, and as a result, smooth tracking can be performed. The purpose is to do.

  In order to solve the above problem, an imaging apparatus according to the present invention divides an image into a plurality of block regions, detects a motion vector for each of the divided block regions, and detects human motion from the images. A face detection unit for detecting a face, and a selection for selecting a predetermined block region located in the face and its peripheral portion from the plurality of divided block regions when a face is detected by the face detection unit Means for setting a predetermined weight for the frequency of the motion vector of the block area selected by the selection means, and the frequency set by the setting means with the magnitude of the motion vector as a class Classifying means for distributing the motion vector for each block area on the frequency distribution based on the frequency distribution, and representative motion based on the magnitude of the frequency It is characterized by having a determining means for determining a vector, and blur correction means for correcting the blur of the image based on the representative motion vector.

  In order to solve the above problem, the imaging method of the present invention divides an image into a plurality of block areas, and detects a motion vector for each of the divided block areas. A face detection step for detecting a human face, and when a face is detected by the face detection step, a predetermined block region located on the face and its peripheral portion is selected from the divided block regions A setting step for setting a predetermined weight for the frequency of the motion vector of the block area selected by the selection step, and setting the weight by the setting step, with the magnitude of the motion vector as a class A classification step of allocating the motion vector for each block area on the frequency distribution based on the frequency, and on the basis of the size of the frequency A determination step of determining a table motion vectors, is characterized by having a blur correction step of correcting the blur of the image based on the representative motion vector.

  As described above, according to the present invention, when a face is detected, the motion with respect to the person is weighted by weighting the frequency with respect to the position of the face and the magnitude of the motion vector of the area corresponding to the body below the face. It becomes easy to detect a vector. In addition, camera shake can be corrected particularly during panning, and it is possible to perform shooting as intended by the photographer.

  As a result, when the subject is photographed while chasing with the camera, the subject is identified, blur correction for the main subject is suppressed, and as a result, smooth tracking can be performed.

Example 1
Hereinafter, the present invention will be described in detail with reference to the drawings. FIG. 1 shows the configuration of an image pickup apparatus having features of the present invention.

  In FIG. 1, reference numeral 101 denotes an image pickup optical system including a lens and a diaphragm, 102 denotes a solid-state image pickup element such as a CCD or CMOS sensor, and 103 denotes a signal processing circuit that performs signal processing necessary for a captured image. Reference numeral 104 denotes an image area dividing unit that divides the image signal-processed by the signal processing circuit 103 into a plurality of block areas, and reference numeral 105 denotes a motion vector detecting unit that detects a motion vector in each of the divided image areas. Reference numeral 109 denotes a face detection unit that detects the position of the face from the image, and performs a known face recognition process on the image to detect the face area of the person in the shooting screen. As face recognition processing, for example, there is a method of extracting a skin color region from the gradation color of each pixel represented by an image and detecting the face with a matching degree with a face contour plate prepared in advance. Also disclosed is a method of performing face detection by extracting facial feature points such as eyes, nose and mouth using a well-known pattern recognition technique. Reference numeral 110 denotes an image region selection unit that selects a predetermined image region including the face from the position of the face on the image whose face position is detected from the image by the face detection unit 109.

  A weight setting unit 106 performs predetermined weighting on the motion vector of the image area selected by the image area selection unit 110. A histogram creation unit 107 classifies the motion vectors for each image area by classifying the vectors by classifying the motion vectors on the frequency distribution based on the frequencies weighted by the weight setting unit 106. Reference numeral 108 denotes a motion vector determination unit that determines a representative motion vector from the motion vectors distributed on the frequency distribution by the histogram creation unit 107. Reference numeral 112 denotes a memory read control unit that performs control for determining the range of an image to be recorded or displayed from the motion vector determined by the motion vector determination unit 108. Reference numeral 111 denotes an image memory for storing signal-processed image data, and reference numeral 113 denotes a recording unit for recording the image data on a recording medium such as a memory card or a hard disk.

  An area for weighting in this embodiment will be described with reference to FIG.

  In the example of FIG. 2, when a subject is present on the screen and the subject is moving to the right, the video camera is moved to the right according to the subject (in a panning state). To do. In addition, when moving images are captured, images are taken into the video camera at a specific cycle, so that motion vectors can be detected from images of the previous frame and the current frame (or the previous field and the current field are also possible).

  Reference numeral 201 denotes a block (vector detection area) divided by the image area dividing unit 104, and a motion vector between fields (between frames) is acquired in units of blocks. A face detection frame 202 is a detection frame of a face detected from an image by the face detection unit 109. A setting frame 203 is a setting frame selected by the image area selection unit 110, and a face detection area and a lower part thereof are selected. The reason why the lower part of the face is also selected is to select the region including the face and the part corresponding to the torso because the person's torso is under the face. The size of the setting frame is about twice the size of the face in the horizontal direction and about four times the size of the face in consideration of the balance between the standard face and body. The reason why only the face portion is not set as the selection region is that the ratio of the selection region to the total number of blocks is reduced only in the face region, and the motion vector detection accuracy for a necessary subject is lowered.

  FIG. 3 shows a flowchart of the weighting change of the present invention. Hereinafter, the vector detection operation according to the present invention will be described with reference to FIG.

  In FIG. 3, in step S300, the face detection unit 109 is used to detect the face position from the image.

  In step S301, if a face is detected in the image in step S300 by the image region selection unit 110, it is determined that a face has been detected, and the process proceeds to step S302. If no face is detected in the image in step S300, the image area selection unit 110 determines that no face is detected from the image, and the process proceeds to step S306.

  In step S302 in FIG. 3, the size of the setting frame is determined by the image area selection unit 110 based on the detected face size. Note that the setting frame selected by the image area selection unit 110 corresponds to the setting frame 203 in FIG.

  Next, after an area is selected in step S302, the weighting amount in the setting frame is changed by the weighting setting unit 106 in step S303. This weighting amount is empirically about four times the weight from the size in the screen when chasing a person. However, if the same weighting is performed when the size of the setting frame is large and small, if the setting frame is small, the frequency for determining the motion vector of the main subject is necessary for detection even if the weighting is performed. There is a possibility that the amount will not increase. Therefore, when the setting frame is small, that is, when the number of motion vector detection frames in the selected region is small, the weighting amount may be increased. By this operation, the frequency for the magnitude of the motion vector necessary for creating the histogram can be secured.

  In step S304, a histogram is created based on the frequency weighted by the histogram creation unit 107. FIG. 4 is a diagram showing how the frequency on the histogram changes with respect to the magnitude of the motion vector when the present invention is applied and weighting is performed. First, a plurality of movement vectors are detected as movement vectors of each block (vector detection area). Thereafter, as shown in FIG. 4A, the frequency distribution (frequency) of each class can be obtained by sorting and plotting the frequency distribution corresponding to the frequency corresponding to the number of vectors on the frequency distribution. As the class value of each class having a range, an intermediate value between the upper limit value and the lower limit value of the class may be used. In addition, since processing in the Y direction (vertical screen direction) and processing in the X direction (horizontal direction in the screen) perform the same processing, only the processing in the X direction will be described below as an example.

  Before weighting as shown in FIG. 4A, the histogram of the motion vector for the main subject has a low frequency, and the representative vector obtained from this histogram is the vector for the motion of the background area shown in FIG. On the other hand, FIG. 4B weights the face and its peripheral part. In this case, since the frequency of the main subject portion increases, the detected representative vector can be a motion vector for the original subject. In this way, by selecting a predetermined area including the face and changing the weighting setting according to the size of the area, it is possible to perform shake correction on the main subject without depending on the size of the main subject. It becomes possible.

  In step S305, the motion vector determination unit 108 determines a representative vector using the histogram created in step S304.

  If no face is detected from the image in step S300, the process proceeds from step S301 to step S306. In step S306, the image area selection unit 110 invalidates the setting frame for weighting.

  In step S307, the weighting setting unit 106 resets the weighting amount. As a result, when a face is not detected, a conventional histogram without weighting as shown in FIG. 4A is created, and a representative motion vector is determined therefrom.

  As described above, the face area detected from the image using face detection and the area corresponding to the body are selected (setting frame is set), and the frequency for the magnitude of the vector of the selected area is weighted. By performing the above, it becomes possible to detect the motion vector for the main subject with high accuracy. As a result, it is possible to correct only the shake with respect to the subject as intended by the photographer.

(Example 2)
FIG. 5 is a block diagram showing a second embodiment of the present invention. In FIG. 5, the same components as those in FIG.

  Reference numeral 501 denotes a panning detection unit. The panning detection unit 501 observes temporal changes in vectors with respect to each motion vector detection region, and determines that a panning operation is performed when there are a predetermined number or more of motion vectors in the same direction in the motion vector detection region. .

  Reference numeral 502 denotes a face detection position determination unit. The face detection position determination unit 502 determines whether or not the face detected from the image by the face detection unit 109 is within a predetermined region for a predetermined time. When a face is detected during panning detection and the position of the detected face is within a predetermined area for a predetermined time, the probability that the user is shooting while chasing a person is increased. For this reason, in this embodiment, panning information and face detection information are used to reflect the photographer's intention more accurately and to follow the subject.

  FIG. 6 shows a flowchart of the main control unit of the present invention. The second embodiment of the present invention will be described below with reference to FIGS.

  In step S600, the panning detection unit 501 determines whether panning is detected. If panning determination is made here, the process proceeds to step S601. Step S601 is a part where the face detection unit 109 detects a face from an image. If a face is detected in the image, it is determined in step S602 that a face has been detected, and the process proceeds to step S603. In step S603, it is determined whether or not the face detected by the face detection position determination unit 502 is within a predetermined area.

  Furthermore, in step S604, it is similarly determined whether the face detection position determination unit 502 detects the face from the image for a predetermined time or more. The determination in step S603 and step S604 is a determination as to whether the face detected from the image is in a stopped state in the screen even though it is in the panning state. If this condition is satisfied, chasing shooting is performed. Can be considered. When the conditions are met, the image area selection unit 110 determines the size of the image setting frame from the detected face size in step S605.

  After selecting an area in step S605, the weighting amount in the setting frame is changed in step S606. The weighting amount here is the same as in the first embodiment. In step S607, a histogram is created based on the weighted frequency as in the first embodiment, and a representative vector is determined in step S608.

  If panning is not detected in step S600, a face is not detected in step S602, or if predetermined conditions are not met in steps S603 and S604, a setting frame is set in step S609. To disable. In step S610, the weighting amount is reset, thereby creating a conventional histogram and determining a representative vector.

  As described above, when panning is detected, a face is detected on the screen, and by confirming that the face position is within a predetermined area in the screen for a predetermined time, the photographer can It is possible to determine with high accuracy that the chase shooting is being performed. As a result, a photographed image as intended by the photographer can be obtained.

(Example 3)
In the third embodiment, after detecting a face and performing weighting, an operation in the case where face detection cannot be performed, for example, when a person turns backward will be described.

  FIG. 7 is a block diagram showing a third embodiment of the present invention. In FIG. 7, the same components as those in FIG.

  In FIG. 7, the vector area selection unit 701 according to the third embodiment is the same block area selection unit as the image area selection unit 110 in FIG. 1, and can also specify the block area of the body part. A color detection and storage unit 702 detects the color in the block region of the body portion selected by the vector region selection unit 701 and stores it. A color monitoring unit 703 monitors whether the color stored in 702 has changed with respect to the position of the body selected by the vector region selection unit 701. Here, when the face is detected, the color of the torso, that is, the color of the clothes, is memorized, and even if the person who is the subject turns back and the face cannot be detected, the subject is determined based on the color of the clothes. It is possible to shoot a chasing without a sense of incongruity by determining whether or not the camera is in the middle.

  FIG. 8 shows a flowchart of the main control unit of the present invention. In the following, a third embodiment of the present invention will be described with reference to FIGS.

  In FIG. 8, step S <b> 800 is a part in which the face detection unit 109 detects a face from an image. In step S801, it is determined whether a face is detected from the image. If a face is detected from the image, the process proceeds to step S802. In step S802, the size of the setting frame is determined by the vector area selection unit 701 from the size of the face detected from the image.

  After selecting an area in step S802, in step S803, the color detection and storage unit 702 detects and stores the color of the area corresponding to the human torso portion. In step S804, the weighting setting unit 106 changes the weighting amount in the setting frame.

  The setting of the weighting amount in the weighting setting unit 106 is the same as in the first embodiment. In step S805, a histogram is created based on the frequency weighted by the histogram creation unit 107. In step S806, the representative vector is determined by the motion vector determination unit 108.

  If no face is detected in step S801, it is determined in step S807 whether or not weighting is set. If the weighting has already been set, the face cannot be detected because the subject turns backward after the face is detected once.

  In step S808, the color monitoring unit 703 determines whether the color corresponding to the body has changed from the color stored in the color detection and storage unit 702. If the color has not changed, the process proceeds to step S804, and the weighting amount remains changed. By this operation, even if the face is not detected in the middle of shooting, it is possible to perform chasing shooting as intended by the photographer by detecting the color of clothes of the subject.

  If no weighting is performed in step S807, or if it is determined in step S808 that the color of the body portion has changed, it is determined that there is no subject to be chased. In step S809, the setting frame is invalidated, and the weighting amount is reset in step S810, whereby a conventional histogram is created and a representative vector is determined.

  As described above, when shooting while chasing a person, the color of the part corresponding to the body (for example, the color of the clothes) is monitored even if the face is not detected, and the setting frame until the color changes Do not change the weighting amount. Thereby, it is possible to perform photographing intended by the photographer.

Example 4
The configuration diagram showing the fourth embodiment of the present invention is the same as FIG. As a fourth embodiment, the operation when the detected face is very small relative to the image area will be described. If the detected face is very small, it is considered that the person is moving or the subject's movement itself is being taken while photographing a landscape, not following the person. Therefore, when the face is very small relative to the image area, it is desirable not to perform weighting.

  FIG. 9 shows a flowchart of the main part of the fourth embodiment. In FIG. 9, in step S900, the face detection unit 109 detects a face from the image. If a face is detected in the image, it is determined in step S901 that a face has been detected, and the process proceeds to step S902.

  In step S <b> 902, the size of the setting frame is determined by the image area selection unit 110 based on the detected face size. After selecting an area in step S902, it is next determined in step S903 whether the size of the setting frame is greater than or equal to a predetermined value. If it is greater than or equal to the predetermined value, the weighting amount in the setting frame is changed in step S904. This weighting amount is the same as in the first embodiment. In step S905, a histogram is created based on the weighted frequency, and a representative vector is determined in step S906.

  If no face is detected in step S900, and if the size of the setting frame is less than the predetermined value in step S903, the process proceeds to step S907. In step S307, the setting frame for weighting is invalidated, In step S908, the weighting amount is reset. As a result, when a face is not detected from the image or when a face is detected from the image but the size is very small, a conventional histogram is created, and a representative motion vector is determined therefrom. It will be.

  As described above, even if a face is detected from an image, if the size is very small, it is possible to perform shooting as intended by the photographer by not performing weighting.

(Other examples)
As another embodiment, an operation when a plurality of faces are detected in an image area will be described. In the other embodiments, the configuration diagram is the same as that in FIG.

  When a plurality of faces are detected, the weights for the face located at the center of the image and the body part below the face are basically changed. Usually, the photographer places the subject he / she wants to shoot at the center of the screen. Therefore, when a plurality of faces are detected, weighting is performed using the face closest to the center, thereby making it possible to perform photographing that best matches the photographer's intention.

  In addition, this invention can be achieved also with the following structures.

  It is assumed that a storage medium storing software program codes for realizing the functions of the above-described embodiments is supplied to a computer or an apparatus. Thereby, the control unit of the computer or apparatus can also be achieved by reading and executing the program code stored in the storage medium. In this case, the program code itself read from the storage medium realizes the function of the above-described embodiment, and the storage medium storing the program code of the software realizing the function of the above-described embodiment constitutes the present invention. Will do.

  As a storage medium for supplying the program code, for example, a flexible disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD-R, DVD-ROM, magnetic tape, nonvolatile memory card, ROM, or the like is used. be able to.

It is a block diagram of the 1st Example of this invention. It is the figure which showed the area | region which performs weighting. It is the flowchart which showed operation | movement of the principal part of the 1st Example of this invention. It is the figure which showed the change of the histogram at the time of changing weighting. It is a block diagram of the 2nd Example of this invention. It is the flowchart which showed operation | movement of the principal part of the 2nd Example of this invention. It is a block diagram of the 3rd Example of this invention. It is the flowchart which showed operation | movement of the principal part of the 3rd Example of this invention. It is the flowchart which showed the operation | movement of the principal part of the 4th Example of this invention. It is the figure which showed the conventional histogram creation method. It is a block diagram of the conventional image correction system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 201 Block divided | segmented 202 Face detection frame 203 Set block area | region 501 Panning detection part 502 Face detection position determination part 702 Color detection and memory | storage part 703 Color monitoring part

Claims (11)

A vector detecting means for dividing an image into a plurality of block areas and detecting a motion vector for each of the divided block areas;
Face detection means for detecting a human face from the image;
A selection means for selecting the face and a predetermined block area located in the periphery thereof from the plurality of divided block areas when a face is detected by the face detection means;
Setting means for setting a predetermined weight for the frequency of the motion vector of the block area selected by the selection means;
Classifying means for assigning the motion vector for each of the respective block areas on the frequency distribution based on the frequency set by the setting means with the magnitude of the motion vector as a class,
Determining means for determining a representative motion vector based on the magnitude of the frequency;
An image pickup apparatus comprising: a shake correction unit that corrects a shake of an image based on the representative motion vector. Panning detection means for detecting whether it is in a panning state;
Position determination means for determining whether or not the position of the face detected by the face detection means has moved,
When the panning is detected by the panning detection unit and the position determination unit determines that the position of the face is not moved, the selection unit selects the face and the face from the plurality of divided block areas. The imaging apparatus according to claim 1, wherein a predetermined block region located in the peripheral portion is selected.   The imaging apparatus according to claim 1, wherein the area selected by the selection unit is a face area detected by the face detection unit and a block region of a torso portion corresponding to the face.   4. The imaging apparatus according to claim 1, wherein the size of the area selected by the selection unit changes based on the size of the detected face.   5. The imaging apparatus according to claim 1, wherein the weighting amount set by the weighting setting unit varies depending on a size of an area selected by the selection unit. Color detecting means for detecting the color of an image of a predetermined block area located in the peripheral part of the face detected by the face detecting means;
Monitoring means for monitoring whether the color detected by the color detection means has changed,
2. The weighting amount set by the weight setting unit is changed when it is detected by the monitoring unit that the color has changed after the face detection unit no longer detects the face. 6. The imaging device according to any one of items 5 to 5.   7. The method according to claim 1, wherein when there are a plurality of faces detected by the face detection means, the weight setting means weights the face in the center of the image and the surrounding area. The imaging device according to any one of the above. 7. The imaging apparatus according to claim 1, wherein when the face size is smaller than a predetermined size, weighting by the weight setting unit is not performed. A vector detection step of dividing an image into a plurality of block regions and detecting a motion vector for each of the divided block regions;
A face detection step of detecting a human face from the image;
When a face is detected by the face detection step, a selection step of selecting a predetermined block region located on the face and its peripheral portion from the divided block regions;
A setting step for setting a predetermined weight for the frequency of the motion vector of the block region selected by the selection step;
A classifying step of classifying the motion vectors for the respective block areas on the frequency distribution based on the frequency set by the setting step, with the magnitude of the motion vector as a class;
A determination step of determining a representative motion vector based on the magnitude of the frequency;
A blur correction step of correcting a blur of the image based on the representative motion vector.   A program for causing a computer to execute the imaging method according to claim 9.   A computer-readable storage medium storing the program according to claim 10.

Images