JP2014090227A - Imaging device and control method for the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress exposure unevenness in a shutter direction by an image correction processing in an imaging device having a shutter mechanism using both of a mechanical shutter and an electronic shutter function.SOLUTION: An imaging device has a shutter mechanism using both of a mechanical shutter 105 and an electronic shutter 107. An image processor 109 determines occurrence of exposure unevenness on the basis of information of an eye relief, a diaphragm value, an image height, a focal state detection result, the distance from an imaging face to the mechanical shutter and the slit width of the shutter. When a lack determination unit 122 determines occurrence of lack in a blurred image, correction processing for symmetrizing the brightness distribution of an image in an occurrence direction of the lack is executed. A correction filter applying unit 129 applies a correction filter with a filter coefficient having a characteristic whose shape corresponds to a shape obtained by reversing the brightness distribution of an achieved image 120 in the shutter running direction, and outputs data of a corrected image 130.

Description

  The present invention relates to an image processing technique when imaging is performed using both a mechanical shutter and an electronic shutter function.

A single-lens reflex digital camera includes a device that performs an imaging operation using a mechanism that uses a mechanical shutter and an electronic shutter function together (hereinafter referred to as a combined mechanism). In this case, the mechanical shutter is a focal plane shutter and constitutes the rear curtain. Prior to the running of the trailing curtain, the electronic shutter of the image sensor is driven. Driving an electronic shutter that performs charge accumulation start scanning of the pixels of the image sensor instead of the mechanical shutter front curtain is hereinafter referred to as an electronic front curtain. In the case of an imaging operation using an electronic shutter, for example, in an imaging device using a CMOS (complementary metal oxide semiconductor) sensor, reset scanning is performed to reduce the accumulated charge amount of a pixel to zero. This is performed for each pixel or for each region or line (hereinafter referred to as a pixel region) composed of a plurality of pixels. After a lapse of a predetermined time from the reset scanning, scanning for reading out signals for each pixel or each pixel region is performed, and an imaging operation using an electronic shutter is performed.
When the exposure of the image sensor is controlled by the combined mechanism, first, as the charge accumulation start scan of the image sensor, reset scanning is sequentially performed for each pixel area of the image sensor in the running direction of the rear curtain. After a predetermined time has elapsed, the image sensor is shielded from light by the rear curtain, and then scanning is performed to sequentially read out the charges accumulated in each pixel.

By the way, in an interchangeable lens type single-lens reflex digital camera, the focal length and the exit pupil distance (the distance from the image plane to the exit pupil position of the lens) differ depending on the photographing lens mounted on the camera body. For this reason, the combined mechanism has the following problems.
The electronic shutter functions on the image sensor surface, and the mechanical shutter is disposed at a position slightly away from the image sensor surface in the optical axis direction due to its configuration. Therefore, the light-shielding position on the imaging surface by the mechanical shutter changes due to differences in focal length, exit pupil position, and the like of the photographing lens. In particular, when the exposure time from the reset scanning time to the light shielding time by the mechanical shutter is short, uneven exposure may occur in the shutter travel direction. Although this phenomenon will be described in detail later, it changes according to the focal length, exit pupil position, and the like of the photographic lens mounted on the camera body. For example, a blurred image that is out of focus is not circular but partly missing. When shooting various scenes and checking them, especially when the light and darkness is clear, such as the sunbeams, the image lacking a part of the circle is clearly shown, resulting in an image unintended by the photographer. there is a possibility.

  Also, the exit pupil distance may change depending on the aperture value and focus position of the photographic lens, the focus of the macro lens, etc., and the light shielding position on the imaging surface by the mechanical shutter may change. Due to these causes, a phenomenon in which a part of the image in the shutter scanning direction is lost may occur. With respect to this problem, Patent Document 1 discloses a technique for avoiding an imaging condition in which image loss occurs by regulating the ISO sensitivity and exposure time of an image sensor.

JP 2010-41510 A

However, limiting the shooting conditions as in Patent Document 1 is disadvantageous in that the photographer cannot freely select the shooting settings.
An object of the present invention is to suppress exposure unevenness in the shutter traveling direction by image correction processing in an imaging apparatus having a shutter mechanism using both a mechanical shutter and an electronic shutter function.

  In order to solve the above-described problems, an apparatus according to the present invention includes an imaging unit that converts light incident through an imaging optical system into an electrical signal, an electronic shutter by the imaging unit is used as a front curtain, and a mechanical shutter is provided. An imaging apparatus that performs exposure control using a shutter mechanism as a rear curtain, and receives focus light that is incident through the imaging optical system to detect a focus state, and detection information of the focus state detection unit , Calculating a defocus amount, and performing focus adjustment control of the imaging optical system, image data by the imaging unit, detection information of the focus state detection unit, and imaging conditions from the control unit Is acquired, and when it is determined that uneven exposure occurs in the image related to the image data, image correction is performed to symmetrize the luminance distribution of the image in the traveling direction of the shutter mechanism. An image processing means for outputting the corrected image data.

  According to the present invention, exposure unevenness in the shutter travel direction can be suppressed by the image correction process.

It is a figure explaining the profit image in a combined use type mechanism. FIG. 8 is a block diagram showing a system configuration example (A) and a configuration example (B) of an image processing unit for explaining the first embodiment of the present invention in conjunction with FIG. 3 to FIG. 7. It is a figure explaining the focus state detection part which an imaging part has. It is a figure explaining the example of a calculation process about an image defect. It is a figure explaining the filter coefficient for image correction. It is a figure explaining correction of the filter coefficient at the time of distance competition. It is a figure explaining the calculation process of the image at the time of tilt determination. It is a block diagram which shows the system configuration example which concerns on 2nd Embodiment of this invention. It is a figure which shows the example of a division | segmentation of an image.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
Before describing the configuration of the imaging apparatus, the operation of the electronic front curtain and mechanical shutter in the combined mechanism and the phenomenon in which uneven exposure occurs in the shutter traveling direction will be described with reference to FIG. Here, in particular, in the case where the time from the reset scanning execution time to the light shielding time by the mechanical shutter is short (for example, a short exposure time such as 1/4000 seconds), refer to FIGS. To explain. Note that the subject side is defined as the front side, and the front-rear relationship of each unit will be described.

FIG. 1A illustrates the state of the imaging light flux. (A) In the figure, a mechanical shutter surface 2 is located slightly in front of the image pickup surface 1, and an exit pupil surface 3 is shown further forward, and an optical axis 4 of the image pickup optical system is indicated by a two-dot chain line. . FIG. 5A shows a so-called “rear pin” state in which the focal point 6 of the lens is located behind the imaging surface 1 at an image height 5 shifted from the optical axis 4 to the exposure end side (upward). The reset line 7 corresponds to the front end (lower end) of the electronic front curtain, and is shown as a member for convenience in FIG. The leading end (upper end) of the rear curtain 8 indicates a slit forming portion of the rear curtain of the mechanical shutter. Rays 9 and 10 are an upper ray 9 and a lower ray 10 of the light beam passing through the exit pupil.
The state shown in FIGS. (B-1) to (b-3) shows a state in which the spread of the light flux passing through the exit pupil is projected onto the imaging surface when viewed from the direction along the optical axis 4. A circle 11 represents the outer shape of the image obtained by projecting the spread of the light beam on the imaging surface 1, and a circle 12 represents the outer shape of the image obtained by projecting the spread of the light beam on the mechanical shutter surface 2. (B-1) In the figure, after the reset line 7 travels from the bottom to the top and exposure starts at the timing of crossing the circle 11, the rear curtain 8 travels from the bottom to the top and crosses the circle 12. The exposure ends at the timing.

A case where the time from the execution of reset scanning to the light shielding by the mechanical shutter is extremely short will be described with reference to FIG. In this case, the upper end of the trailing curtain 8 may pass through the lower end 12a of the circle 12 before the reset line 7 reaches the lower end 11a of the circle 11. A light ray passing through the upper ends of the circles 11 and 12 corresponds to the upper light ray 9, and a light ray passing through the lower ends of the circles 11 and 12 corresponds to the lower light ray 10. In the lower light ray 10, since the exposure end timing comes before the exposure start time, the image on the imaging surface 1 has a shape in which a part on the lower side is missing.
(B-3) The figure shows the state in which a little time has passed compared to (b-2). The reset line 7 reaches almost the center of the circle 11, and the upper end of the trailing curtain 8 does not reach the center of the circle 12. That is, a circular blurred image is exposed after image loss occurs. As described above, if the timings at which the electronic front curtain and the rear curtain pass through the lower end, the center, and the upper end of the light beams on the imaging surface 1 and the mechanical shutter surface 2 vary, exposure unevenness occurs. . As a result, as shown in FIG. 3C, the image 13 exposed on the imaging surface 1 has a missing portion on the lower side, and uneven exposure occurs in the vertical direction (shutter travel direction).

  FIG. 1B shows a so-called “front pin” state in which the focal point 6 is positioned forward of the imaging surface 1 at the same image height 5 as in FIG. (B) The circle 11 shown in the figure represents the spread of the light flux on the imaging surface 1, the light ray passing through the upper end thereof corresponds to the lower light ray 10, and the light ray passing through the lower end of the circle 11 corresponds to the upper light ray 9. On the other hand, the circle 12 represents the spread of the light flux on the mechanical shutter surface 2, the light ray passing through the upper end corresponds to the upper light ray 9, and the light ray passing through the lower end corresponds to the lower light ray 10. That is, in the “front pin” state, the relationship between the upper and lower rays corresponding to the upper and lower ends of the circles 11 and 12 is reversed. As a result, as shown in FIG. 7C, the position where the image 13 exposed on the imaging surface 1 is missing is opposite to the “rear pin” state, and the exposure unevenness is also opposite.

  FIG. 1C shows a “rear pin” state as in FIG. 1A, but the image height 5 is below the optical axis 4. (B) As shown in the figure, the relationship between the centers of the circle 11 and the circle 12 is different from the state shown in the (b-1) diagram in FIG. Specifically, in FIG. 1C (b), the circle 11 and the circle 12 are not concentric, and the center of the circle 11 is shifted downward from the center of the circle 12. On the other hand, when the image height 5 is above the optical axis 4 as shown in FIG. 1A, the center of the circle 11 is shifted above the center of the circle 12. (B) In the state shown in the figure, due to the difference in the deviation, the reset line 7 can pass through the lower end of the circle 11 before the upper end of the trailing curtain 8 takes the lower end of the circle 12 too much. As a result, the image 13 exposed on the imaging surface 1 becomes a circle without a chip, but exposure unevenness occurs.

From the above, it can be seen that the following phenomenon is caused.
・ The out-of-focus image is not circular and part of it is missing.
-The direction in which the blurred image is missing differs between the distant view and the close view, with the subject in focus as the reference.
-The missing state of the image (including the presence or absence of a missing part) differs depending on the screen position (image height).
For example, when the camera posture in the shooting state is the normal position and the shutter travel direction is from the bottom to the top in FIG. 1, the blurred image on the far side is missing on the bottom (see FIG. 1 (C)). In the near-field side, the upper side of the blurred image is missing (see FIG. 1B (c)). In particular, a phenomenon in which an image is missing at the lower side of the screen (a missing blurred image in the foreground) is remarkable.

[First Embodiment]
Hereinafter, a first embodiment of the present invention will be described. The imaging apparatus according to the present embodiment includes means for suppressing exposure unevenness due to the above-described phenomenon that occurs when exposure control is performed using a combined mechanism by image correction.

FIG. 2A is a block diagram illustrating the system and image processing of this embodiment. An example of an interchangeable lens type imaging device that includes a camera body 101 and a photographing lens device 102 is illustrated.
The imaging optical system 103 includes a plurality of lens groups and optical members. In FIG. 2A, a diaphragm 103a and a focus lens 103b for focus adjustment included in the imaging optical system are shown on the optical axis 104, but a zoom lens or the like is provided. The lens driving unit 110 includes an actuator that drives the movable optical member, and performs aperture control of the diaphragm 103a and position control of the movable lens in accordance with a control signal from the lens control unit 111. The electrical contact 112 is a communication and power supply connection unit that connects the camera body 101 and the photographing lens device 102 attached thereto, and mediates between the lens control unit 111 and the camera control unit 114. These control units include, for example, a CPU (Central Processing Unit), and control each unit by executing a predetermined program read from the memory.

The mechanical shutter 105 is a focal plane shutter, and drive control is performed by the camera control unit 114. The rear curtain of the mechanical shutter 105 constitutes a shutter mechanism together with the electronic front curtain. The imaging unit 106 includes an imaging element such as a CMOS sensor. The imaging element converts light incident through the imaging optical system 103 into an electrical signal, and outputs the captured image data to the image processing unit 109. The imaging unit 106 has a function of an electronic shutter 107 and includes a focus state detection unit 108. The focus state detection information related to the imaging optical system 103 is sent from the focus state detection unit 108 to the camera control unit 114 to perform focus adjustment control. The detection information of the focus state detection unit 108 includes information on whether or not the subject is in focus, and information on the amount of shift when the subject is not in focus. The camera control unit 114 calculates the defocus amount from the detection information. To do.
The image data processed by the image processing unit 109 is stored in the memory unit 116. The display unit 115 includes a display device such as a liquid crystal panel, and displays captured images and various types of information according to control commands from the camera control unit 114.

Next, an outline of the operation of the imaging apparatus will be described.
In the photographic lens device 102, the aperture 103a and the focus lens 103b constituting the imaging optical system 103 are controlled by the lens driving unit 110, and light from the subject is guided to the camera body 101 in this state. In the camera body 101, the focus state detection unit 108 performs focus state detection. That is, whether or not the subject is in focus and the amount of shift are detected, and the detection signal is output to the camera control unit 114. The subject light guided to the camera body 101 by the photographic lens device 102 is exposure controlled by the mechanical shutter 105 and the electronic shutter 107, and the imaging unit 106 outputs an imaging signal to the image processing unit 109. The image processing unit 109 performs image processing based on image data of an imaging signal acquired from the imaging unit 106 and imaging information including focus state detection information sent via the camera control unit 114. Details of the image correction processing performed by the image processing unit 109 will be described later. The image data after image processing is stored in the memory unit 116, and the camera control unit 114 receives an operation instruction from an operation unit (not shown), reads out data such as a photographed image from the memory unit 116, and displays the data on the display unit 115. Control the display on the screen.

  Next, the configuration of the focus state detection unit 108 in the imaging unit 106 will be described with reference to FIG. In the present embodiment, each pixel 131 of the image sensor has a plurality of focus state detection elements. These detection elements respectively detect light that has passed through different areas of the exit pupil. In the example of FIG. 3, one pixel 131 has a pair of sub-pixels 108a and 108a. The dividing lines of the sub-pixels 108a and 108a are different from each other with an angle difference of 90 ° between adjacent pixels, and a photodetector (photodiode) and a microlens (not shown) constituting each sub-pixel 108a are provided. Yes. The plurality of pixels constitute the focus state detection unit 108 as a whole. In the example of FIG. 3, the image shift amount is calculated by comparing the signals of the respective images (A image and B image) detected by the pair of sub-pixels 108a and 108a, and the defocus amount (focus shift amount). ) Is converted. The camera control unit 114 calculates a lens driving amount from the defocus amount, and transmits a control signal to the lens control unit 111 via the electrical contact 112. Receiving this control signal, the lens control unit 111 controls the lens driving unit 110 and moves the focus lens 103b along the optical axis to perform a focus adjustment operation. Note that the number of sub-pixel divisions in one pixel portion may be three or more.

  FIG. 2B shows the flow of image correction processing performed by the image processing unit 109. In order to make the flow of data easy to understand, the processing means is represented by a rectangular frame, and the data is distinguished by a chamfered frame with rounded corners. The acquired image 120 indicates image data acquired by the image processing unit 109 from the imaging unit 106, and the shooting information 121 is reference information acquired by the image processing unit 109 from the camera control unit 114, and includes focus state detection information (defocus amount). including. The correction filter table 123 includes correction filter coefficient data.

  The lack determination unit 122 determines whether or not the blurred image is missing based on the acquired image 120 and the shooting information 121. When the blur image is missing (see FIGS. 1A and 1B), the correction filter coefficient 124 is acquired from the correction filter table 123. The correction filter correction unit 125 corrects the correction filter coefficient 124 based on the acquired image 120 and the shooting information 121. Although details of the correction process will be described later, the correction process is executed according to the determination results by the perspective conflict determination unit 126 and the tilt determination unit 127. The correction filter correction unit 125 outputs the corrected correction filter coefficient 128 to the correction filter application unit 129. The correction filter application unit 129 is an image correction unit that actually corrects the acquired image 120. With this correction, a corrected image 130 is obtained from the acquired image 120. The data of the corrected image 130 is image data in which the occurrence of uneven exposure due to the above phenomenon is suppressed.

  The image correction process is not always executed with reference to the acquired image 120 and the photographing information 121, and correction is not necessary when a missing image is not lost. The lack determination unit 122 does not select the correction filter coefficient 124 from the correction filter table 123, and the correction filter application unit 129 does not apply the corrected correction filter coefficient 128. That is, the image data is output with the acquired image 120 without any correction. In the present embodiment, the configuration example in which the image processing unit 109 of the camera body 101 performs image correction processing has been described. However, the present invention is not limited to this, and the camera main body 101 and an information processing apparatus provided separately may be connected, and the acquired image 120 and shooting information 121 may be transmitted to the information processing apparatus to perform the same image correction processing as described above. In this case, the camera body 101 and the information processing apparatus constitute an imaging apparatus (system).

Next, with reference to FIG. 4, calculation processing related to missing blur images will be described. In the calculation, various parameters used are as follows.
-Exit pupil distance.
Aperture value (so-called F value).
-Image height and defocus amount at image height.
-Distance from the imaging surface to the mechanical shutter surface.
-The slit width, which is the distance when the electronic front curtain and rear curtain (mechanical shutter) travel.

In FIG. 4A, FIG. 4A corresponds to FIG. 1A and shows the “rear pin” state. FIG. 4B is a graph showing time on the horizontal axis and image height on the vertical axis.
(A) The angle 14 shown in the figure represents an angle formed by the upper ray and the lower ray determined from the F value with reference to the position of the imaging surface 1 on the optical axis. The exit pupil diameter 16 is obtained from this angle 14 and the exit pupil distance 15 that is the distance between the exit pupil plane 3 and the imaging plane 1. Next, at a certain image height 5, the position of the focus 6 is determined from the defocus amount 17 based on the detection result of the focus state detection unit 108. Thereby, the circle 11 representing the spread of the light beam on the imaging surface 1 can be determined. In the figure, the length (diameter) 11c in the vertical direction with respect to the circle 11 is shown. Further, from the distance 18 between the imaging surface 1 and the mechanical shutter surface 2 which is a design amount, a circle 12 representing the spread of the light beam on the mechanical shutter surface 2 can be similarly determined. In the figure, the length (diameter) 12c in the vertical direction with respect to the circle 12 is shown.
In the following, only the image height position in the vertical direction of the imaging surface 1 will be discussed. Depending on the image height position in the left-right direction of the image pickup surface, the relationship between the spread of light fluxes on the image pickup surface 1 and the mechanical shutter surface 2 changes, but it does not affect the chipping. The reason is that the shift direction in the positional relationship between the spread of the two light beams is perpendicular to the shutter scanning direction.

  In FIG. 5B, the lengths 11c and 12c of the circles representing the spread of the light beam in the vertical direction on the imaging surface 1 and the mechanical shutter surface 2 are projected onto the image height axis which is the vertical axis. The image height 11a is set at the lower end position of the circle 11, and the image height 12a is set at the lower end position of the circle 12. The image height 11b is set at the upper end position of the circle 11, and the image height 12b is set at the upper end position of the circle 12. The graph line 7a that rises to the right exemplifies the travel timing of the tip of the electronic front curtain, and the time at the foot of the perpendicular line drawn from the point on the graph line 7a to the time axis corresponds to the exposure start timing. Times 19a and 19b correspond to the image heights 11a and 11b, respectively. Further, the graph line 8a that rises to the right exemplifies the travel timing of the leading edge of the rear curtain, and the time at the foot of the perpendicular line drawn down from the point on the graph line 8a to the time axis corresponds to the exposure end timing. Times 20a and 20b correspond to the image heights 12a and 12b, respectively.

First, the exposure start timing will be described with reference to the graph line 7a. The time 19a when the electronic front curtain crosses the image height 11a is the exposure start timing at the lower end of the image. The time 19b when the electronic front curtain crosses the image height 11b is the exposure start timing at the upper end of the image. Assuming that the curtain speed of the electronic front curtain is quasi-constant, the exposure starts sequentially from time 19a to 19b in accordance with the graph line 7a indicating linear characteristics from the lower end to the upper end of the image.
Next, the exposure end timing will be described with reference to the graph line 8a. The time 20a when the trailing curtain crosses the image height 12a is the exposure end timing at the lower end of the image. The time 20b when the trailing curtain crosses the image height 12b is the exposure end timing at the upper end of the image. Assuming that the curtain speed of the trailing curtain is quasi-constant, the exposure is sequentially terminated from time 20a to 20b according to the graph line 8a indicating the linear characteristic from the lower end to the upper end of the image.

  From the above times 19a, 19b, 20a, and 20b (20a <19a <19b <20b), the cause of image loss and exposure unevenness can be explained. For example, when the exposure start time 19a at the lower end of the image is compared with the exposure end time 20a in the optical state shown in FIG. 9A, the time 20a is earlier in time. In this case, since the end time has passed before the exposure starts at the lower end of the image, the image lack occurs as a result of the exposure time becoming zero. On the other hand, at the upper end of the image, the exposure start time 19b is first and the exposure end time 20b is later. The temporal relationship between the exposure start timing and the exposure end timing is normal. Exposure is performed at the upper end of the image, and no image loss occurs.

  As described above, the chipped state, that is, the presence / absence of the chipping is different at the lower end and the upper end of the image, and the exposure time is naturally different. With regard to the exposure start timing and the exposure end timing, the exposure time in the vertical direction should have a uniform or linear distribution in consideration of linearity from the lower end to the upper end of the image. In the case of FIG. 4A, the exposure time at the lower end of the image is zero, and the exposure time is at the upper end of the blurred image. Accordingly, the light quantity distribution is such that a chipped state with zero exposure time continues from the lower end of the image, the exposure time becomes zero (no chipping) from a certain position (image height), and the exposure time gradually increases linearly toward the upper end. Is obtained.

Here, in the case where the exposure end time 20a is earlier than the exposure start time 19a, the concept of “minus exposure time” is introduced. Normal exposure time is zero or positive. When the exposure time is further extended to include a negative value, the exposure time determined from “exposure end time−exposure start time” shows a negative value when exposure end time <exposure start time. As a result, the exposure time can be linearly interpolated even in a region other than the upper end to the lower end of the image. This concept is shown in FIG.
The image 13 shown in the left figure is missing, and the hatched portion 13e indicates the missing portion. The intersections with a straight line extending in the vertical direction through the circle center of the image 13 correspond to the image height 13a at the lower end and the image height 13b at the upper end, respectively. The position of the intersection of the straight line extending in the vertical direction through the circle center of the image 13 and the boundary line of the hatched portion 13e corresponds to the image height 13d. The right figure is a graph showing the relationship between each position of the image 13 projected onto the image height axis which is the vertical axis and the exposure time on the horizontal axis. The exposure time axis is the positive axis on the right side of the origin O.

  The exposure time at the image height 13a is a negative value and indicates the time corresponding to the point 21a. The exposure time at the image height 13b is a positive value and indicates the time corresponding to the point 21b. The exposure time distribution from the image height 13a to the image height 13b is linear as shown by the graph line 21, and exhibits a distribution obtained by linear interpolation. Looking at this distribution, the exposure time is zero at the image height 13d corresponding to the intercept of the vertical axis, and the image is missing in the range of the image height 13a to the image height 13d located below (see the hatched portion 13e). I understand that.

Through the above procedure, the chipping determination unit 122 (see FIG. 2B) calculates the chipped state of the image and the exposure time distribution. Note that image defect and exposure unevenness are uniquely determined by the parameters. Therefore, it is not necessary to perform calculation every time. In the present embodiment, a calculation result is obtained in advance, and this is dealt with by performing image correction using the correction filter table 123. Details of the correction filter coefficient 124 will be described below with reference to FIG.
FIG. 5A illustrates the image 13, where a part of the lower side is missing and an axis extending in the vertical direction through the center of the circle is _denoted by yn. This axis is an axis orthogonal to the x axis set in the horizontal direction of the imaging surface. 5B illustrates the luminance distribution 22 of the image 13 with y _n on the horizontal axis and luminance on the vertical axis. As shown in the figure, the luminance distribution 22 is asymmetric in the vertical direction (y _n direction), and the range 23 is missing.

The brightness distribution function f (x, y) of the blurred image can be defined where x is the horizontal coordinate of the imaging surface and y is the vertical coordinate. FIG. 4B illustrates filter coefficients for correcting chipping and vertical asymmetry with respect to a blurred image, with the coordinate y _n on the horizontal axis and the filter coefficient on the vertical axis. The graph line 24 represents the change of the correction filter coefficient, and the coefficient value distribution has a characteristic of a shape obtained by inverting the luminance distribution 22 shown in (a) in the vertical direction (y _n direction). The symmetry axis is an axis that passes through the origin of y _n and extends in a direction parallel to the vertical axis of FIG.
In order not to change the sum of the luminances of the entire image 13, the sum of the filter coefficients is set to 1. The correction filter can be defined as a function g (x, y) using the x coordinate and the y coordinate. When the brightness distribution function of the blurred image after correction is defined as F (x, y), it is expressed by the following equation.

[Formula 1]
F (x, y) = g (x, y) * f (x, y)
“*” Represents convolution (convolution operation). The corrected luminance distribution 25 corresponding to F (x, y) is shown in FIG. The horizontal axis is the coordinate y _n, the vertical axis represents luminance. The shape of the luminance distribution 25 is not chipped and is symmetric in the vertical direction (y _n direction).

  The missing portion of the image 13 is compensated by the symmetrized image correction process, and the luminance distribution is corrected to a vertically symmetric shape. Such correction may be applied under specific conditions in consideration of detection errors in the focus state detection unit 108 and the load of correction processing. The specific condition is, for example, a shooting scene in a back-pin state such as a sun leaking day in which an image defect is clearly seen.

Next, a special case of correction filter selection will be described with reference to FIG.
FIG. 6A illustrates a so-called perspective conflict state in which there are regions that have different focus state detection results within a close range. It is assumed that the blurred images 13K and 13L are close to each other, and the blurred image 13L has a larger defocus amount than 13K. The blurred images 13K and 13L are both missing, and a point P25 is shown in the overlapping region.
When correcting a blurred image centered on the point P25 in a region where both images overlap, it may be difficult to accurately detect the focus state in a region where two images having different defocus amounts are close to each other. That is, as a result of the difference between the blurred images 13K and 13L and the blurred image 13x (see the two-dot chain line) calculated based on the focus state detection result obtained at the point P25, the correction filter coefficient corresponding to the blurred image 13x. Then, an appropriate correction cannot be performed for an actual blurred image. A solution to this problem is described below. The perspective conflict determination unit 126 (see FIG. 2B) determines that the blurred images 13K and 13L calculated from different defocus amounts overlap as perspective conflict, and performs an avoidance process described later.

  While it is difficult to detect the focus state at the point P25 shown in FIG. 6A, the blurred images 13K and 13L are close to each other across the point P25. Accordingly, when the defocus amounts at the point P25 and the blurred images 13K and 13L are denoted as DEF (P25), DEF (13K), and DEF (13L), respectively, “DEF (13K) <DEF (P25) <DEF (13L) ) ”Is established. Since the defocus value is larger than the blurred image 13K and smaller than the blurred image 13L at the point P25, a selection criterion is required as to which of the correction filter coefficients corresponding to the blurred images 13K and 13L is to be applied. . In the following, it will be described which one can be used for correction without any trouble and the influence can be avoided.

  As in the present embodiment, when image data is subjected to filter processing composed of only positive filter coefficients, the sharpness of the image is reduced, resulting in a blurred image. The degree of blurring is greater for filters with longer tap lengths. When a filter having a blurring effect is applied to a blurred image whose diameter and chipping cannot be predicted, the deterioration is smaller when the degree of blurring is kept low. Therefore, it is only necessary to pay attention to the tap length of the correction filter corresponding to the blurred images 13K and 13L and to adopt a criterion of selecting a shorter tap length.

First, the perspective conflict determination unit 126 compares the blurred images 13K and 13L. The left images in FIG. 6B show the blurred images 13K and 13L in a concentric state. The right figure is a graph with y _n on the horizontal axis and luminance on the vertical axis. The graph line 22K indicates the luminance distribution of the blurred image 13K with a relatively small defocus amount, and the graph line 22L indicates the luminance distribution of the blurred image 13L with a relatively large defocus amount. Although the graph lines 22K and 22L overlap, there is a slight deviation. Since the blurred images 13K and 13L are blurred images in one captured image, the captured information other than the image height and the focus state detection result is the same. However, since both images are close to each other, the image heights are considered to be substantially equal, and therefore only the defocus amount as a focus state detection result is different. The diameters 26K and 26L indicate the diameters of the blurred images 13K and 13L, respectively. These diameters change in proportion to the defocus amount, but the ratios of the respective missing areas 23K and 23L in the blurred image are not so different.

  Next, correction filter coefficients for the blurred images 13K and 13L will be described with reference to FIG. A graph line 24K indicates a filter coefficient for the blurred image 13K having a relatively small defocus amount, and a graph line 24L indicates a filter coefficient for the blurred image 13L having a relatively large defocus amount. These have inverted shapes in which the graph lines 22K and 22L shown in FIG. Considering that the difference between the diameters of the images and the ratio of missing parts are almost the same, it can be seen that the filter tap length is longer for a filter with a large defocus amount and shorter for a filter with a small defocus amount. . Based on this and the above-described argument, the filter that is preferable for the point P25, that is, the selection result of the perspective conflict determination unit 126 is a filter corresponding to the blurred image 13K with a small defocus amount. The use of the filter can solve the above problem at the time of perspective conflict.

Next, another example of a special case relating to correction filter selection will be described with reference to FIG.
FIG. 7A shows an optical relationship between the imaging surface 1, the mechanical shutter surface 2, the exit pupil surface 3, and the like, and is a case where the photographing lens device 102 mounted on the camera body 101 is a tilt lens. That is, the exit pupil plane 3 and the central axis 28 thereof are not perpendicular to the imaging plane 1 and the mechanical shutter plane 2 because they are inclined with respect to the optical axis 4. In this case, the blurred image to be corrected is elliptical due to the tilt of the lens. Hereinafter, a preferable correction filter in the tilt state of the imaging optical system will be described. Note that the tilt determination unit 127 (see FIG. 2B) calculates a tilt amount (a lens tilt angle with respect to the imaging surface 1), and reflects the tilt determination result in the correction filter correction result.

  When the photographic lens device 102 shown in FIG. 2A is attached to the camera body 101, the lens control unit 111 sends lens information to the camera control unit 114 via the electrical contact 112. Thereby, when the tilt lens is attached, the camera control unit 114 can recognize that the photographing lens is the tilt lens. On the other hand, many tilt lenses are not configured to detect the amount of tilt and notify the camera body. Therefore, in the correction procedure described below, the information that the tilt lens is attached to the camera body can be used, but the description will be made on the assumption that the information on the tilt amount cannot be acquired. If the camera control unit 114 can acquire information on the tilt amount, the tilt amount may be reflected in the calculation process of the blurred image.

FIG. 7B is an enlarged view showing the vicinity of the imaging surface in FIG. In a state where the central axis 28 of the exit pupil plane 3 is inclined with respect to the optical axis 4, the focal plane 6 f where the focal point 6 exists is perpendicular to the central axis 28. The defocus amount is defined as a distance with reference to the focal plane 6f. It should be noted that the spread of the light flux on the imaging surface 1 is not taken into consideration, and a blurred image is simply discussed from the spread of the light flux on the imaging surface 1.
The blurred image 30 represents the spread of the light flux on the imaging surface 1 and is a blurred image when the chipping is ignored. This is an ellipse that is longer in the vertical direction (major axis direction) than in the left-right direction (minor axis direction). The defocus amount 32a at the lower end 30a of the blurred image 30 is equal to the distance from the focal plane 6f perpendicular to the central axis 28 of the exit pupil plane 3 to the plane 31a. The plane 31 a is a plane that includes the lower end 30 a of the blurred image 30 on the imaging surface 1 and is perpendicular to the central axis 28 of the exit pupil plane 3. On the other hand, the defocus amount 32b at the upper end 30b of the blurred image 30 is equal to the distance from the focal plane 6f to the plane 31b. The plane 31 b is a plane that includes the upper end 30 b of the blurred image 30 on the imaging plane 1 and is perpendicular to the central axis 28 of the exit pupil plane 3.

When the defocus amounts 32a and 32b are compared, “32a> 32b” is obtained. This is because the focal plane 6f is inclined with respect to the imaging surface 1, and the defocus amount is continuously shortened from the lower end 30a to the upper end 30b of the blurred image 30.
Here, when three focus state detection points (see 33a, 33b, and 33c) along the vertical direction are set in the blurred image 30, it is assumed that an accurate focus state detection result is obtained at each detection point. . According to the blur image calculation procedure described above, blur images at the detection points 33a, 33b, and 33c become circles 34a, 34b, and 34c having different diameters, respectively. Since the blurred image 30 when the tilt lens is mounted is elliptical, the calculated shape of the blurred image is different from the actual shape of the blurred image. A solution to this problem is described below.

  As described above, the defocus amount at each detection point in the vertical direction changes continuously because the focal plane 6f is inclined with respect to the imaging surface 1 when the tilt lens is mounted. Moreover, the region where the defocus amount changes is a narrower region as the blurred images overlap. Since the camera control unit 114 recognizes that the tilt lens is attached to the camera body 101, the tilt determination unit 127 originally creates a blurred image at the detection points 33a, 33b, and 33c based on this information. Specifically, it is determined to be an elliptical blur image.

  In the elliptical blur image calculation process, since the change in the defocus amount in the vertical direction on the imaging surface 1 is known, the inclination (see angle 35) between the imaging surface 1 and the exit pupil surface 3 is calculated from this information. Is done. When it is assumed that the detection points 33a, 33b, and 33c from which different defocus amounts are calculated are inclined in the depth direction with respect to the paper surface of FIG. 7, based on the inclination angle and the geometric shape of the light beam described above, The length 36 of the major axis of the ellipse 38 can be calculated. The short axis length 37 of the ellipse 38 is equal to the diameter of the blurred image (see the circle 34b) calculated at the detection point 33b. From the above, it is possible to correct the blurred image (see the circle 34b) calculated at the detection point 33b into a blurred image (see the ellipse 38).

  The correction filter correction unit 125 considers the influence of deformation from a circular image generated when the tilt lens is mounted to an elliptical image, and corrects the filter coefficient based on the elliptical blur image calculated by the tilt determination unit 127. I do. 4 and 5, the circular blur image 13 changes to become an ellipse, but the basic processing is the same as described above. Note that the tilt of the exit pupil in the pan direction orthogonal to the tilt direction is also different only in the tilt direction, and thus correction processing may be performed by the same tilt determination as described above. Further, the present embodiment has been described on the assumption that the camera control unit 114 cannot directly acquire information on the tilt amount of the photographing lens device. When the information can be acquired, the elliptical shape can be directly calculated from the tilt amount, and this may be reflected in the filter coefficient correction process.

  According to the first embodiment, the combined mechanism performs image correction based on the exit pupil distance, aperture value, image height, focus state detection result, distance from the imaging surface to the mechanical shutter, and shutter slit width. Is called. Without limiting the shooting conditions, image defects and the like can be corrected and uneven exposure can be suppressed, so that the photographer can shoot with the intended settings without being restricted.

[Second Embodiment]
Next, a second embodiment of the present invention will be described. The present embodiment is different from the first embodiment in that the focus state detection unit 108 is provided separately from the imaging unit 106. Therefore, by using the same reference numerals already used for the same configuration and processing as in the case of the first embodiment, description thereof is omitted.

  FIG. 8 is a block diagram illustrating a system configuration example according to the present embodiment. The imaging unit 106 does not have a focus state detection unit, and the focus state detection unit 108 is provided separately. The optical member 113 is a so-called quick return mirror, and switches to a state in which light guided from the photographing lens device 102 to the camera body 101 is guided to the focus state detection unit 108 at a predetermined timing. The optical member 113 is driven by a drive mechanism (not shown) controlled by the camera control unit 114. The focus state detection unit 108 receives the reflected light of the quick return mirror, performs focus state detection, and outputs a detection signal to the camera control unit 114. The camera control unit 114 calculates the defocus amount and transmits a control signal to the lens control unit 111 via the electrical contact 112. The lens control unit 111 performs drive control of the focus lens 103 b via the lens driving unit 110.

  As in the present embodiment, the focus state detection unit 108 provided individually often cannot detect the focus state over the entire imaging surface, and the defocus amount at consecutive detection positions cannot be obtained. That is, the defocus amount can be acquired only at discrete positions. As described in the first embodiment, the blur image calculation process requires a defocus amount at the position of the blur image. In the present embodiment, a blur image calculation process is performed using defocus amounts at discrete positions.

  FIG. 9 is a conceptual diagram illustrating defocus amounts at discrete positions. A plurality of areas 202 indicate divided areas with respect to the image area 201. In other words, the image is divided into N × M regions 202 by dividing the entire image into N in the horizontal direction and M in the vertical direction. The image processing unit 109 calculates a blurred image using one representative value (defocus amount) for the divided area 202, and applies a correction filter having a filter coefficient calculated from the representative value to the entire divided area. It applies to.

  In this method, the number of divisions is sufficient, and it is desirable to assign one division region to one discrete detection position (the position of the focus state detection point). However, if the divided areas cannot be associated with the discrete positions of the detection points on a one-to-one basis, there may be a plurality of detection points in one divided area. In such a case, for the same reason as the avoidance of perspective conflict described in the first embodiment, processing for setting a representative value at a detection point with a relatively small defocus amount is performed. As a result, the degree of blur can be kept small, and adverse effects caused by calculating the filter coefficient using the representative value instead of the strict defocus amount can be avoided.

  In the second embodiment, the same effect as in the first embodiment can be obtained even when only the defocus amount at discrete detection positions can be obtained by the combined mechanism. In other words, the user can shoot without being restricted by the shooting conditions by suppressing image loss and exposure unevenness in the shutter scanning direction due to the positional relationship between the shooting lens, the imaging device and the mechanical shutter, and the situation of the subject.

DESCRIPTION OF SYMBOLS 103 Imaging optical system 106 Imaging part 105 Mechanical shutter 107 Electronic shutter 122 Defect determination part 126 Perspective competition determination part 127 Tilt determination part 129 Correction filter application part

Claims (10)

An imaging apparatus having imaging means for converting light incident through an imaging optical system into an electrical signal, and performing exposure control by a shutter mechanism having an electronic shutter by the imaging means as a front curtain and a mechanical shutter as a rear curtain. And
A focus state detecting means for detecting the focus state by receiving the light incident through the imaging optical system;
Control means for obtaining detection information of the focus state detection means, calculating a defocus amount, and performing focus adjustment control of the imaging optical system;
When the image data obtained by the imaging unit, the detection information of the focus state detection unit, and the information indicating the photographing condition from the control unit are acquired, and it is determined that exposure unevenness occurs in the image related to the image data, the shutter An image pickup apparatus comprising image processing means for performing image correction for symmetrizing the luminance distribution of the image in the traveling direction of the mechanism and outputting the corrected image data. The image processing means includes a chipping determination means for determining whether or not a chip occurs in the image related to the image data.
When it is determined by the missing determination means that the image is missing, a process of symmetrizing the luminance distribution of the image in the traveling direction of the shutter mechanism by correcting a filter coefficient applied to the image data. The imaging device according to claim 1, wherein the imaging device is performed.   The missing determination means acquires an exit pupil distance, an aperture value, an image height, a distance from the imaging surface to the mechanical shutter, and a slit width of the shutter mechanism as information indicating imaging conditions from the control means, and the image The imaging apparatus according to claim 2, wherein occurrence of chipping is determined.   The image processing means uses the filter coefficient having a characteristic of a shape obtained by inverting the luminance distribution of the image in the traveling direction of the shutter mechanism when the lack determination means determines that the image is defective. The imaging apparatus according to claim 2, wherein the image correction is performed. The image processing unit includes a perspective conflict determination unit that determines whether different perspective conflict states occur within a range in which detection information of the plurality of focus state detection units is close to each other,
When it is determined by the perspective conflict determination means that perspective conflict occurs, the image correction is performed according to detection information with a relatively small defocus amount among detection information of the plurality of focus state detection means. The imaging apparatus according to claim 1, wherein the imaging apparatus is characterized. The image processing unit includes an inclination determination unit that determines whether or not the imaging optical system is inclined with respect to the imaging surface of the imaging unit,
The image correction is performed by calculating a deformation of an image that changes according to an inclination when the inclination determination unit determines that the imaging optical system is inclined with respect to an imaging surface of the imaging unit. The imaging device according to any one of 1 to 5.   7. The imaging apparatus according to claim 1, wherein the imaging unit includes a plurality of detection elements that respectively detect light that has passed through different areas of an exit pupil of the imaging optical system as the focus state detection unit. The imaging device according to item.   The image processing means divides an image area related to the image data into a plurality of areas, and for each divided area, a defocus amount calculated from detection information of the focus state detection means is used as a representative value. The imaging apparatus according to claim 1, wherein the image correction is performed after making the correspondence.   In the image processing unit, each of the divided areas and the positions of discrete detection points related to focus state detection do not correspond one-to-one, and a plurality of detection points exist in one divided area. In this case, the representative value is set at a detection point with a relatively small defocus amount. An imaging apparatus having imaging means for converting light incident through an imaging optical system into an electrical signal, and performing exposure control by a shutter mechanism having an electronic shutter by the imaging means as a front curtain and a mechanical shutter as a rear curtain A control method to be executed,
Obtaining image data by the imaging means, detection information of a focus state related to the imaging optical system, and information indicating imaging conditions;
Determining whether exposure unevenness occurs in the image according to the image data using the acquired information; and
When it is determined that exposure unevenness occurs in the image related to the image data, the image correction is performed to symmetrize the luminance distribution of the image in the traveling direction of the shutter mechanism, and the corrected image data is output. A method for controlling an image pickup apparatus.

Images