JP4621991B2 - Image blur correction apparatus and correction method thereof - Google Patents

Description

  The present invention relates to an image shake correction apparatus and a correction method therefor, and more particularly to an image shake correction apparatus that corrects an image shake caused by a camera shake or the like that occurs at the time of taking an image of a single frame, such as taking an image (still image) with a digital camera. It relates to the correction method.

  Various correction techniques have been proposed for camera shake, which is one of the causes of shooting failures with digital cameras. For example, in Patent Documents 1 and 2, when images for one frame are captured, images for a plurality of frames are continuously captured. Then, it has been proposed to reduce image blur due to camera shake or the like by shifting and synthesizing the acquired images of a plurality of frames by the amount of image blur (pixel shift synthesis).

In Patent Document 3, a correction lens disposed in a photographic optical system is driven, and the image blur is reduced by changing the relative position between the subject image and the image pickup means by the amount of image shake within the imaging plane. Has been proposed.
JP 2000-341577 A JP-A-9-261526 Japanese Patent No. 3530696

  However, even if the image blur between the images of a plurality of frames is corrected by pixel shifting synthesis as in the cited references 1 and 2, the influence of the image blur that has occurred during the exposure period when each image is imaged is removed. I can't. For this reason, there has been a problem that even if pixel-shifted composition of images for a plurality of frames is performed, the sharpness of each image can be improved only to some extent because the sharpness of each image is low.

  In addition, when the image blur correction is performed by driving the correction lens as in the cited document 3, there are problems such as an increase in power, an increase in cost, and a correction failure of high frequency components due to the influence of inertia of the correction lens.

  The present invention has been made in view of such circumstances, and an object thereof is to provide an image blur correction apparatus and a correction method thereof that can capture a still image with high sharpness using an inexpensive apparatus.

  In order to achieve the above object, an image shake correction apparatus according to a first aspect of the present invention is an image pickup means for picking up an image formed by a photographing optical system, and a plurality of frames of images are continuously captured by the image pickup means at a predetermined shutter speed. Imaging control means for acquiring images for a plurality of frames by imaging, image blur detection means for detecting image blur of an image formed by the photographing optical system, and each acquired by the imaging control means Image blur correction comprising: an image compositing unit that generates an image for one frame in which image blur is corrected by synthesizing a frame image by shifting by an amount corresponding to the image blur detected by the image blur detection unit. In the apparatus, for each frame image acquired by the imaging control unit and used to synthesize an image for one frame generated by the synthesis unit, each pixel is sequentially set as a target pixel. By setting a value obtained by weighted average of a pixel value of each pixel group including a plurality of pixels in a predetermined range including the eye pixel and a predetermined weighting coefficient as a new pixel value of the target pixel, Weighting conversion means for weight-converting the image of each frame, and weighting coefficients used for weight conversion of the image of each frame in the weighting conversion means, in the exposure period when the image of each frame is imaged by the imaging means Weighting coefficient determining means for determining based on the image blur detected by the image blur detecting means.

  According to the present invention, when images of a plurality of frames are imaged to correct image blur and pixel shift is combined, the images of each frame are taken into consideration in the image blur during the exposure period of each frame image. Weighted conversion is performed, and the converted images of each frame are combined by shifting pixels. Therefore, even if the sharpness of the original image is low, an image with high sharpness can be obtained. Further, since image blur correction is performed without using an optical member such as a correction lens, the apparatus is inexpensive.

  According to a second aspect of the present invention, in the image blur correction apparatus according to the first aspect, the weighting coefficient determination unit is configured to determine the pixel of interest at the start of exposure when the image of each frame is captured by the imaging unit. It is characterized in that it is determined on the basis of the trajectory that the image point imaged in the period moves during the exposure period and the length of time existing in each pixel of the pixel group.

  The present invention shows an aspect of a method for determining a weighting coefficient used for weighting conversion.

  According to a third aspect of the present invention, in the image blur correction device according to the first or second aspect of the invention, the weight conversion unit is configured to perform the attention according to a shutter speed when an image of each frame is captured by the imaging unit. The size of the pixel group range is changed with respect to the pixels, and the slower the shutter speed, the larger the pixel group range.

  The present invention takes into account that when the shutter speed is low, that is, when the exposure time is long, the magnitude of image blur during that period increases.

  According to a fourth aspect of the present invention, in the image blur correction device according to the first, second, or third aspect, the image blur detection unit detects an angular velocity in a vertical direction and a horizontal direction of the photographing optical system. It is characterized by comprising detection means and calculation means for calculating the direction and magnitude of image blur based on the angular velocity detected by the angular velocity detection means.

  The present invention shows an aspect in which image blur is detected by detecting an angular velocity of a photographing optical system. As another method, there is a method of comparing images of respective frames.

  The image blur correction method according to claim 5, a plurality of frames imaging step of continuously capturing images of a plurality of frames at a predetermined shutter speed by an imaging unit that captures an image formed by the photographing optical system; An image blur detection step for detecting an image blur of the photographing optical system when images of the plurality of frames are captured in the multi-frame imaging step, and an image of each frame is captured in the multi-frame imaging step A weighting coefficient determining step for determining a weighting coefficient used for weighting conversion based on image blur during the exposure period of the image, and for each frame image captured by the multiple-frame imaging step, each pixel is sequentially focused on As a weighted average of the pixel value of each pixel of a pixel group including a plurality of pixels in a predetermined range including the target pixel and the weighting coefficient determined by the weighting coefficient determination step By using the calculated value as a new pixel value of the pixel of interest, a weighting conversion step for weighting conversion of the image of each frame, and an image of each frame weight-converted by the weighting conversion step are converted into the image blur. And an image composition step of generating an image for one frame in which the image blur is corrected by combining the images by shifting by an amount corresponding to the image blur detected in the detection step.

  The image blur correction method according to a sixth aspect of the present invention is the image blur correction method according to the fifth aspect, wherein the weighting coefficient determination step includes the pixel of interest at the start of exposure when the image of each frame is captured by the imaging unit. It is characterized in that it is determined on the basis of the trajectory that the image point imaged in the period moves during the exposure period and the length of time existing in each pixel of the pixel group.

  The image blur correction method according to a seventh aspect of the present invention is the image blur correction method according to the fifth or sixth aspect, wherein the weighting conversion step is performed according to a shutter speed when an image of each frame is captured by the imaging unit. The size of the pixel group range is changed with respect to the pixels, and the slower the shutter speed, the larger the pixel group range.

  An image blur correction method according to an eighth aspect of the present invention is the invention according to the fifth, sixth, or seventh aspect, wherein the image blur detection step includes an angular velocity detection unit that detects angular velocities in the vertical and horizontal directions of the photographing optical system. And the direction and magnitude of image blur are calculated based on the detected angular velocity.

  The image blur correction methods according to claims 5 to 8 are method inventions corresponding to the image blur correction apparatuses according to claims 1 to 4, respectively.

  According to the present invention, an image with high sharpness in which image blur due to camera shake or the like is reduced can be taken with an inexpensive apparatus.

  DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of an image blur correction apparatus and a correction method thereof according to the present invention will be described below with reference to the accompanying drawings.

  FIG. 1 is a block diagram showing an embodiment of an internal configuration of an imaging apparatus (digital camera) 10 to which the present invention is applied.

  As shown in the figure, the digital camera 10 according to the present embodiment includes a photographing optical system 12, a image sensor 14 (hereinafter referred to as "CCD") composed of a solid-state imaging device such as a CCD, a timing generator (TG) 16, an analog signal processing. Unit 18, A / D converter 20, image input controller 22, digital signal processing unit 24, encoder 28, image display unit 30, compression / decompression processing unit 32, media controller 34, storage medium 36, auto exposure (AE) It comprises a detection unit 38, an autofocus (AF) detection unit 40, a central processing unit (CPU) 42, a ROM 44, a RAM 46, a flash ROM 48, an operation unit 50, a flash device 52, an angular velocity sensor 70, and the like.

  The operation unit 50 includes a shutter button, a power switch, a shooting / playback mode selection switch, a back switch, a menu / OK switch, a multi-function cross key, and the like. The switch S1 is turned on when the button is half-pressed to prepare for photographing such as AF and AE, and the switch S2 is turned on when the button is fully pressed to take in an image.

  The overall operation of the digital camera 10 is comprehensively controlled by the CPU 42, and the CPU 42 controls each unit of the digital camera 10 according to a predetermined program based on an input from the operation unit 50.

  In addition to the program executed by the CPU 42, the ROM 44 stores data necessary for various controls such as a program diagram. The CPU 42 develops the program stored in the ROM 44 in the RAM 46, and executes various processes while using the RAM 46 as a working memory. The flash ROM 48 stores various setting information relating to the operation of the digital camera 10 such as user setting information.

  The photographic optical system 12 includes a zoom lens 12z, a focus lens 12f, and a diaphragm (for example, iris diaphragm) 12i, which are driven and operated by a zoom motor 60z, a focus motor 60f, and an iris motor 60i, respectively. That is, the zoom lens 12z is driven by the zoom motor 60z to move back and forth on the photographing optical axis, thereby changing the focal length. The focus lens 12f is driven by the focus motor 60f to move back and forth on the photographic optical axis, thereby changing the imaging position. Further, the aperture 12i is driven by the iris motor 60i, and the aperture amount changes continuously or stepwise, thereby changing the aperture value. The CPU 42 controls the driving of the zoom motor 60z, the focus motor 60f, and the iris motor 60i via the zoom motor driver 62z, the focus motor driver 62f, and the iris motor driver 62i, and performs operations of the zoom lens 12z, the focus lens 12f, and the aperture 12i. Control.

  The CCD 14 is composed of a color CCD having a predetermined color filter array (for example, honeycomb array). Light incident on the light receiving surface of the CCD 14 via the photographing optical system 12 is converted into signal charges of an amount corresponding to the amount of incident light by each photodiode arranged on the light receiving surface. The signal charges accumulated in each photodiode are read according to the timing signal applied from the timing generator (TG) 16 and sequentially output from the CCD 14 as a voltage signal (image signal).

  The CCD 14 includes a shutter gate and a shutter drain, and the signal charges accumulated in each photodiode can be swept out to the shutter drain by applying a shutter gate pulse to the shutter gate. The CPU 42 controls application of a shutter gate pulse to the shutter gate via the TG 16 to control the charge accumulation time of signal charges accumulated in each photodiode (so-called shutter speed by an electronic shutter).

  The analog signal processing unit 18 includes a CDS circuit and an analog amplifier. The CDS circuit performs a correlated double sampling process on the CCD output signal based on the CDS pulse applied from the TG 16, and the analog amplifier has an imaging sensitivity applied from the CPU 42. The image signal output from the CDS circuit is amplified by the gain set accordingly. The A / D converter 20 converts the analog image signal output from the analog signal processing unit 18 into a digital image signal (image data).

  The image input controller 22 has a built-in buffer memory of a predetermined capacity, accumulates one frame of image data output from the A / D converter 20, and stores it in the RAM 46.

  The digital signal processing unit 24 includes a white balance correction circuit, a gamma correction circuit, a synchronization circuit, a contour correction circuit, a luminance / color difference signal generation circuit, etc., and processes the image data stored in the RAM 46 in accordance with a command from the CPU 42, A YUV signal composed of a luminance signal and a color difference signal is generated.

  When a through image is displayed on the image display unit 30, images are continuously captured by the CCD 14, and the obtained image data is continuously processed to generate a YUV signal. The generated YUV signal is applied to the encoder 28 via the RAM 46, converted into a signal format for display, and output to the image display unit 30. As a result, a through image is displayed on the image display unit 30.

  In the case of recording an image, the CCD 14 captures an image in response to a shooting command from the shutter button, and the obtained image data is processed to generate a YUV signal. The generated YUV signal is added to the compression / decompression processing unit 32, converted into predetermined compressed image data (for example, JPEG), and then stored in the storage medium 36 via the media controller 34.

  As will be described in detail later, when recording an image, if the image blur correction function is turned on, an image blur correction process is performed. Images for a plurality of frames are continuously picked up by the CCD 14 in accordance with a shooting command, and the image data for the plurality of frames is temporarily stored in the RAM 46. Then, the image data is read into the CPU 42 and an image blur correction process described later is performed. At this time, image data for one frame is generated by combining the acquired image data of a plurality of frames and stored in the RAM 46. Then, the image data stored in the RAM 46 is processed by the digital signal processing unit 24 as described above to generate a YUV signal, which is converted into predetermined compressed image data by the compression / decompression processing unit 32, and then the media controller. It is stored in the storage medium 36 via 34.

  The compressed image data stored in the storage medium 36 is read from the storage medium 36 in accordance with a reproduction command, converted into an uncompressed YUV signal by the compression / expansion processing unit 32, and then an image display unit via the encoder 28. 30 is output. As a result, the image recorded in the storage medium 36 is reproduced and displayed on the image display unit 30.

  The AE detection unit 38 calculates a physical quantity necessary for AE control from image data captured by the CCD 14 and taken into the RAM 46 at that time in accordance with a command given from the CPU 42 when the shutter button is half-pressed. For example, as a physical quantity necessary for AE control, one screen is divided into a plurality of areas (for example, 8 × 8), and an integrated value of R, G, and B image data is calculated for each divided area. The CPU 42 detects the brightness of the subject based on the integrated value obtained from the AE detection unit 38, the aperture value at the time of image data acquisition, and the shutter speed, and obtains the EV value for obtaining the optimum exposure. Based on the value and the program diagram, an appropriate aperture value and shutter speed (appropriate exposure time in the CCD 14) when the shutter button is fully pressed to record an image are determined.

  The AF detection unit 40 calculates a physical quantity necessary for AF control from the image data captured in the RAM 46 imaged by the CCD 14 at that time in accordance with a command given from the CPU 42 when the shutter button is half-pressed. In the digital camera 10 of the present embodiment, AF control is performed based on the contrast of the image, and the AF detection unit 40 calculates an AF evaluation value indicating the sharpness of the image from the image data. The CPU 42 controls the movement of the focus lens 12f by driving the focus motor 60f via the focus motor driver 62f so that the AF evaluation value calculated by the AF detection unit 40 is maximized.

  The flash unit 52 controls the flash light emission unit 54 including a xenon tube, the flash light emission amount (light emission time) emitted from the flash light emission unit 54 by the light emission command and the light emission stop command from the CPU 42, and charging of a main capacitor (not shown). And a flash control unit 56 that performs control.

  The flash device 52 operates in accordance with a flash mode selected from flash modes such as auto flash that automatically emits light at low luminance, forced flash, red-eye reduction flash, flash emission inhibition, and slow sync.

  Hereinafter, processing relating to image blur correction applied in the digital camera 10 will be described. FIG. 2 is a schematic configuration diagram showing functional blocks of the configuration of the image blur correction apparatus for performing the image blur correction. The following description will be given with reference to FIG. 1 as appropriate.

  In the present digital camera 10, the image blur correction function can be enabled (ON) / disabled (OFF) by a predetermined switch operation. When the image blur correction function is turned on, when the shutter button is fully pressed and a shooting command is received, the CPU 42 passes through the TG 16 so that the shutter speed (exposure time) determined by the AE detection unit 38 is obtained. Thus, the charge accumulation time (exposure time) of each photodiode in the CCD 14 is controlled to capture an image, and images for a plurality of frames are continuously captured with the same exposure time.

  Here, the number of frames of images to be continuously captured and the exposure time at the time of capturing each image are not limited to specific cases. For example, as described above, there is a mode in which an image corresponding to a predetermined number of frames is captured by continuously repeating the imaging process with an appropriate exposure time determined by the AE detection unit 38 for a predetermined number of times. Conceivable. In the present embodiment, the following description will be made on the assumption that an image for three frames is captured when an image for one frame is recorded according to this aspect. However, when recording an image for one frame, an image for two frames or four frames or more may be captured instead of three frames.

  As another embodiment, a value (T / T) obtained by dividing the exposure time T by the number of frames (N) of the image to be captured with respect to the optimal exposure time (T) obtained in the AE detection unit 38. N) may be an exposure time at the time of capturing an image of each frame. For the number of frames (N) of the image to be captured, for example, an exposure time value (T / n) when the optimal exposure time (T) is equally divided by a predetermined number of frames (natural number n) is determined in advance. It is also possible to determine the maximum value (N) among the number of frames (n) that satisfies the condition that the exposure time is longer than the minimum exposure time (Tmin).

  In the present embodiment, the image data for three frames sequentially captured by the CCD 14 is temporarily stored in the RAM 46 via the analog signal processing unit 18, the A / D converter 20, and the image input controller 22 as described above. Stored in This processing corresponds to storing image data of three frames of images captured by the photographing optical system 12 and the CCD 14 in the configuration diagram shown in FIG. 2 in the primary storage unit 100. Are means embodied by the RAM 46 of FIG.

  On the other hand, the CPU 42 detects the amount of image blur due to camera shake or the like during the imaging period in which the imaging process is performed by the CCD 14 as described above. An angular velocity sensor 70 is installed in the digital camera 10. The angular velocity signal output from the angular velocity sensor 70 is read by the CPU 42, and the CPU 42 performs processing such as integration on the angular velocity signal. As a result, a shake signal indicating the shake amount of the camera is generated. The angular velocity sensor 70 is composed of two angular velocity sensors that detect the angular velocity in the left-right direction and the angular velocity in the up-down direction, and a shake signal in the left-right direction and the up-down direction is generated by these angular velocity signals.

  In the digital camera 10, a potentiometer (not shown) for detecting the position is connected to the zoom lens 12z. The zoom position signal output from the potentiometer is read by the CPU 42, and the CPU 42 determines the amount of image shake (the magnitude of image shake in the vertical and horizontal directions) on the imaging surface of the CCD 14 based on the zoom position signal and the shake signal. Is calculated. That is, since the image shake amount on the imaging surface of the CCD 14 is determined by the camera shake amount and the focal length of the photographing optical system 12, a zoom position indicating the position of the zoom lens 12z as information on the focal length of the photographing optical system 12. A signal is acquired, and an image shake amount is calculated from the zoom position signal and the shake signal.

  In the configuration diagram of FIG. 2, the zoom position is detected by the lens information detection unit 102. The image shake amount detection unit 104 is supplied with the zoom position signal detected by the lens information detection unit 102 and the angular velocity signal from the angular velocity sensor 70, and the image shake amount detection unit 104 determines the image based on the information. The shake amount is detected (calculated). The image blur amount detecting means 104 is a means embodied by the arithmetic processing of the CPU 42 in FIG.

  The calculation of the image shake amount may be performed sequentially during the imaging period simultaneously with the reading of the angular velocity signal, but by storing the angular velocity signal and the zoom position signal obtained during the imaging period, the shake signal can be calculated from the angular velocity signal. The process of generating the image and the process of calculating the image shake amount from the zoom position signal and the shake signal may be performed after the imaging period. Further, only the process of generating the shake signal from the angular velocity signal may be performed during the imaging period simultaneously with the reading of the angular velocity signal.

  In addition, the detection of the image shake amount is not limited to the case of using the angular velocity sensor. For example, the image shake amount may be detected from an image captured using an imaging system of a different system from that for image capturing, or may be detected using other means. May be.

  When the imaging processing by the CCD 14 is completed and image data for three frames is accumulated in the RAM 46 in FIG. 1, the image data is read into the CPU 42. The CPU 42 performs the following image blur correction process on the read image data for three frames. In the configuration diagram of FIG. 2, the image blur correction calculating means 106 is a means embodied by the CPU 42 in FIG. 1, and the image blur correction calculating means 106 is for three frames stored in the primary storage means 100. The image blur correction process is performed by reading the image data and the image blur amount detected by the image blur detection unit 104.

  The image blur correction process includes a weighting conversion process for each image data of three frames and a pixel shift composition process. First, the weight conversion process is performed on the image data of each frame, and then the weight conversion process is performed. Pixel shifted composition processing is performed using the image data of each frame.

  First, the weighting conversion process will be described with reference to FIGS. In FIG. 3, an image of one frame (for example, the first frame) read from the RAM 46 is illustrated as the original image 150. In the original image 150, the pixel of interest (the pixel of interest) and its surroundings A 3 × 3 pixel group 152 composed of 8 pixels is extracted and shown. In the 3 × 3 pixel group 152, the center pixel is the target pixel 152A.

  In the weighting conversion process, a 3 × 3 filter (weighting table 154) is prepared for the 3 × 3 pixel group 152, and a weighted average using the weighting table 154 is obtained. For example, each pixel value (luminance value) of the pixel group 152 is a1, a2, a3, b1, b2, b3, c1, c2, c3 (the pixel value of the pixel of interest is b2) as shown in FIG. When the value of each element 154 is A1, A2, A3, B1, B2, B3, C1, C2, C3 (An, Bn, Cn: n = 1 to 3), the weighted average T is obtained by the following equation.

  Then, the weighted average T obtained in this way becomes the pixel value b2 ′ (= T) after weight conversion of the pixel of interest 152A.

  On the other hand, the value of each element of the weighting table 154 is set based on the image blur amount (change in image blur amount) during the imaging period (exposure period) when the original image 150 is captured by the CCD 14. That is, the locus of the image point that was imaged on the pixel of interest 152A at the start of imaging (at the start of exposure) of the original image 150 until the end of imaging (at the end of exposure) and the time that existed in each pixel Set based on length. For example, the value of each element of the weighting table 154 to be multiplied by each pixel is determined so that the image point becomes a large value in proportion to the proportion of time that the image point exists in each pixel. As described above, the image blur amount during the exposure period is calculated based on the angular velocity signal from the angular velocity sensor 70 and the zoom position signal (focal length of the photographing optical system 12).

  When the image point of the pixel of interest 152A moves, for example, as shown in the locus shown in FIG. 4A, based on the amount of image blur during the exposure period of the original image 150, the ratio of the time during which the image point exists in each pixel Accordingly, the value of each element of the weighting table 154 is set to a value as shown in FIG.

  In the CPU 42, the above weighting conversion processing is sequentially performed with all the pixels other than the one peripheral edge pixel of the original image 150 as the target pixel in order, and the image data of the original image 150 is converted as image data of the weighted conversion image. A similar weighting conversion process is performed on all the image data for three frames stored in the RAM 46. At this time, the value of each element of the weighting table is set for each image data of each frame based on the amount of image blur detected during the exposure period of each image data.

  Next, the pixel shift composition process will be described with reference to FIGS. When the weight conversion processing of the image data for three frames imaged by the CCD 14 is completed as described above, the CPU 42 synthesizes the pixel shift from the image data of the weight conversion images 150A to 150C for the three frames as shown in FIG. By processing, a composite image 160A for one frame for recording is generated.

  Therefore, the CPU 42 sets the origin for the weighted converted image 150A of the first frame as shown in FIG. Although the origin can be set at an arbitrary point on the image, the center position of the image is set as the origin in the present embodiment.

  Then, the horizontal shift amount and the vertical shift amount (displacement amount) due to the image blur of the second frame weight conversion image 150B and the third frame weight conversion image 150C are detected with respect to the first frame weight conversion image 150A. This deviation amount is obtained from the image blur amount detected by the angular velocity signal from the angular velocity sensor 70 (and the focal length of the photographing optical system 12) as described above. For example, when the amount of image blur detected at the start of image capture (exposure) for the first frame is set to 0, the image blur amount detected at the start of image capture for the second frame is 2. This is the shift amount of the weighted converted image 150B of the frame. The shift amount of the weighted converted image 150C in the third frame is the same. When the shift amounts of the weighted converted image 150B of the second frame and the weighted converted image 150C of the third frame with respect to the weighted converted image 150A of the first frame are obtained in this way, the image is shifted in the opposite direction by the shift amount. Thus, the weighted conversion images 150A to 150C from the first frame to the third frame are synthesized.

  Specifically, the following processing is performed. 6A, first, for the weighted converted image 150A of the first frame, assuming two-dimensional coordinates (X in the horizontal direction and Y in the vertical direction) with the center position of the image as the origin, Two-dimensional coordinates shall be assigned. On the other hand, with respect to the weighted converted image 150B of the second frame in FIG. 5B and the weighted converted image 150C of the third frame in FIG. 4C, the shift amount (image blur amount) obtained as described above is obtained. According to the above, the position of the origin is displaced from the center position of the image. For example, if the weighted converted image 150B of the second frame is shifted by ΔX in the X direction and ΔY in the Y direction with respect to the weighted converted image 150A of the first frame, the position of the origin is similarly shifted from the center position of the image. Displace by ΔX in the X direction and ΔY in the Y direction. Then, two-dimensional coordinates are assigned to each pixel by the coordinate system at the position of the new origin. In FIGS. 6B and 6C, a frame F indicated by a broken line with respect to the weighted converted image 150B of the second frame and the weighted converted image 150C of the third frame indicates a shift amount (image blur amount) of each image. In the two-dimensional coordinates at the new origin changed accordingly, the range of pixels to which the same coordinate value is assigned as each pixel of the weighted converted image 150A of the first frame is shown and an image having the same angle of view as the first frame image Indicates the range where.

  Thereby, in the weighted converted images 150A to 150C from the first frame to the third frame, the position (pixel) of the image point with respect to the same object point has the same coordinate value. Therefore, the pixel values of the pixels having the same coordinate value are added and averaged in the weighted converted images 150A to 150C from the first frame to the third frame. As a result, the weighted converted images of the respective frames are superimposed so that the image points of the same object point overlap as shown in FIG. 4D, and a composite image for one frame to be recorded on the storage medium 36 is generated. The

  Here, as shown by a thick frame (correctable range 160) in FIG. 4D, a pixel range in which the weighted converted images 150A to 150C for three frames overlap, and a pixel range that does not overlap the image at all for two frames or the other. Will occur. In the present embodiment, for example, only the image in the correctable range 160 in which the weighted converted images for three frames overlap is treated as the composite image 160A subjected to image blur correction, and the storage medium is actually stored in the correctable range 160. The image of the image size to be recorded in 36 is cut out. The details of the image cutting process will be described later.

  In the present embodiment, since each frame image is captured with an exposure time at which each has an appropriate brightness, the pixel values of the pixels of the weighted converted images 150A to 150C are averaged (added). Therefore, a composite image for one frame of proper exposure is obtained. However, what kind of composition processing is performed when a composite image for one frame is generated from a plurality of frames is not limited to this. For example, when images of a plurality of frames are captured by dividing the exposure time at which the brightness is appropriate, a composite image is generated by simply adding the pixel values of each pixel of the image of each frame. When processing such as signal amplification is performed on the image of each frame, luminance adjustment processing is performed on a composite image obtained by averaging the pixel values of each pixel or adding the pixel values. In other words, the image of each frame is synthesized by the process of adding the pixel values of each pixel and the process of making the brightness of the synthesized image obtained by this process appropriate. It depends on the conditions for acquiring the image and the processing content applied to the image of each frame.

  FIG. 7 is a flowchart showing a processing procedure of the above-described pixel shift composition processing. First, the CPU 42 sets the origin of a weighted converted image (hereinafter simply referred to as an image) for the first frame (step S10). For example, as described above, the center position of the first frame image is set as the origin. Next, the image shake amount at the start of exposure of the second and third frame images is calculated based on the angular velocity signal from the angular velocity sensor 70 (and the focal length of the photographing optical system 12) (step S12). Subsequently, the origins of the second and third frame images are determined as described above based on the image shake amount, and the origins of the first to third frame images are aligned (step S14). Then, a composite image for one frame is generated by calculating an average of the images from the first frame to the third frame (step S16).

  Image data of a composite image for one frame obtained by performing the image blur correction process as described above is output from the CPU 42 to the RAM 46. Then, as in the case where image blur correction is not performed, the image data is converted into a YUV signal by the digital signal processing unit 24, and subsequently converted into predetermined compressed image data by the compression / decompression processing unit 32. It is stored in the storage medium 36 via the controller 34. In FIG. 2, the image data of one frame of the composite image obtained by performing the image blur correction process by the image blur correction calculation unit 106 is recorded in the storage unit 108. The storage unit 108 Means embodied by the storage medium 36 are shown.

  Next, a more detailed embodiment and other embodiments of the image blur correction process will be described.

  First, an embodiment of weight conversion processing in image blur correction processing will be described. In the above embodiment, the weighting conversion process in the image blur correction process weights and converts the pixel value of the pixel of interest using the 3 × 3 weighting table 154 for the 3 × 3 pixel group 152 as shown in FIG. However, the weight conversion process can be performed with a pixel group having an arbitrary number of elements n × m (n and m are natural numbers) and a weighting table.

  Considering that the amount of image blur within the exposure period increases as the shutter speed at the time of capturing an image of each frame becomes slower, that is, as the exposure time becomes longer, the pixels used for weight conversion processing It is preferable to increase the group range and increase the number of elements in the pixel group and the weighting table. For example, when the exposure time is equal to or greater than a predetermined threshold value, weight conversion processing is performed using a 5 × 5 pixel group and a weighting table. When the exposure time is less than the threshold value, the above embodiment is performed. As described above, it is conceivable to perform weight conversion processing using a 3 × 3 pixel group and a weighting table. FIG. 8 shows an example of the weighting conversion process in this case. For example, when the exposure time is shorter than 1/60 seconds, the same as that shown in the above embodiment as shown in FIG. In addition, by calculating a weighted average using the 3 × 3 pixel group 152 and the weighting table 154, the pixel value b2 ′ after weighting conversion of the pixel of interest 152A is calculated.

  On the other hand, when the exposure time is 1/60 second or more, as shown in FIG. 5B, the weighted average is obtained from the 5 × 5 pixel group 152 and the weighting table 154, thereby performing the weighting conversion of the pixel of interest 152A. The pixel value b2 ′ is calculated. As shown in the figure, the pixel values of the 5 × 5 pixel group 152 are a1 to a5, b1 to b5, c1 to c5, d1 to d5, and e1 to e5 (the pixel value of the pixel of interest is c3), and the weighting table 154 When the values of each element of A1 are A1 to A5, B1 to B5, C1 to C5, D1 to D5, E1 to E5 (An, Bn, Cn, Dn, En: n = 1 to 5), T is required.

Then, the weighted average T thus obtained becomes the pixel value C3 ′ after weight conversion of the pixel of interest 152A.
Also in the 5 × 5 weighting table 154, the value of each element is set similarly to the value of each element of the 3 × 3 weighting table described in the above embodiment. That is, each pixel is multiplied so that the image point imaged on the pixel of interest 152A becomes a large value in proportion to the proportion of time existing in each pixel by the end of imaging (at the end of exposure). The value of each element of the weighting table 154 is determined.

  When the exposure time is extremely short, for example, when the exposure time is less than 1/250 seconds, the image shift amount during the exposure period of each frame image is extremely small, so that the pixel shift is not performed without performing the weighting conversion process. It is preferable to perform the synthesis.

  FIG. 9 is a flowchart showing a procedure for selecting a pixel group and the number of elements of the weighting table in the weighting conversion process for the exposure time of the image of each frame. When the CPU 42 sets an appropriate shutter speed (exposure time) for capturing an image of each frame based on information from the AE detection unit 38 (step S30), the CPU 42 determines whether the exposure time is 1/60 second or more. (Step S32). If it is determined YES, the pixel group and the number of elements in the weighting table in the weighting conversion process are set to 5 × 5 (step S34). On the other hand, if it is determined as NO in step S32, it is subsequently determined whether or not the exposure time is 1/250 second or longer (step S36). If it is determined YES, the pixel group and the number of elements in the weighting table in the weighting conversion process are set to 3 × 3 (step S38). If it is determined NO in step S36, the weight conversion process is not performed.

  Next, an embodiment related to pixel shift composition processing in image blur correction processing will be described.

  As shown in FIG. 6, when the weighted converted images 150A to 150C for three frames are shifted during the pixel shift composition, the pixel range where the weighted converted images 150A to 150C for the three frames overlap is image blur. As a result, the image is reduced to a part of the entire pixel range of the weighted converted image of each frame (correctable range 160 indicated by a thick line). On the other hand, the weighted converted images 150 </ b> A to 150 </ b> C of each frame are the same size as the image captured by the entire imaging surface of the CCD 14 and are larger in image size (pixels) than the image (recorded image) actually recorded on the storage medium 36. Number). As a result, the image size of the composite image 160A in the correctable range 160 is usually larger than the image size of the recorded image, and the recorded image is cut out from the composite image 160A in the correctable range 160.

  In the image blur correction process, the image of each frame continuously picked up by the CCD 14 is effective as image blur correction even when the pixels are combined by shifting without weighting conversion. Accordingly, the pixel shift composition processing in this specification can be applied to whether the image of each frame to be pixel shifted composite is a weighted conversion image obtained by weight conversion or an image that is not subjected to weight conversion. Even if the image of each frame to be shifted and combined is a weighted conversion image, it is simply referred to as an image.

  The cutout range of the recorded image is set to a rectangular range 170 with the center position of the first frame image 150A as the center, as shown in FIG. That is, the cutout range of the recorded image is set in the center with respect to the entire first frame image 150A obtained from the entire imaging surface (light receiving surface) of the CCD 14. The cutout area 170 is called a scheduled recording area 170. As shown in FIG. 6, when the images 150A to 150C for three frames are combined by shifting the pixels, the correctable range in which the images 150A to 150C for three frames overlap as shown in FIGS. 6D and 10B. 160 composite images 160A are generated. At this time, as shown in FIG. 10B, when the scheduled recording range 170 set in the first frame image 150A is included in the correctable range 160, the image in the scheduled recording range 170 is the recorded image. Cut out as.

  On the other hand, if the image capturing time (exposure time) of each frame image becomes long, the amount of image shake between the frames increases. Therefore, the correctable range 160 is largely displaced vertically and horizontally with respect to the center of the first image 150A. Alternatively, the size may be reduced, and the scheduled recording range 170 set at the center of the first frame image 150A may not be included in the correctable range 160.

  FIG. 11 shows the entire range of the imaging surface 180 of the CCD 14, and the pixel range of each frame corresponding to the scheduled recording range 170 set at the center position of the image 150A of the first frame at the time of pixel shift composition. This is shown on the imaging surface 180. A pixel range corresponding to the scheduled recording range 170 of the image 150A of the first frame is set to the pixel range 170A at the center position on the imaging surface 180 of the CCD 14. Note that, on the imaging surface 180 of the CCD 14, similarly to the first frame image, a two-dimensional coordinate is assumed in which the center position is the origin, the horizontal direction is X, and the vertical direction is Y.

  On the other hand, the pixel range corresponding to the scheduled recording area 170 of the images 150B and 150C in the second frame and the third frame is displaced vertically and horizontally corresponding to the shift amount (image blur amount) of these images, and the CCD 14 captures the image. The pixel ranges 170B and 170C are shifted from the center position on the surface 180, respectively. The pixel ranges 170B and 170C indicate ranges in which an image having an angle of view range that matches the pixel range 170A is captured. At this time, when the image blur amount increases as shown in the pixel range 170B of FIG. 9, the pixel range of the second frame or the third frame corresponding to the scheduled recording range 170 may protrude from the range of the imaging surface 180 of the CCD 14. . In this case, the scheduled recording range 170 is not included in the correctable range 160 as described above. In addition, when the pixel range of the second frame or the third frame corresponding to the scheduled recording range 170 is greatly deviated from the range of the imaging surface 180 of the CCD 14, the size of the correctable range 160 is smaller than the scheduled recording range 170. Occurs.

  In such a case, mainly the method of changing the position of the scheduled recording range 170 set at the center in the first image 150A, the method of changing the size of the scheduled recording range 170, and the scheduled recording range 170 of three frames. A method is conceivable in which the images 150A to 150C of minutes (all frames) are not limited to within the correctable range 160 where they are superimposed.

  First, a method for changing the position of the scheduled recording range 170 will be described. As shown in FIG. 12 (A), an initial scheduled recording area 170 set in the center of the first frame image 150A is included with respect to the correctable range 160 in which images of three frames are superimposed by pixel shift composition. If not, the scheduled recording range 170 is changed to a position included in the correctable range 160 as shown in FIG. At this time, for example, it is preferable to displace to a position where the minimum amount of change is in the X direction and the Y direction. In the case of FIG. Are aligned.

  FIG. 13 is a flowchart showing a procedure in the CPU 42 for changing the position of the scheduled recording range 170. As described in the flowchart of FIG. 7, the origin positions of the second and third frames are shifted from the center position based on the image blur amount (image shift amount), and the origin positions of the first to third frames are shifted. Then, the CPU 42 calculates the coordinate values of the four corners of the correctable range 160 in which the images for three frames are superimposed (step S20). As shown in FIG. 14A, when assuming an XY coordinate whose center is the origin for the first frame image 150A, the upper left corner, the upper right corner, the lower right corner, and the lower left corner of the correctable range 160 The coordinate values (XC0, YC0), (XC1, YC0), (XC1, YC1), and (XC0, YC1) are calculated.

  Subsequently, the CPU 42 determines whether or not the positions of the four corners of the standard scheduled recording range 170 set at the center of the first frame image 150A are within the correctable range 160 (step S22). That is, as shown in FIG. 14A, the coordinate values of the upper left corner, the upper right corner, the lower right corner, and the lower left corner of the initial scheduled recording area 170 are set to (X0, Y0), (X1, Y0), (X1, If Y1) and (X0, Y1), it is determined whether or not all the conditions of X0 ≧ XC0, Y0 ≧ YC0, X1 ≦ XC1, and Y1 ≦ YC1 are satisfied. If YES is determined, the scheduled recording range 170 is included in the correctable range 160. Therefore, the image of the scheduled recording range 170 is cut out as a recorded image without changing the position of the scheduled recording range 170. (Step S26).

  On the other hand, if NO is determined in step S22, the position of the scheduled recording area 170 is moved so that the positions of the four corners of the scheduled recording area 170 are within the correctable range 160 (stop S24). For example, if the scheduled recording area 170 is gradually moved in a direction that satisfies all the conditions of X0 ≧ XC0, Y0 ≧ YC0, X1 ≦ XC1, and Y1 ≦ YC1, all the conditions are satisfied. The scheduled recording area 170 is moved to the position where the As a result, the recordable range 170 is included in the correctable range 160 as shown in FIG. Then, the CPU 42 cuts out the image of the changed scheduled recording area 170 as a recorded image (step S26).

  Next, a method for changing the size of the scheduled recording range 170 will be described. As shown in FIG. 15A, an initial scheduled recording area 170 set at the center of the first frame image 150A is included with respect to the correctable range 160 in which images of three frames are superimposed by pixel shift composition. If not, the scheduled recording area 170 is reduced to a size included in the correctable area 160 as shown in FIG. For example, the image size of a recorded image cut out from the scheduled recording range 170 is defined in advance as a plurality of sizes such as 1600 × 1200 pixels and 1280 × 960 pixels, and the user selects a desired image size from among these image sizes. It can be selected. In the image size selected by the user, the size of the scheduled recording range 170 for cutting out a recorded image of that image size is also changed. Assuming that the scheduled recording range 170 having a size corresponding to the image size selected by the user is the initial scheduled recording range, if the initial scheduled recording range 170 is not included in the correctable range 160, the pixel size is set to that. The maximum scheduled recording range of the size that is included in the correctable range 160 without changing the center position of the first frame image 150A is the corresponding scheduled recording range. The new recording schedule range 170 is changed.

  FIG. 16 is a flowchart showing a procedure in the CPU 42 for changing the position of the scheduled recording range 170. As described in the flowchart of FIG. 7, the origin positions of the second and third frames are shifted from the center position based on the image blur amount (image shift amount), and the origin positions of the first to third frames are shifted. Then, the CPU 42 calculates the coordinate values of the four corners of the correctable range 160 in which the images for three frames are superimposed (step S30). As shown in FIG. 17A, when assuming an XY coordinate whose center is the origin for the image 150A of the first frame, the upper left corner, the upper right corner, the lower right corner, and the lower left corner of the correctable range 160 The coordinate values (XC0, YC0), (XC1, YC0), (XC1, YC1), and (XC0, YC1) are calculated.

  Subsequently, the CPU 42 determines whether or not the positions of the four corners of the standard scheduled recording range 170 set at the center of the first frame image 150A are within the correctable range 160 (step S32). That is, as shown in FIG. 17A, the coordinate values of the upper left corner, the upper right corner, the lower right corner, and the lower left corner of the initial scheduled recording area 170 are set to (X0, Y0), (X1, Y0), (X1, If Y1) and (X0, Y1), it is determined whether or not all the conditions of X0 ≧ XC0, Y0 ≧ YC0, X1 ≦ XC1, and Y1 ≦ YC1 are satisfied. If the determination is YES, the scheduled recording range 170 is included in the correctable range 160, and therefore the image of the scheduled recording range 170 is used as a recorded image without changing the size of the scheduled recording range 170. Cut out (step S36).

  On the other hand, if NO is determined in step S32, the scheduled recording area 170 is reduced to a prescribed size so that the positions of the four corners of the scheduled recording area 170 are within the correctable range 160 (stop S34). For example, prescribed image sizes 1600 × 1200 (pixels), 1280 × 960 (pixels),... That can be recorded as recorded images are determined in advance, and a table that records these image sizes is prepared. The CPU 42 refers to the data in the table and changes the scheduled recording range 170 to a size corresponding to the prescribed image size. Then, when the size of the scheduled recording range 170 is changed without changing the center position, the largest size among the sizes satisfying all the conditions of X0 ≧ XC0, Y0 ≧ YC0, X1 ≦ XC1, and Y1 ≦ YC1 The scheduled recording area 170 is changed (reduced). As a result, the scheduled recording range 170 is included in the correctable range 160 as shown in FIG. Then, the CPU 42 cuts out the image of the changed scheduled recording area 170 as a recorded image (step S36).

  When the scheduled recording range 170 is not included in the correctable range 160, both the position and the size of the scheduled recording range 170 are changed as described above, and the changed size of the scheduled recording range 170 is corrected. Both the position and size of the scheduled recording area 170 may be changed so as to be the largest of those included in the possible range 160.

  Further, when the size of the scheduled recording range 170 is changed (reduced) from the size corresponding to the initially planned image size as described above, an image cut out from the reduced scheduled recording range 170 is initially scheduled. The clipped image may be enlarged by pixel interpolation so as to match the image size.

  Next, a method will be described in which the scheduled recording range 170 is not limited to the correctable range 160 in which the images 150A to 150C for three frames are superimposed. As shown in FIG. 18, in the case where the recordable range 170 set in the center of the image 150A of the first frame is not included in the correctable range 160 in which the images of three frames are superimposed by pixel shift composition, Then, a combined image of the scheduled recording range 170 is generated without being limited within the correctable range 160.

  As shown in the figure, the scheduled recording range 170 is a range in which images 150A to 150C for three frames overlap (that is, a correctable range 160), and the first frame and the second frame or the first frame and the second frame. It is divided into ranges 162 and 164 where the images of the frames overlap, and a range 166 of only the image 150A of the first frame.

  When the images 150A to 150C of each frame are combined by shifting the pixels, as described in the above embodiment, a combined image is generated by calculating an average of pixel values of overlapping pixels (pixels having the same coordinate value) of each frame. . At this time, the pixel values are added to the overlapping pixels of each frame for all the pixels in the scheduled recording range 170, not limited to the correctable range 160 where the images of three frames overlap. Then, the added value is divided by 3 to obtain an average of pixel values of each pixel. If the scheduled recording range 170 is included in the correctable range 160, the obtained pixel value becomes the pixel value of each pixel of the composite image (recorded image) as it is. On the other hand, when the ranges 162 and 164 where the images of two frames overlap as shown in FIG. 18 and the range 166 of only the image of one frame occur, the pixel values of the pixels in the ranges 162 and 164 are as described above. The obtained pixel value is set to a value multiplied by 3/2, and the pixel value of the pixel in the range 166 is set to a value obtained by multiplying the obtained image value by 3 times as described above. As a result, the pixel value of the composite image in the scheduled recording range 170 becomes an average value of the pixel values of each frame corresponding to the number of overlapping frames. Then, the image in the scheduled recording range 170 is used as a recorded image.

  Next, image blur correction processing (pixel shift composition processing) when the flash mode is set to the slow sync mode will be described. In the slow sync mode, the flash unit 52 is turned on and shooting is performed with a low-speed shutter (long exposure). In this case, if an image for one frame is divided into a plurality of images, the image is divided into an image obtained when the flash is turned on and an image obtained when the flash is turned off.

  Further, in the image (subject) to be picked up, there are a part affected by the flash and a part not.

  For example, as shown in FIG. 19, when a person A in the foreground and a night scene B in the background are photographed in the slow sync mode, the person A in the foreground is captured as a dark image when the flash is turned off, whereas the person A in the foreground is captured when the flash is turned on. Since it is taken as a bright image and its luminance change is large, it becomes a part affected by the lighting of the flash. On the other hand, the night view B is a portion that is not affected by the lighting of the flash because the luminance change is small between when the flash is turned on and when the flash is turned off.

  When performing shooting in the slow sync mode (when recording one frame image), assuming that three frames are continuously captured as shown in FIG. 20, the flash device 52 is lit. The frame image captured when the flash is on (for example, the first frame image 150A) and the frame image captured when the flash device 52 is not lit (when the flash is off) (for example, two frames) Eyes and third frame images 150B and 150C). The part affected by the flash lighting (the part of the person A) is a bright area in the frame image 150A captured when the flash is on, and is a dark area in the frame images 150B and 150C captured when the flash is off.

  When pixel shift composition using all the images 150A to 150C in the area affected by the flash is performed, there is a problem that the area becomes unnecessarily dark due to the influence of the dark image when the flash is turned off.

  Therefore, by comparing the image of the frame captured when the flash is on and the image of the frame captured when the flash is off, the area affected by the flash is determined by detecting the area where the luminance change exceeds a predetermined value. To detect. For the area affected by the flash, only the frame image (first frame image) captured when the flash is turned on is used. For the area not affected by the flash emission, all frame images are used. It is assumed that the pixel shift composition is performed using.

  That is, in the image blur correction process, a weighting conversion process and a pixel shift composition process are performed on the area that is not affected by the flash as in the above embodiment. On the other hand, image blur correction processing is not performed on the area affected by the flash, and the image of the area is extracted from the frame image (first frame image) captured when the flash is turned on. Then, a combined image is generated by combining the images of the respective areas, and an image in the recording scheduled range is cut out from the combined image to be a recorded image.

  FIG. 21 is a flowchart showing a photographing process in the slow sync mode. When the slow sync mode is set and the shutter button is fully pressed and a shooting command is received, the number of shot frames is determined based on the required exposure amount (step S40). Then, the flash unit 52 emits light, and images for a plurality of frames are continuously captured (step S42). Subsequently, among the images of each frame, the brightness distribution of each of the images taken when the flash is turned on and the images taken when the flash is turned off is detected, and the brightness changes (brightness difference) greater than or equal to a predetermined value by comparing them. The pixel range in which the occurrence of the pixel is determined as a masking range (step S44). That is, an area where a luminance difference of a predetermined value or more is generated between when the flash is turned on and when the flash is turned off is an area affected by the flash as described above, and this area is set as a masking range.

  Subsequently, the CPU 42 performs an image blur correction process for an image range other than the masking range (step S46). As described above, the image blur correction process is performed by the weighting conversion process and the pixel shift composition process. In addition, in the pixel shift composition process, the value obtained by averaging the pixel values of the overlapping pixels of each frame image is not used as the pixel value of the composite image, but simply the added value (total value) is used as the pixel value of the composite image. It is desirable.

  Then, the CPU 42 extracts the image of the masking range from the frame image captured when the flash is turned on, and combines the extracted image of the masking range and the image other than the masking range subjected to the image blur correction processing in step S46. (Addition) is performed to generate a composite image for one frame, and brightness adjustment is performed to obtain optimum brightness (step S48). An image in the scheduled recording area is cut out as a recorded image from the composite image generated in this way, and recorded in the storage medium 36 (step S50).

  It should be noted that the image blur correction process of the present embodiment at the time of shooting in the slow sync mode is effective even when only the pixel shift composition process is performed without performing the weighting conversion.

1 is a block diagram showing an embodiment of an internal configuration of an imaging apparatus (digital camera) according to the present invention. FIG. 2 is a schematic configuration diagram showing functional blocks of a configuration as an image shake correction apparatus for performing image shake correction processing. Explanatory drawing used for description of the weighting conversion process in image blur correction. The figure which illustrated the weighting table used by weighting conversion processing. Explanatory drawing used for description of pixel shift composition processing in image blur correction. Explanatory drawing used for description of pixel shift composition processing in image blur correction. The flowchart which showed the process sequence of the pixel shift composition process. Explanatory drawing used for description of the other form of weight conversion processing. The flowchart which showed the selection procedure of the pixel group in the weighting conversion process, and the number of elements of a weighting table. Explanatory drawing used for description of pixel shift composition processing. Explanatory drawing used for description of the relationship between a recording scheduled range and the imaging surface of CCD. Explanatory drawing used for description when changing the position of the scheduled recording range when the scheduled recording range is not included in the correctable range. The flowchart which showed the process sequence in the case of changing the position of a recording scheduled range, when the recording scheduled range is not included in the correction | amendable range. Explanatory drawing used for description when changing the position of the scheduled recording range when the scheduled recording range is not included in the correctable range. Explanatory drawing used for description when changing the magnitude | size of a recording scheduled range when the recording scheduled range is not included in the correction | amendable range. The flowchart which showed the process sequence in the case of changing the magnitude | size of a recording scheduled range, when the recording scheduled range is not included in the correction | amendable range. Explanatory drawing used for description when changing the magnitude | size of a recording scheduled range when the recording scheduled range is not included in the correction | amendable range. Explanatory drawing used for description when a recording plan range is not restrict | limited to the correction | amendable range. Explanatory drawing used for description of image blur correction processing when the flash mode is set to the slow sync mode. Explanatory drawing used for description of image blur correction processing when the flash mode is set to the slow sync mode. 6 is a flowchart showing a processing procedure for image blur correction processing when the flash mode is set to a slow sync mode.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Imaging device (digital camera), 12 ... Imaging optical system, 12f ... Focus lens, 12z ... Zoom lens, 14 ... CCD, 16 ... Timing generator (TG), 18 ... Analog signal processing part, 20 ... A / D conversion 22 ... Image input controller, 24 ... Digital signal processing unit, 28 ... Encoder, 30 ... Image display unit, 32 ... Compression / decompression processing unit, 34 ... Media controller, 36 ... Storage medium, 38 ... AE detection unit, 40 ... AF detection unit, 42 ... Central processing unit (CPU), 44 ... ROM, 46 ... RAM, 48 ... Flash ROM, 50 ... Operation unit, 52 ... Flash device, 70 ... Angular velocity sensor, 100 ... Primary storage means, 102 ... lens information detection means, 104 ... image shake amount detection means, 106 ... image shake correction calculation means, 108 ... storage means, 150 ... Image, 152 ... pixel group 160 ... correctable range, 170 ... recording expected range

Claims (8)

An image pickup means for picking up an image formed by the photographing optical system; and an image pickup control means for acquiring an image for a plurality of frames by causing the image pickup means to continuously pick up an image for a plurality of frames at a predetermined shutter speed. An image blur detecting unit for detecting an image blur of an image formed by the photographing optical system, and an image of each frame acquired by the imaging control unit according to the image blur detected by the image blur detecting unit. In an image blur correction apparatus comprising: an image synthesizing unit that generates an image for one frame in which image blur is corrected by synthesizing by shifting by an amount;
For each frame image acquired by the imaging control unit and used to synthesize one frame image generated by the combining unit, each pixel is set as a target pixel in order, and a predetermined range including the target pixel The image of each frame is weighted by using a value obtained by weighted averaging of the pixel value of each pixel group of the plurality of pixels and a predetermined weighting coefficient as a new pixel value of the target pixel. Weighting conversion means for converting;
Based on the image blur detected by the image blur detection unit during the exposure period when the image of each frame is captured by the imaging unit, the weighting coefficient used for the weight conversion of the image of each frame in the weight conversion unit. Weighting coefficient determination means determined by
An image blur correction apparatus comprising:   The weighting coefficient determining unit is configured to detect a trajectory in which an image point formed on the pixel of interest moves during an exposure period at the start of exposure when an image of each frame is captured by the imaging unit, and each of the pixel groups. 2. The image blur correction device according to claim 1, wherein the image blur correction device is determined based on a length of time existing in the pixel.   The weight conversion unit changes the size of the range of the pixel group with respect to the pixel of interest according to the shutter speed when the image capturing unit captures an image of each frame, and the slower the shutter speed, 3. The image blur correction device according to claim 1, wherein a range of the pixel group is increased.   The image blur detection unit detects an angular velocity detection unit that detects vertical and horizontal angular velocities of the imaging optical system, and calculates the direction and magnitude of the image blur based on the angular velocity detected by the angular velocity detection unit. The image blur correction apparatus according to claim 1, 2 or 3, further comprising: A multi-frame imaging step of continuously capturing images of a plurality of frames at a predetermined shutter speed by an imaging unit that captures an image formed by the imaging optical system;
An image blur detection step of detecting an image blur of the photographing optical system when images of the plurality of frames are being imaged in the multiple frame imaging step;
A weighting factor determination step for determining a weighting factor used for weighting conversion based on an image blur during an exposure period when an image of each frame is captured in the multiple frame imaging step;
With respect to each frame image captured by the multiple frame imaging step, each pixel is set as a target pixel in order, and the pixel value of each pixel of a pixel group including a plurality of pixels in a predetermined range including the target pixel, and the weighting A weight conversion step of performing weight conversion of the image of each frame by setting a value obtained by weighted averaging with the weighting coefficient determined by the coefficient determination step as a new pixel value of the pixel of interest;
The image of each frame subjected to the weight conversion by the weight conversion step is synthesized by shifting by an amount corresponding to the image blur detected by the image blur detection step, thereby generating an image for one frame in which the image blur is corrected. Image compositing process;
An image blur correction method comprising:   The weighting coefficient determining step includes: a step in which an image point formed on the pixel of interest moves during an exposure period at the start of exposure when the image of each frame is imaged by the imaging unit; and each of the pixel groups 6. The image blur correction method according to claim 5, wherein the determination is based on a length of time existing in the pixel.   In the weighting conversion step, the size of the pixel group range is changed with respect to the pixel of interest in accordance with the shutter speed when the imager captures an image of each frame. 7. The image blur correction method according to claim 5, wherein a range of the pixel group is increased.   The image blur detection step is characterized in that an angular velocity detection unit detects an angular velocity in a vertical direction and a horizontal direction of the photographing optical system, and calculates a direction and a magnitude of the image blur based on the detected angular velocity. Item 5. The image blur correction method according to 5, 6, or 7.

Images