JPH10177135A - Focus detector - Google Patents

Abstract

PROBLEM TO BE SOLVED: To minimize arithmetic calculation amount and arithmetic calculation time for calculating relative positional deviation between a subject and a camera, and to perform optimum focus detection considering the calculated relative positional deviation between the subject and the camera, by setting the range to detect the moving amount of an image and changing the position of an image signal used to detect a phase difference. SOLUTION: A diaphragm for focus detection 65 is moved and held in the optical path of a photographing lens by a motor 66, and a light shielding plate for focus detection 67 is moved by a motor 68 and covers over either the aperture part 65a or 65b of the diaphragm 65 so as to shield light. Then, the maximum moving amount of the image signal is predicted from the focal distance of an optical system, the interval of an image signal fetching time and further a driving time. A correlative value showing the coincident degree of two images obtained time sequentially in this range is arithmetically calculated by a correlative arithmetic calculation, and the position of the image signal having the high coincident degree of two images is searched from the correlative value, then the phase difference is arithmetically calculated by using the image signal at this position.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital camera or the like, which is capable of transmitting at least two luminous fluxes passing through an optical system.
By dividing the optical image formed in time series on the imaging means through each of the regions into an image signal, and detecting a phase difference between the image signals. The present invention relates to an improvement of a focus detection device that calculates a focus state of the optical system.

[0002]

2. Description of the Related Art Hitherto, a large number of phase difference detection type focus detection devices have been used as automatic focus adjustment devices used in single-lens reflex type silver halide cameras.

FIG. 48 is an arrangement diagram of an optical system of a single-lens reflex camera having a conventional phase difference detection type focus detection device. A part of a light beam 9a emitted from the taking lens 1 is a main mirror which is a half mirror. The light 2 is reflected as a light beam 9b in the direction of the focus plate 3, and is formed as an object image on the mat surface. The photographer is configured to observe the subject image on the focus plate via the eyepieces 5a and 5b and the pentaprism 4.

On the other hand, the light beam 9e of the light beam 9a transmitted through the main mirror 2 is reflected by the sub-mirror 6 and guided to the focus detecting device 7 as a light beam 9f. The focus state (defocus amount) with respect to the photographing silver halide film 8 is detected.

When the detected defocus amount is larger than a predetermined focusing width and it is determined that the image is out of focus, a control circuit (not shown) focuses the photographing lens 1 so as to cancel the detected defocus amount. The adjustment lens is driven to adjust the focus.

Next, the principle of focus detection of a conventional focus detection device will be described with reference to FIGS.

FIG. 49A shows a focused state, that is, an in-focus state. Light beams 16a and 16b transmitted through two different pupils of the photographing lens 10 form an image on the primary image forming surface 14, and the primary image is formed. The subject image on the imaging plane is a secondary imaging lens 12
Due to a and 12b, re-imaging is performed on the sensor surface 13 on which two sets of line sensors are arranged. Here, the field lens 11 is disposed near the primary image forming plane of the photographing lens 10 and efficiently guides a light beam having a predetermined image height to the sensor surface 13 to prevent a decrease in the amount of light generated as the image height increases.
In general, 2 that passes through different pupils of the taking lens 10
The two light beams 16a and 16b are defined by apertures (not shown) disposed immediately before or immediately after the secondary imaging lenses 12a and 12b, and the photographing lens 10 does not have a member for performing pupil division. Assuming that the distance (phase difference) between the relative positions of the two images in the focused state is δ ₀ , the defocus amount and direction at that time can be determined from the difference between δ ₀ and the actually given phase difference. .

FIG. 49 (b) shows a state in which the front focus is set by the defocus amount d1, and the phase difference δ ₁ between the two images.
Becomes smaller than δ ₀ , and when d ₁ becomes larger, “δ ₀ −δ
₁ "also increases.

FIG. 49C shows a state in which the image is focused on by the defocus amount d ₂ , and the phase difference δ ₂ between the two images is obtained.
Is greater than δ ₀ . Then, as d ₂ increases, the value of “δ ₂ −δ ₀ ” also increases. In this way, by detecting the phase difference between the two images formed on the sensor surface 13, the focus state of the photographing lens, that is, the magnitude and direction of the defocus amount can be detected.

[0010]

In a conventional single-lens reflex camera, the size of the conventional single-lens reflex camera is increased to secure a space for disposing the focus detection device, and the cost required for the focus detection device is increased. In addition, since the relative positional relationship between the photographing film and the focus detection device changes due to a temperature change or a durability change of the quick return mirror (main mirror), the focusing accuracy is reduced accordingly. There are drawbacks.

In an image pickup apparatus such as a digital camera, a solid-state image pickup device is used as an image pickup medium. By using this solid-state image pickup device as a focus detection sensor, the above-mentioned disadvantage can be solved. One method of using a solid-state imaging device as a focus detection device is to perform focus adjustment by detecting the contrast of a subject image on a solid-state imaging device used in video cameras and the like. However, since accurate defocus amount cannot be detected, high-speed and high-precision focus adjustment cannot be performed.

Therefore, a pupil dividing means for selectively transmitting one of two different pupil regions in the photographing lens is provided, and a subject image obtained by a light beam transmitted through one pupil region and the other are provided. Of the relative position of the subject image obtained by the luminous flux transmitted through the pupil region of
By detecting the phase difference and detecting the defocus amount, the above problem can be solved, and high-speed, high-precision focus adjustment can be performed.

However, in the above conventional example, after the image signal of the first subject image obtained by the light beam passing through one pupil region is fetched, the second object obtained by the light beam passing through the other pupil region is acquired. This means that the image signals of the images are captured, and the operations of capturing the two image signals are temporally different. Here, if the subject image on the image sensor moves between the time when the first image signal is captured and the time when the second image signal is captured, there is a problem that a detection error occurs due to the movement.

The above problem will be described with reference to FIG. This example is an example in which the subject image on the image sensor moves in the vertical direction due to camera shake or movement of the subject.

FIG. 13A shows a first subject image L.
When the camera is tilted downward due to camera shake after capturing the subject image L and before capturing the subject image R shown in FIG. 13B, the subject image R moves above L. Such image signals R _c of the position on the same image pickup element and the image signals L _c case will be looking at the underside of the object from the L _c,
An image signal having a different shape results. When a correlation operation is performed using the image signals having different shapes, the obtained phase difference also has an erroneous value.

As described above, when the subject image on the image sensor moves in the vertical direction due to camera shake or movement of the subject, there is a problem that an error occurs in the detected phase difference.

(Object of the Invention) It is an object of the present invention to minimize the amount of calculation and the calculation time for calculating the relative displacement between the subject and the camera, and to reduce the calculated relative position between the subject and the camera. An object of the present invention is to provide a focus detection device capable of performing optimum focus detection in consideration of a position shift.

[0018]

In order to achieve the above object, the present invention according to claims 1 to 4, according to the present invention, comprises a photographing means for photographing, a photographing optical system, and a photographing light beam passing through the optical system. Is chronologically divided into at least two different regions, and a pupil shape changing unit for forming an image on the imaging unit, and an optical element which is formed on the imaging unit in a time series manner through each of the regions. A focus detecting device comprising: an image converting unit that converts an image into an image signal, and calculates a focus state of the optical system by detecting a phase difference of the image signal. Movement amount detection means for detecting a movement amount in a direction orthogonal to the direction in which the regions are arranged; a range setting means for setting a range in which the movement amount is detected by the movement amount detection means; For the phase difference detection performed by the arithmetic means. That is to focus detection apparatus provided with an image signal position changing means for changing the position of the image signal.

More specifically, under general photographing conditions, the movement of the subject image in the vertical direction of the camera, for example, is caused by camera shake (or fluctuation of the subject itself). The movement of the subject image caused by camera shake becomes larger as the focal length of the optical system becomes longer, for example.
There is such a characteristic that it becomes larger as the image signal capturing time interval is larger. Therefore, the maximum movement amount of the image signal is predicted from the focal length of the optical system and the time interval for capturing the image signal, the accumulation time of the optical image and the drive time of the pupil shape changing means. The required range is limited except for the required range, and a correlation value indicating the degree of coincidence of the two images obtained in time series within this range is calculated by a correlation calculation.
The position of the image signal having a high degree of coincidence between the two images is searched from the correlation value, and the phase difference is calculated using the image signal at this position.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail based on illustrated embodiments.

FIG. 1 to FIG. 3 are explanatory diagrams for explaining the detection principle of the focus detection device according to the present invention.

FIG. 1 shows a light beam in a focused state. In the focused state shown in FIG. 1A, a light beam 23a emitted from a photographing lens 20 is a photographing lens on a light receiving surface of an image pickup device 22. The optical axis 24 is focused, and the defocus amount at this time is zero. Then, as shown in FIG. 1B, an aperture (first pupil area) 21b having an opening above the taking lens 20
Light beam 2 emitted from the taking lens 20 even if
3b is incident on the light receiving surface of the image sensor 22 at the position of the taking lens optical axis 24. Then, as shown in FIG. 1C, a diaphragm (second pupil region) 21c having an opening below the taking lens 20
, The light flux 23c is incident on the light receiving surface of the image sensor 22 at the position of the photographing lens optical axis 24.

In this way, in the focused state, the light beam transmitted through the first pupil region (hereinafter simply referred to as "first pupil") and the second pupil region (hereinafter simply referred to as "second pupil") are used. ) Is transmitted to the same position on the image sensor 22,
In this case, the phase difference between the two images is zero.

FIG. 2 shows a light beam in the front focus state. In the front focus state of FIG.
Light beam 25a that has been injected is focused in front by d _a from the light receiving surface of the image sensor 22, before defocus amount d _a has a focus state from. FIG. 2B shows a light beam when a stop (first pupil) 21 b having an opening above the photographing lens 20 is placed. A light beam 25 b emitted from the photographing lens 20 is a light receiving surface of the image sensor 22. Above, it is incident downward by "δ _a / 2" from the optical axis 24 of the taking lens. When a stop (second pupil) 21c having an opening below the taking lens 20 is placed as shown in FIG. 2C, the light beam 25c emitted from the taking lens 20 is received by the image sensor 22. "Δ _a / 2" from the optical axis 24 of the taking lens 20 on the surface
Incident only upward.

[0025] During pre-focus state in this manner, the light beam having passed through the light beam transmitted through the first pupil and a second pupil is on the image pickup element so that the phase difference [delta] _a is generated.

FIG. 3 shows the luminous flux in the back focus state. In the back focus state of FIG.
Light beam 26a which has been emitted from 0 focused just behind d _b from the light receiving surface of the imaging device 22, in a state of defocus amount d _b after pin. FIG. 3B shows the photographing lens 20.
Is a light beam when an aperture (first pupil) 21b having an opening above the lens is placed. A light beam 26b emitted from the photographing lens 20 is transmitted to the image sensor 22 from the photographing lens optical axis 24 by “δ”.
_b / 2 ". Then, as shown in FIG. 3C, when a stop (second pupil) 21c having an opening below the taking lens 20 is placed, the light flux 26c emitted from the taking lens 20 is applied to the image pickup device 22 by the taking lens. The light enters the optical axis 24 downward by “δ _b / 2”.

[0027] During post-focus state in this manner, so that the phase difference between the first light flux transmitted through the pupil of the light beam transmitted through the second pupil on the imaging device [delta] _b is generated.

Here, it is assumed that the phase difference when the subject image by the light beam transmitted through the second pupil is above the position of the subject image by the light beam transmitted through the first pupil is a positive value. The phase difference in the pre-focus state of FIG.
_a ”and the phase difference in the back focus state of FIG. 3 is“ −δ _b ”
The magnitude and direction of the defocus amount can be determined from the magnitude and sign of the phase difference.

FIG. 4 is a perspective view showing a part of an optical system of a camera using the focus detecting device according to the embodiment of the present invention, and FIG. 5 is a central sectional view of FIG.

In these figures, reference numerals 60 and 61 denote photographing lenses, and a lens 61 is a focus adjusting lens, which is driven and controlled by a lens driving motor (not shown) to perform focus adjustment. 62 is an optical low-pass filter, 63 is an infrared cut filter, 64 is a CCD of a solid-state image sensor that converts an optical image into a video signal by photoelectric conversion, and 65 is a focus detection aperture, which is provided in the optical path of the photographing lens by a motor 66. You can go back and forth. 67 is a light shielding plate for focus detection,
It is held by the motor 68 so as to be able to advance and retreat in the optical path of the taking lens.

FIG. 6 is a block diagram schematically showing an electrical configuration of a camera using the focus detection device according to the embodiment of the present invention.

In the figure, reference numeral 70 denotes a CCD (corresponding to the CCD 64 in FIG. 4 and the like) which is a solid-state image sensor, and an analog video signal from the CCD 70 is converted into a digital signal by an A / D converter 71. Then, it is sent to the digital signal processing unit 72. 73 is an electronic viewfinder (EV
F) A VRAM as storage means for holding image data to be displayed in 76, a D / A converter 74 for converting the image data from the VRAM 73 into an analog signal, 75
Is an LCD driver, and an electronic viewfinder (EV)
F) to cause the EVF 76 to display the image output from the VRAM 73.

Reference numeral 77 denotes a buffer memory such as a DRAM for temporarily storing digital video signals and various data;
Reference numeral 78 denotes a memory for recording and storing photographed image data; 79, a system control unit for controlling the entire camera; 80, a display member for displaying settings and control status of the camera; and an LCD driver for driving and controlling 81. And 82, a CCD driver for driving and controlling the CCD 70; 83, a lens control unit for driving and controlling an aperture and a light shielding plate for focus detection; an aperture for photography, a lens for focus adjustment, etc .; And an operation switch for detecting a release operation.

The above is the main configuration of the camera according to the present invention.

FIG. 7 is a diagram for explaining the operation of the focus detecting aperture 65 and the focus detecting light shielding plate 67 shown in FIGS. 4 and 5. When the focus detecting operation is not performed, FIG.
As shown in (a), the focus detection aperture 65 and the light shielding plate 67 are retracted outside the pupil shape 69 (the area through which the effective light beam passes) when the imaging aperture of the imaging lens is opened.
When the focus detection operation is performed, as shown in FIGS. 7B and 7C, the focus detection diaphragm 65 is moved and held in the optical path of the photographing lens by the motor 66, and the focus detection light shielding plate 67 is moved by the motor 68. It moves to cover one of the openings 65a and 65b of the focus detection diaphragm 65, and shields the light.

Here, an image signal obtained by a light beam passing through the left pupil in the state of FIG. 7B is denoted by L, and FIG.
Let R be an image signal obtained by a light beam that has passed through the right pupil in the state of FIG.

Next, a method of driving a CCD which is a solid-state image sensor will be described with reference to FIGS. 8, 9 and 10. FIG.

In the pupil time-division phase difference focus detection apparatus according to this embodiment, the image signal L based on the light beam passing through the left pupil
After capturing the image signal, the image signal R is captured by the light beam that has passed through the right pupil. The shorter the time interval for capturing the image signals L and R, the smaller the focus detection error due to the movement of the subject image due to camera shake or the like. Therefore, it is desirable that this time interval be as short as possible.

Therefore, like reading an image at the time of photographing,
When pixel data of the entire screen of the CCD is read, there is a problem that the time required for reading becomes long and the time interval for taking in the image signal becomes long. Therefore, at the time of capturing the image signal for focus detection, the reading is performed at a higher speed than the normal reading of all the pixels. The method will be described below.

FIG. 8 is a schematic diagram of an interline type CCD. In the figure, 31 is a pixel, 32 is a vertical charge transfer element, 33 is a horizontal charge transfer element, and 34 is an output unit.

The signal charge photoelectrically converted by the pixel 31 is sent to the vertical charge transfer element 32, where the four-phase drive pulses φV1, φV
V2, φV3, and φV4 sequentially transfer in the direction of the horizontal charge transfer element 33. The horizontal charge transfer element 33 outputs the signal charges for one horizontal row transferred from the vertical charge transfer element 32 to the output unit 34 by the two-phase drive pulses φH1 and φH2.
Where it is converted to a voltage and output.

FIG. 9 shows the image pickup area of the CCD. In this embodiment, in order to speed up the read operation, only the area 41 used for focus detection is read at a normal read speed. The other areas 42 and 43 perform sweep transfer for reading at high speed.

FIG. 10 shows the vertical charge transfer device 3 of the CCD.
6 is a timing chart for one vertical synchronization period when 2 is a four-phase drive.

In the figure, VD is a vertical synchronizing signal, and a vertical blanking period is indicated by a Low potential. HD is a horizontal synchronizing signal, and a horizontal blanking period is indicated by a Low potential. φV1, φV2, φV3, φV4 are four-phase driving pulses of the vertical charge transfer element 32, and 51 and 52 are pixels 31
Is a read pulse for transferring the signal charges photoelectrically converted to the vertical charge transfer element 32. Of the four-phase drive pulse,
Numerals 53 and 54 are high-speed sweeping transfer pulses for transferring the signal charges transferred to the vertical charge transfer elements 32 in the regions 42 and 43 in FIG. 9 at high speed, respectively, and generate drive pulses faster than a normal drive pulse.

As described above, by sweeping out the area not used for focus detection at a high speed, the time required for the operation of reading the image signal can be shortened, and the focus detection accuracy can be prevented from lowering. Can be

Next, a method of obtaining a phase difference by correlation calculation will be described with reference to FIGS.

FIG. 11 shows an image signal L obtained by a light beam passing through the left pupil and an image signal R of the right pupil, and the two image signals have a phase difference of δ. here,
The image signal L is composed of values l ₁ to l ₂₃ corresponding to the signal charges of each pixel, and the image signal R is also composed of r ₁ to r ₂₃ . First, a correlation amount C (τ) is calculated from the image signals L and R. Here, C (τ) is calculated as in the following equation.

Here, max [l (i + 6), r (i +
τ + 6)] selects the larger value of l (i + 6) and r (i + τ + 6). Then, the value of τ is changed from −6 to 6.

FIG. 12 is a graph showing a change in the correlation amount C (τ). When the value of τ approaches the phase difference δ, C (τ)
Becomes smaller. Then, the correlation amounts C (3), C (4), C
Using (5), C (δ) at which the correlation between C (3) and C (4) is minimal is calculated by interpolation, and when C (δ) at which this correlation is minimal, Is the phase difference δ. In this manner, the phase difference between the image signals L and R can be calculated, and the maximum and minimum values of τ (maximum shift amount) and the number of pixel data are not limited to those in the above embodiment.

Next, a method of correcting a phase difference when the subject image on the CCD moves in the vertical direction due to camera shake will be described.

FIG. 13A shows a case where the camera is tilted downward due to camera shake before capturing the subject image R after capturing the subject image L by the light beam passing through the left pupil. Thus, the subject image moves upward. In this case, the image signal R _c in the same position as the image signal L _c is, L
_Since the user is looking at the object below _c , an image signal having a different shape is obtained. When a correlation operation is performed using the image signals having different shapes, the obtained phase difference also has an erroneous value.

Then, the image signal _Lc and the image signals R _a , R _b ,
R _c, R _d, performs a correlation calculation of R _e, the correlation amount minimum value C
_a (δ), C _b (δ), C _c (δ), C _d (δ), C _e (δ) and phase difference δ
_{a, δ b, δ c,} δ d, and calculates the [delta] _e. A minimum value among the minimum values C _a (δ) to C _e (δ) of the correlation amounts indicates that the degree of coincidence between the two images is the highest. Therefore, an image signal in which the minimum value of the correlation amount is the minimum is looking at the same part of the object, and if the phase difference at this time is used, the error generated by the vertical camera shake is also minimized. Moving the position of the image signal for which the phase difference is calculated using this is the principle of the correction method for reducing the detection error due to camera shake in the vertical direction.

The image signal _Lc and the image signals R _a , R _b ,
R _c, R _d, the minimum value of the correlation of _{R e C a (δ),} C b (δ), C
FIG. 14 plots _c (δ), C _d (δ), and C _e (δ). Since R _b is looking at the same position as L _c ,
The value of _b (δ) is the smallest. Then, the image signal L
By using the phase difference [delta] _{b c} and R _b in the calculation of the defocus amount, it is possible to significantly reduce the error caused by the deflection vertical hand.

Here, in the above embodiment, CC
Although the correction was performed every pixel of D line, the actual amount of movement due to camera shake is not every single line. Therefore, in order to perform correction with higher accuracy, an image signal generated every 0.5 line is generated by interpolation. Just do it.

FIG. 15 shows a method of generating an image signal R _ab by interpolation using the image signals R _a and R _b, and shows an image signal at intervals of 0.5 lines.

[0056] Each pixel of the image signal R _a output R _a1 to R _an, and R
_Using each of the pixels R _{b1 to} R _bn of _b , the average value of the output of the corresponding pixel can be generated as R _{ab1 to} R _abn, and R _ab1 = (R _a1 + R _b1 ) / 2 and , R _abi = (R _ai + R _bi ) / 2 i = 1 to n. Then, in order to perform an interpolation calculation in increments of 0.5 lines, the calculation of the average value may be changed to a weighted average. By doing so, finer correction in the vertical direction can be performed, and the detection accuracy can be improved.

By the way, in an object whose shape in the vertical direction changes little (an object whose spatial frequency is low), even if the minimum value C (δ) of the correlation amount changes with the position of the image signal, it does not change much. Also, C is determined by the noise component of the image signal.
Since the value of (δ) also fluctuates to some extent, using an image signal at a location where C (δ) is minimized in such a case may use an image signal at an incorrect position.

Therefore, as shown in FIG. 16, the minimum value C _c (δ) of the correlation amount of the image signal R _{c at} the same position as the image signal L _c and the most
When the difference of C _b (δ) where (δ) is small is smaller than a predetermined value, there is a possibility that the above-mentioned erroneous judgment is made. Therefore, the accuracy is improved by using the image signal R _c. Although improvement cannot be made, it is possible to prevent a situation in which an incorrect decision causes an adverse effect and further lowers the accuracy. This is particularly effective when the subject has an oblique line pattern. In this case, although the value of C (δ) is the same at any position, the value of the phase difference δ is greatly different, so there is a possibility that an erroneous determination may be made. This is because if the determination is too high, the detection accuracy is greatly reduced if an erroneous determination is made. In the case of such a subject, the detection error of the phase difference due to the movement in the vertical direction behaves in the same manner as the error of the phase difference due to the movement in the horizontal direction. This is not a problem because the error is corrected.

The amount of vertical movement of the image signals L and R caused by camera shake increases as the focal length of the photographic lens increases, and increases as the time interval between the capture of the image signals L and R increases. Therefore, by setting the width of the calculation area for performing vertical correction from the capture time interval of the image signal obtained from the focal length information and the accumulation time of the photographing lens and the drive time of the focus detection light shielding plate,
It is not necessary to perform an operation on an unnecessarily large area, the operation time can be reduced, and the amount of image signals to be read can be reduced, so that less memory is used and the time required for reading is reduced. effective.

FIG. 17 is a flowchart of the "shift range calculation" subroutine for setting the calculation area. The operation starts from step (2402) via step (2401).

First, at step (2402), the accumulation time TC is calculated from the luminance information of the subject, and the next step (2)
In 403), the driving time TD of the focus detection light shielding plate 67 is calculated, and in the following step (2404), the image signal capturing time interval TA is calculated as the sum of the accumulation time TC and the driving time TD. In step (2405), the focal length FA of the photographing lens is detected by reading the zoom encoder of the zoom lens.

In step (2406), it is determined whether or not the image signal capturing time interval TA is smaller than 10 msec. If "TA <10 msec", step (2407) is performed.
The process proceeds to step (2412) if not. In step (2407), it is determined whether or not the focal length FA is smaller than 30 mm. If “FA <30 mm”, the flow proceeds to step (2410); otherwise, the flow proceeds to step (2408). In step (2410), 1 is input to LS so that the calculation area is set for each of the upper and lower lines. In step (2808), it is determined whether or not the FA is smaller than 60 mm. If “FA <60 mm”, the flow proceeds to step (2411), otherwise, to step (211).
Go to 409). In step (2411), 2 is input to LS to set the calculation area and the upper and lower two lines, and in step (2409), 3 is input to LS.

In this way, when TA is 10 msec and “F
If A <30 mm, “LS = 1” and “30 mm ≦ F
If A <60 mm, “LS = 2” and “FA ≧ 60”
mm ”,“ LS = 3 ”. Step (241
2) to (2417), similarly, “10 msec ≦ TA ≦
20 msec ”, if“ FA <30 mm ”,“ LS = 2 ”
And "LS = 4" if "30 mm ≦ FA <60 mm"
If “FA ≧ 60 mm”, “LS = 5”.
Similarly, in steps (2418) to (2422), if “TA ≧ 20 msec” and “FA <30 mm”, “LS = 3” is set; if “30 mm ≦ FA <60 mm”, “LS = 5 ”, and if“ FA ≧ 60 mm ”,“ LS ”
= 6 ".

As described above, by changing the calculation area LS according to the focal length of the photographing lens and the time interval for capturing the image signal, it is possible to eliminate the need for performing unnecessary calculations and reading out the image signal. is there. Here, the method of determining the calculation area LS is such that the calculation area LS is reduced (the number of lines is reduced) when the focal length of the imaging lens is short or the capture time interval of the image signal is short. Needless to say.

FIG. 18 is a diagram for explaining the principle of removing a detection error due to camera shake in the first embodiment.
It is assumed that the subject image is moving from left to right at a constant speed due to camera shake. Here, it is assumed that the capture interval of the image signal is constant. The image signals actually captured are L ₁ , R ₂ , and L ₃ , and R ₁ , L ₂ , and R ₃ are the other image signals when they are captured at the same timing.

After the image signal L ₁ is captured, the image signal R
_{Since the} subject image has moved by δ _m before the image is captured, the phase difference δ ₁₂ between L ₁ and R ₂ becomes the true phase difference δ ₁₁
Δ _m , and δ ₁₂ = δ ₁₁ + δ _m (2) The subject image similarly moves by δ _m from the time when the image signal R ₂ is captured to the time when the image signal L ₃ is captured, but the phase difference δ ₂₃ between L ₃ and R ₂ is smaller by δ _m. Δ ₂₃ = δ ₁₁ −δ _m (3) The [delta] _12, to determine the true phase difference [delta] ₁₁ from [delta] _23, may be as follows.

Δ = (δ ₁₂ + δ ₂₃ ) / 2 = (δ ₁₁ + δ _m + δ ₁₁ -δ _m ) / 2 = δ ₁₁ (4) By calculating the phase difference δ in this way, It is possible to remove a detection error caused by the movement of the subject due to the fluctuation of the speed. For this reason, when the time interval of taking in the image signal is sufficiently short, the above assumption is satisfied. Therefore, when the driving speed of the focus detection light shielding plate 67 is high and the accumulation time is short, it is very effective. Is the way.

In the above description, the time required to drive the focus detecting light shield 67 from the right pupil to the left pupil of the focus detecting aperture 65 and the focus detecting light shielding from the left pupil to the right pupil are described. The time required to drive the plate 67 is the same, and the accumulation time of the CCD is the same for all three times, that is, when the image signal capturing time interval is the same. The accumulation time may fluctuate. Therefore, a description will be given of a calculation method when the capture time interval of the image signal changes.

FIG. 19 shows the relationship between the image signal fetch time and the position of the subject image. The time interval between the image signals L ₁ and R ₂ is T ₁₂ , and the time interval between R ₂ and L ₃ is T. _23, the phase difference between L ₁ and ₁₂ phase difference R ₂ is [delta], R ₂ and L ₃ is [delta] _23.

First, the slope of the solid line of the image signal L is (δ ₂₃ −δ ₁₂ ) / (T ₁₂ + T ₂₃ ), and the phase difference between L ₁ and L ₂ is T ₁₂ (δ ₂₃ −δ ₁₂ ) / ( T ₁₂ + T ₂₃ ).

The phase difference δ between L ₂ and R ₂ is And a weighted average by the time intervals T ₁₂ and T ₂₃ is obtained.

Next, the automatic focusing apparatus for a camera having the above configuration will be described with reference to the following flowchart.

FIG. 20 is a flow chart of the sequence of the entire camera. When the power switch is turned on and power supply to the circuit is started, steps (001) to (00) are performed.
Proceed to 2) to display the viewfinder. Here, CCD
Then, after performing various signal processing on the video signal, the video signal is displayed on the electronic viewfinder EVF. In the next step (003), the state of the switch SW1 that is turned on by pressing the release button in the first stage is detected, and if it is off, the process proceeds to step (004), and the flag JF for detecting the state of focus adjustment is initialized. . If the switch SW1 is on, the process proceeds to step (005).

In step (005), the output of the video signal captured by the CCD and the gain of the signal processing circuit, CC
The brightness of the subject is calculated from the accumulation time of D and the F value of the photographing lens. In the next step (006), an appropriate shutter time (exposure time) is obtained from the subject brightness calculated in the above step (005), the photographing mode of the camera, and the exposure correction information.
And an aperture value (F value) are calculated and stored in a predetermined memory area. When performing a release operation described later, the shutter and the aperture are controlled based on the data stored in the memory area.

In the next step (007), the state of the flag JF for detecting the state of focus adjustment is determined, and "JF =
If it is “1”, it indicates that the focus adjustment has already been completed and the camera is in focus. In such a case, the process proceeds to step (012), while “JF = 0” indicates that the camera is still in focus. If not, step (0) is performed to perform focus adjustment.
08).

Step (008) is a "focus detection" subroutine for detecting the amount of defocus of the photographing lens.
A detailed description will be given later. In the next step (009)
The defocus amount detected in the above step (008) is compared with the allowable defocus amount obtained from the permissible circle of confusion and the F value of the photographing lens. Proceeds to step (010), 1 is input to the flag JF to store that the camera is in focus, and the routine proceeds to step (012).

If it is determined in step (009) that the object is not in focus, the process proceeds to step (011), and the focus adjusting lens of the photographing lens is removed so as to eliminate the defocus amount detected in step (008). The group is driven and controlled, the focus is adjusted, and the process returns to step (002).

As described above, the focus adjustment is repeated until the focusing state is achieved or the switch SW1 is turned off.

When the focus state has been reached and the process proceeds to step (012), the state of the switch SW2 which is turned on when the release button is pressed in the second stage is detected.
If W2 is off, the process returns to step (002), while if the switch SW2 is on, the process proceeds to step (013) to perform a release operation.

In step (013), the aperture of the photographing lens is driven and controlled to the aperture value calculated in step (006). In the next step (014), the shutter is closed once, and the charge in the CCD is refreshed. The shutter is controlled at the shutter time calculated in step (006) to expose the CCD. The following steps (01
In 5), the CCD is driven to read out a video signal and perform predetermined signal processing. Then, the video signal is subjected to compression processing in step (016), and the compressed video signal is recorded in the memory of the recording medium in step (017).

In step (018), the shutter is driven to the open state, and when the shutter is returned to the initial state, step (002) is performed.
Return to

Next, the operation of the "focus detection" subroutine will be described with reference to the flowchart of FIG.

When the focus detection subroutine is called in step (008) of FIG. 20, the process proceeds to step (102) via step (101).

In step (102), the focus detection diaphragm 65 in the state shown in FIG. 7A is driven to move into the optical path of the photographing lens. In the next step (103), FIG.
The focus detection light-shielding plate 67 in the state of (a) is moved into the optical path of the taking lens. The above steps (102), (1)
03), the focus detection aperture 65 and the focus detection light shielding plate 67 are in a state as shown in FIG. 7B, and only the light beam that has passed through the left opening 65a of the focus detection aperture 65 is on the CCD. To form an image.

[0085] At step (104), the accumulation of the CCD charge, in the next step (105), reads out the image signal L ₁ accumulated in the step (104),
It is stored in a predetermined memory area. Here, the read operation image signal L ₁ is different from the normal read operation, as described above, the read only the image signal of the area required for the focus detection,
An image signal in an unnecessary area is swept out at a high speed, thereby reducing an image signal readout time.

In step (106), the focus detection light shielding plate 67 is driven again, and only the light beam that has passed through the opening 65b on the right side of the focus detection stop 65 as shown in FIG.
An image is formed on D. Then, the following steps (1
In step (07), electric charges are accumulated in the same manner as in step (104). In step (108), step (1) is performed.
05) as well as read the image signals R _2, stored in a predetermined memory area.

In step (109), the light-shielding plate 67 for focus detection is driven again to bring the state as shown in FIG.
Only the light beam that has passed through the opening 65a on the left side of the focus detection diaphragm 65 forms an image on the CCD. Then, in step (110), the CCD of the storage operation, the next step (111), reads out the image signals L _3, and stores in a predetermined memory area.

In step (112), the focus detection diaphragm 65 is driven to retract, and in the next step (113), the focus detection light shield 67 is driven to retract. When the operations of steps (112) and (113) are completed, FIG.
The state returns to the initial state as shown in FIG.

Step (114) is a "defocus amount calculation" subroutine, which will be described in detail later.
_The defocus amount of the photographing lens is calculated using ₁ , R ₂ and L ₃ . When the step (114) is completed, the "focus detection" subroutine is returned in the step (115).

The steps (104) and (10)
By making the accumulation times and gains in the accumulation operations 7) and (110) the same, the output levels of the image signals can be made the same, and the correlation in performing the correlation operation can be increased. The detection accuracy can be made high and stable.

Next, the operation of the "defocus amount calculation" subroutine will be described with reference to the flowchart of FIG.

When the "defocus amount calculation" subroutine is called in step (114) of FIG. 21, the process proceeds to step (202) via step (201).

[0093] At step (202), the correlation operation described earlier, calculates the phase difference [delta] ₁₂ of the image signals L ₁ and R _2. In step (203), in the same way to calculate the phase difference [delta] ₂₃ of the image signals L ₃ and R _2. In the following step (204), a phase difference δ from which a detection error due to a finally obtained constant-speed camera shake component is removed is calculated. The arithmetic expression at this time is as follows.

Δ = (δ ₁₂ + δ ₂₃ ) / 2 (6) In the next step (205), the above step (204)
The defocus amount DF of the photographing lens is calculated based on the phase difference δ obtained in the above, the sensitivity K of the focus detecting system determined by the focus detecting aperture and the photographing lens optical system, and the pixel pitch P of the CCD. The arithmetic expression at this time is as follows.

DF = δ · K · P (7) Then, when the above step (205) is executed, the subroutine returns in step (206).

Next, another example in which the driving time of the focus detection light shielding plate 67 fluctuates and the time interval for taking in the image signal fluctuates will be described. Here, since the configuration of the camera and the sequence of the entire camera are the same, different portions of the focus detection method will be described.

FIG. 23 is a flowchart of the "focus detection" subroutine. When the "focus detection" subroutine is called in step (008) of FIG. 20, step (3) is executed.
01), and proceeds to step (302).

In steps (302) and (303), similarly to steps (102) and (103) in FIG. 20, the focus detection aperture 65 and the focus detection light shielding plate 67 are driven.
In the state (b), a preparation operation for focus detection is performed.

In the next step (304), the timer value TIMER of the self-running timer of the system control unit is set to RA
By storing in the storage area T ₁ of the on M, the image signal L
₁ stores the start time of the accumulation operation. Step (3
In 05), the storing operation image signals L _1, the following step (306), executes the read operation image signal L _1.

[0100] At step (307), by the state shown in FIG. 7 (c) and drives the focus detecting light shielding plate 67, which stores the following step (308), the timer value TIMER in a memory area T _2, the image storing the starting time of the accumulation operation signal R _2. In steps (309) and (310),
To accumulate and read an image signal R _2.

[0102] At step (311), the state of FIG. 7 (b) by driving the focus detecting light shielding plate 67, the following step (312), the accumulating operation starting time of the image signal L ₃ in the storage area T ₃ stored, step (313), the accumulation and reading of the image signals L ₃ at (314).

In steps (315) and (316), the focus detecting aperture 65 and the focus detecting light shielding plate 67 are retracted to the state shown in FIG. 7A, and the defocus amount is calculated in step (317). In step (318), this subroutine is returned.

Here, the method of calculating the defocus amount will be described in detail with reference to FIG.

FIG. 24 shows a "defocus amount calculation" subroutine in the case where the time interval of image signal capture varies. When this subroutine is called in step (317) of FIG. 23, the process goes through step (401). , Proceed to step (402).

In step (402), the image signals R ₂ and L
₁ calculates the accumulation start time interval T _12, the next step (403), calculates the image signals L ₃ and the accumulation start time interval T ₂₃ of R _2. In the next step (404), determined by the correlation calculation of the image signals L ₁ and the phase difference [delta] ₁₂ in R _2, the following step (405), obtains the image signal L ₃ and the phase difference [delta] ₂₃ of R ₂ by correlation calculation .

In step (406), the phase differences δ ₁₂ and δ
_The phase difference δ from which the detection error caused by the movement of the optical image at a constant speed is removed by the following equation _{23 is obtained} by using the time intervals T ₁₂ and T ₂₃ .

In step (407), the defocus amount DF is calculated from the phase difference δ, the sensitivity K of the focus detection system, and the pixel pitch P of the CCD obtained in step (406). Step (408)
To return this subroutine.

As described above, when the time interval at which the accumulation operation is started fluctuates, a memory for measuring the time interval is required, and the arithmetic processing becomes complicated. Therefore, if the control is performed so that the time interval is constant, there is an effect that the arithmetic processing is simplified, the processing time can be shortened, and the memory usage can be reduced.

That is, if the time required for driving the focus detection light-blocking plate 67 fluctuates to some extent, the upper limit value of the drive time is known. , The accumulation operation is started, and the same accumulation time is used, so that the time interval at which the accumulation operation is started can be constant, and the arithmetic processing can be simplified easily. Can be. By providing a timer for stabilizing the start time interval of such an accumulation operation, there is also an effect that it is not necessary to provide a detection means for detecting the completion of driving of the focus detection light shielding plate 67.

(Second Embodiment) A second embodiment of the present invention will be described.
In the embodiment, four image signals are fetched in a time series to remove a detection error caused by movement of an optical image having constant acceleration due to camera shake or movement of a subject.

FIG. 25 is a diagram for explaining the principle of removing a detection error due to a camera shake according to the second embodiment. The image of the subject moves from left to right at a constant acceleration due to the camera shake. Here, it is assumed that the time interval for capturing the image signal is constant.

The image signals L ₁ , R ₂ , L ₃ , and R ₄ are image signals that are actually captured, and are the image signals R ₁ , L ₂ , R ₃ , and L.
Reference numeral ₄ denotes another image signal that is supposed to be taken in at the same timing.

After the image signal L ₁ has been fetched, the image signal R
₂ to the subject image is moved by [delta] _m1 between captured, since the captured image signals R _2, subject image until the image signals L ₃ incorporated is moved by [delta] _{m @ 2,} this time " _δm1 < _δm2 ”. Therefore, as in the first embodiment described above, when obtaining the phase difference [delta] ₁ with the image signal _{L 1, R 2, L 3} , δ is a small value by [delta] _d1 than the true phase difference [delta] ₁₁ Become.

Δ ₁ = (δ ₁₂ + δ ₂₃ ) / 2 = δ ₁₁ −δ _d1 (9) This δ _d1 is an error generated by a change in speed due to acceleration. Next, the image signals R ₂ , L ₃ , R ₄
And the phase difference δ in the same manner as in the first embodiment.
When seeking _2, a large value only [delta] _d2 than the true phase difference [delta] _11. Here, “δ _m2 <δ _m3 ”.

Δ ₂ = (δ ₂₃ + δ ₃₄ ) / 2 = δ ₁₁ + δ _d2 (10) Here, if moving at a constant acceleration, the magnitudes of δ _d1 and δ _d2 are the same. δ _d1 = δ _d2 (11) Thus, the phase difference δ from which the detection errors δ _d1 and δ _d2 have been removed is obtained as in the following equation.

Δ = (δ ₁ + δ ₂ ) / 2 = (δ ₁₁ + δ _d1 + δ ₁₁ + δ _d2 ) / 2 = δ ₁₁ (12) = (δ ₁₂ + 2δ ₂₃ + δ ₃₄ ) / 4 (4) 13) Thus, the four image signals L ₁ , R ₂ , L ₃ , R ₄
Based on the phase differences δ ₁₂ , δ ₂₃ , and δ ₃₄ determined from the above, it is possible to remove a detection error caused by the movement of the uniform acceleration, and to greatly improve the detection accuracy.

Next, a description will be given of a method of calculating the phase difference δ in the case where the time interval for taking in the image signal always changes.

The time required to capture the image signal R ₂ from the capture of the image signal L ₁ is T ₁₂ , the time required to capture the image signal L ₃ from the capture of the image signal R ₂ is T ₂₃ , and the image signal L ₃
The time required from the capture of the image signal to the capture of the image signal R ₄ is T
_Assuming that ₃₄ , the phase differences δ ₁ , δ
₂ is as follows.

Then, the phase difference δ is as follows.

Here, assuming that the time required to move the focus detection light shielding plate 67 in the same direction is constant, and that the accumulation time is the same in one focus detection operation, T ₁₂ = T ₃₄ ... (18) Then, the above arithmetic expression is as follows.

As described above, by giving a certain regularity to the image signal capturing time interval, the arithmetic expression can be greatly simplified.

Since the configuration and the rough operation of the camera according to the second embodiment are the same as those of the first embodiment,
The description here is omitted, and a focus detection operation different from the first embodiment will be described.

FIG. 26 is a flowchart of the "focus detection" subroutine in the case where the time interval of taking in the image signal is constantly changing. When the "focus detection" subroutine is called, the process proceeds from step (501) to step (502). Proceed to.

In steps (502) and (503), the focus detection aperture 65 and the focus detection light shielding plate 67 are driven to perform a preparation operation for focus detection to bring the state shown in FIG. 7B.

[0125] At step (504), by storing the timer value TIMER of the free-running timer of the system controller in the storage area T ₁ of the on RAM, storing a time to start the accumulation operation image signals L ₁ . In step (505), the storing operation image signals L _1, step (506), executes the read operation of the image signal L _1.

In step (507), the light-shielding plate 67 for focus detection is driven to the state shown in FIG.

In the next step (508), the image signal R ₂
The accumulation operation start time stored in the storage area T ₂ of the, step (509), the storing operation image signals R _2, in step (510), performs the operation of reading the image signals R _2.

[0128] At step (511), FIG. 7 so that the state of (b) drives the focus detecting light shading plate 67, in step (512), stores the accumulated operation starting time of the image signal L ₃ region T ₃ stored, in step (513), the accumulation of the image signals L _3, step (514), the read operation of the image signal L _3.

In step (515), the focus detecting light-shielding plate 67 is driven so as to be in the state shown in FIG. 7C. In step (516), the accumulation start time of the image signal R ₄ is stored in the storage area T _4. In step (517), the image signal R ₄ is stored, and in step (518), the image signal R ₄ is read out.

In the following steps (519) and (520), the focus detection aperture 65 and the focus detection light shielding plate 67 are driven to the state shown in FIG. The following step (521)
After calculating the defocus amount in step (522), the subroutine is returned in step (522).

Here, the method of calculating the defocus amount will be described in detail with reference to FIG.

FIG. 27 is a flow chart of the "defocus amount calculation" subroutine, which corresponds to the step (52) in FIG.
When the "defocus amount calculation" subroutine is called in 1), the process proceeds to step (601) through step (601).
Proceed to 2).

In step (602), the image signals L ₁ and R
Time of _second storage start calculates the interval T _12, step (60
In 3), a time interval T ₂₃ between the start of accumulation of the image signals R ₂ and L _3.
, And in step (604), the image signals L ₃ and R ₄
It calculates the time interval T ₃₄ the start of the accumulation.

In the next step (605), the image signal L ₁
And determined by the correlation calculating the phase difference [delta] ₁₂ in R _2, in step (606), determined by a correlation computation with the phase difference [delta] ₂₃ of the image signals L ₃ and R _2, in step (607), image signal L
The phase difference δ ₃₄ between ₃ and R ₄ is obtained by a correlation operation.

In steps (608) and (609), the phase differences δ _{1 and} δ _{2 from} which the error of the constant velocity component has been removed are calculated. In step (610), the phase difference δ from which the error of the constant acceleration component has been removed is calculated. Calculate. In the next step (611), the defocus amount D is obtained from the phase difference δ from which the error of the uniform acceleration component has been removed, the sensitivity K of the focus detection system, and the pixel pitch P of the CCD.
F is calculated, and this subroutine is returned in the next step (612).

The focus detection operation in the case where the start time interval of the accumulation operation changes has been described above.

Next, when the time required to drive the focus detecting light-shielding plate 67 is different, the driving time is different when the driving direction is different, but when driving in the same direction, the driving time is stable to a substantially constant value. That is, there is a certain regularity in the start time interval of the accumulation operation, the time from the start of accumulation of the image signal L to the start of accumulation of the image signal R is constant in one focus detection operation, and The time from the start of the accumulation of R to the start of the accumulation of the image signal L is also constant in one focus detection operation, but in one focus detection operation, the start of the accumulation of the image signal L and the start of accumulation of the image signal R. Is different from the time from the start of accumulation of the image signal R to the start of accumulation of the image signal L. When expressed in the same manner as in the above-described embodiment, T ₁₂ = T ₃₄ , T ₁₂ ≠ T ₂₃ ... 20).

[0138] Figure 28 is a focus detection subroutine when there is the regularity in the time interval of starting accumulation, the image signal R ₄
Since the accumulation operation start time T ₄ the same as the flowchart of the subroutine of FIG. 26, except that no go Since it is not necessary to measure, a detailed description thereof will be omitted.

When the "defocus amount calculation" subroutine in step (720) is called, the flow advances to step (802) via step (801) in FIG.

In the step (802), the image signals L ₁ and R
The _second accumulation start time interval T ₁₂ is calculated, the next step (8
03), an accumulation start time interval T ₂₃ between the image signals R ₂ and L ₃ is obtained.
Is calculated.

In steps (804) to (806), similar to steps (605) to (607) in FIG.
The phase differences δ ₁₂ , δ ₂₃ , δ ₃₄ are calculated, and step (807)
Calculates the phase difference δ from which the error due to the change in the uniform acceleration is removed, and in step (808), calculates the defocus amount D from the phase difference δ, the sensitivity K of the focus detection system, and the pixel pitch P of the CCD.
After calculating F, the subroutine is returned in step (809).

The focus detection operation in the case where the time interval of the start of accumulation has regularity has been described above. By giving the regularity to the time interval in this way, the calculation of the phase difference δ can be performed with the regularity. Since a simpler operation can be performed than in the case where there is no operation, there is an effect that the memory required for the operation processing is reduced and the time required for the operation processing can be shortened.

Next, the focus detection operation in the case where the accumulation start time interval during one focus detection operation is constant will be described.

FIG. 30 is a flowchart of the focus detection subroutine in the case where the time interval of the accumulation start is constant. Since there is no need to measure the time interval, the operation of reading the time of each accumulation start is not performed. 26 is similar to the flowchart of FIG.
5 and the focus detection light shielding plate 67 are driven, and the image signals L ₁ , R
₂ , L ₃ and R ₄ are stored and read out, and detailed description is omitted.

When the "defocus amount calculation" subroutine is called in step (917), the flow advances to step (1002) via step (1001) in FIG. 30, and the "defocus amount calculation" subroutine is executed.

In steps (1002) to (1004), phase differences δ ₁₂ , δ ₂₃ and δ ₃₄ are calculated by the same correlation calculation as in steps (605) to (607) in FIG. Then, the phase difference δ from which the error generated by the movement of the uniform acceleration is removed is calculated. In the next step (1006), the defocus amount DF is calculated from the phase difference δ, the sensitivity K of the focus detection system, and the pixel pitch P of the CCD.
Is calculated, the subroutine is returned in step (1007).

The focus detection operation in the case where the time interval of the accumulation start is constant has been described above. The processing for calculating the phase difference δ in which the error due to the movement of the uniform acceleration is removed is further simplified, and the memory for calculation is reduced. In addition, there is an effect that the calculation time is shortened, and there is an effect that it is not necessary to store the time of the accumulation start.

Here, as a method of giving regularity to the time interval of the accumulation start or to make it constant, it is necessary to wait until a predetermined time has elapsed even if the drive of the focus detecting light shielding plate 67 is completed in a short time. This can be achieved by providing a stabilization timer so that the accumulation operation is started from the beginning. The value of this stabilization timer may be any value as long as it is larger than the maximum value of the time required for driving. By providing such a stabilization timer, there is also an effect that it is not necessary to provide a detecting means for detecting the completion of driving of the focus detection light shielding plate 67.

(Third Embodiment) A third embodiment of the present invention.
In the embodiment, five image signals are fetched in time series, the movement of an optical image caused by camera shake or movement of a subject is approximated by a quadratic function, and the detection error generated by this movement is reduced.

FIG. 32 is a diagram for explaining a method of reducing a detection error according to the third embodiment. The vertical axis represents the position y of the subject image and the horizontal axis represents time t. As the operation of the camera, the image signals L ₁ , R ₂ , L
₃ , R ₄ and L ₅ are taken in time series. Here, the image signal L
Movement of the object image while incorporating ₁ ~L ₅ is to be approximated to a quadratic function. Then, a quadratic function y = At + Bt ² that passes through the image signals L ₁ , L ₃ , and L ₅ is obtained, and the positions of the imaginary image signals L ₂ and L ₄ are obtained by interpolation using the quadratic function. the average value of the phase difference between the ₂ and the phase difference of R ₂ and L ₄ and R ₄ is for a phase difference δ finally determined.

Next, the calculation method will be described.

[0152] The position of the image signal L ₁ as shown in FIG. 32 y
₁ , the time is t ₁ , the position of the image signal L ₃ is y ₃ , and the time is t
_3, the position of the image signal L ₅ y _5, the time t _5, y ₁ =
When the quadratic function y = At + Bt ^{2 is obtained} with 0 and t ₁ = 0, y ₃ = At ₃ + Bt ₃ ² (21) y ₅ = At ₅ + Bt ₅ ² (22) The above two equations Thus, the coefficients A and B are obtained as in the following equations.

Then, the position y _{2 of} the imaginary image signal L _{2 and} the position y ₄ of the image signal L ₄ are as follows.

The phase difference δ between the imaginary image signal L ₂ and the image signal R _2
22 is as follows.

The phase difference δ ₄₄ between the imaginary image signal L ₄ and the image signal R ₄ is given by the following equation.

Here, t ₁ = 0, y ₁ = 0 (33), so that t ₂ = T ₁₂ , y ₂ ′ = −δ ₁₂ (34) t ₃ = T ₁₂ + T ₂₃ , y ₃ = −δ ₁₂ + δ ₂₃ (35) t ₄ = T ₁₂ + T ₂₃ + T ₃₄ , y ₄ ′ = −δ ₁₂ + δ ₂₃ −δ ₃₄ (36) t ₅ = T ₁₂ + T ₂₃ + T ₃₄ + T ₄₅ , Y ₅ = −δ ₁₂ + δ ₂₃ −δ ₃₄ + δ ₄₅ (37) where δ ₁₂ is the phase difference between the image signals L ₁ and R ₂ , δ ₂₃ is the phase difference between the image signals L ₃ and R ₂ , δ ₃₄ the phase difference of the image signals L ₃ and R _4, [delta] ₄₅ is the phase difference of the image signal L ₅ and R _4.

The equations (23) and (24) show that
Substituting Equation (37) gives Further, δ ₂₂ and δ ₄₄ are as follows.

Δ ₂₂ = A · T ₁₂ + BT ₁₂ ² + δ ₁₂ (40) δ _{44 =} A · (T ₁₂ + T ₂₃ + T ₃₄ ) + B · (T ₁₂ + T ₂₃ + T ₃₄ ) ² + δ ₁₂ − δ ₂₃ + δ ₃₄ (41) And the finally obtained phase difference δ is the average value of δ ₂₂ and δ ₄₄ Becomes This is a method for calculating the phase difference δ when the time interval for capturing the image signal is constantly changing.

Next, assuming that the time required to move the focus detection light shielding plate 67 in the same direction is constant and the accumulation time is the same during one focus detection operation, T ₁₂ = T ₃₄ , T _{12 23} = T ₄₅ (44) Substituting this into equations (38), (39) and (43) gives The above is the method of calculating the phase difference δ when the image signal capturing time interval has regularity.

Next, a description will be given of a method of calculating the phase difference δ in the case where the time interval of taking in the image signal is the same when performing one focus detection operation. Under these _{conditions, T 12 = T 23 = T} 34 = T 45 ......... (48) , and the Becomes

As described above, the arithmetic expression can be greatly simplified by giving regularity to the time interval of image signal capture or by setting the same value.

The configuration and main flow chart (rough operation) of the camera according to the third embodiment are the same as those of the first embodiment, and a description thereof will be omitted, and the first and second embodiments will be omitted. A focus detection operation different from the mode will be described.

FIG. 33 is a flowchart of a focus detection subroutine in a case where the time interval of image signal capture is constantly changing. When this subroutine is called, the process proceeds to step (1101) and then to step (1102).

In steps (1102) and (1103), the focus detection aperture 65 and the focus detection light
7 is driven to the state shown in FIG.

[0165] At step (1104), by storing the timer value TIMER of the self timer storage area T ₁ of the on RAM, storing a time to start the accumulation operation image signal L _1. In step (1105), the accumulation operation image signals L _1, step (1106),
It reads the image signal L _1.

[0166] At step (1107), and driven to a state of the focus detection light-shielding plate 67 and FIG. 7 (c), the step (1108), the accumulation start time of the image signals R ₂ T
₂ stores, in step (1109), the accumulation of the image signals R _2, in step (1110), to read the image signals R _2.

[0167] At step (1111), and drives the focus detecting light shielding plate 67 so that the state of FIG. 7 (b), step (1112), the accumulation start time of the image signal L ₃ T
Stored in _3, at step (1113), the accumulation of the image signals L _3, step (1114) reads the image signal L _3.

[0168] At step (1115), and drives the focus detecting light shielding plate 67 so that the state of FIG. 7 (c), the step (1116), the accumulation start time of the image signals R ₄ T
₄ stores, in step (1117), the accumulation of the image signals R _4, in step (1118) reads the image signal R _4.

[0169] At step (1119), so that the state of FIG. 7 (b), and drives the focus detecting light shading plate 67, in step 1120, stores the accumulation start time of the image signal L ₅ to T ₅ Then, in step (1121), the image signal L ₅
Is stored, and in step (1122), the image signal L ₅
Is read.

In steps (1123) and (1124), the focus detection aperture 65 and the focus detection light shielding plate 67 are driven to the state shown in FIG. Subsequent steps (112
When the defocus amount is calculated in 5), step (112)
In 6), this subroutine is returned.

Here, a method of calculating the defocus amount will be described with reference to FIG.

FIG. 34 is a flowchart of the "defocus amount calculation" subroutine.
When the "defocus amount calculation" subroutine is called in step 25), step (1201) is followed by step (1).
Go to 202).

[0173] Step (1202) - (1205), the time interval T ₁₂ the storage start of the image signal, T _23, T _34, T
₄₅ , and in steps (1206) to (1209), the phase differences δ ₁₂ , δ ₂₃ , δ ₃₄ , δ ₄₅ of the image signals are calculated by correlation calculation. In step (1210), the coefficient A of the first-order term of the quadratic function is calculated, and step (1211)
Then, the coefficient B of the quadratic term is calculated. In step (1212), the phase difference δ calculated by the quadratic function approximation is calculated. In step (1213), the defocus amount DF is calculated from the phase difference δ, the sensitivity K of the focus detection system, and the pixel pitch P of the CCD. Is calculated, and the subroutine is returned in the next step (1214).

The focus detection operation in the case where the start time interval of the accumulation operation changes has been described above.

Next, when the time required to drive the focus detection light shielding plate 67 is different when the driving direction is different, the driving time is different. However, when driving in the same direction, the driving time is constant. Become. That is, there is a certain regularity in the accumulation start time interval, and the times T ₁₂ and T ₃₄ from the start of accumulation of the image signal L to the start of accumulation of the image signal R are the same, and the accumulation of the image signal R to the image signal L time T ₂₃ of the accumulation to the start of, T ₄₅ but is the same, the time from the accumulation start time and image signals R from the storage start of the image signal L to the accumulation start of the image signals R to accumulation start of the image signal L Are different, and T ₁₂ = T ₃₄ , T ₂₃ = T ₄₅ , T ₁₂ T ₂₃ (54).

[0176] Figure 35 is a "focus detection" subroutine when the time interval of the accumulation start is the regularity, there is no need to measure the accumulation start time T _4, T ₅ of the image signals R _4, L ₅ Since it is the same as the flowchart of the “focus detection” subroutine of FIG.
The description here is omitted.

When the defocus amount calculation subroutine of step (1323) is called, step (1323) of FIG.
After step 401), the process proceeds to step (1402).

[0178] At step (1402), calculates the image signal L ₁ and the accumulation start time interval T ₁₂ of R _2, steps (14
03), an accumulation start time interval T ₂₃ between the image signals R ₂ and L ₃ is obtained.
Is calculated. In step (1404) - (1407), in the same manner as in step of FIG. 34 (1206) - (1209), calculates the phase difference _{δ 12, δ 23, δ 34} , δ 45, step (1408), 2 Coefficient A of the first-order term of the following function
In step (1409), the coefficient B of the quadratic term is calculated.

In step (1410), the phase difference δ obtained by the quadratic function approximation is calculated, and in step (141)
In 1), the defocus amount DF is calculated from the phase difference δ, the sensitivity K of the focus detection system, and the pixel pitch P of the CCD, and the subroutine is returned in the next step (1412).

The above is the description of the focus detection operation in the case where the time interval of the accumulation start has regularity. By giving the regularity to the time interval, the calculation of the phase difference δ does not have the regularity. Since the calculation is simpler than in the case, the memory required for the calculation processing is reduced, and the calculation time is shortened.

Next, a description will be given of the focus detection operation when the accumulation start time interval during one focus detection operation is constant.

FIG. 37 is a flowchart of the "focus detection" subroutine when the time interval of the accumulation start is constant.
Since there is no need to measure the time interval, the operation of storing the time of each accumulation start is not performed. For the other parts, the drive of the focus detection aperture 65 and the focus detection light shielding plate 67 is controlled in the same manner as in the flowchart of FIG.
₁ , R ₂ , L ₃ , R ₄ , and L ₅ are sequentially stored and read.

When the "defocus amount calculation" subroutine is called in step (1520), the flow advances to step (1602) via step (1601) in FIG. 37 to execute the defocus amount calculation subroutine.

In steps (1602) to (1605), phase differences δ ₁₂ , δ ₂₃ , δ ₃₄ and δ ₄₅ are calculated by the same correlation calculation as in steps (1206) to (1209) in FIG. ) Calculates the phase difference δ corrected by the quadratic function approximation. In the next step (1607), the phase difference δ, the sensitivity K of the focus detection system,
After calculating the defocus amount DF from the pixel pitch P of the CCD, the subroutine returns in step (1608).

The above is a description of the focus detection operation in the case where the time interval of the accumulation start is constant. The calculation processing of the phase difference δ corrected by the quadratic function approximation is further simplified, and the memory and time required for the calculation processing are reduced. Can be significantly reduced,
Furthermore, since there is no need to store the time of the start of accumulation, no memory or calculation time is required.

If the time required for driving the focus detecting light-shielding plate 67 fluctuates to some extent, a stabilization timer that starts accumulation and waits after a predetermined time has elapsed even if driving is completed early. With this arrangement, the start time interval of the accumulation operation can be easily made regular or constant. The time of the stabilization timer at this time may be a value larger than the maximum value of the drive time, and in this case, there is no need to provide a detection unit for detecting the completion of driving of the focus detection light shielding plate 67. There is also an effect.

In each of the above embodiments, the image signal is set to 3
Although the example in which the phase difference is calculated by correcting the movement of the subject image on the CCD due to the movement of the subject and the camera shake by using up to five images has been described, the movement of the subject image is corrected by using six or more image signals. May be. For example, the image signal accumulation time interval is 6
The phase differences δ ₁₂ , δ ₂₃ , δ ₃₄ , δ ₄₅ , δ ₅₆ are obtained from the image signals L ₁ , R ₂ , L ₃ , R ₄ , L ₅ , R ₆ and the phase difference δ is calculated as You may.

Δ = (δ ₁₂ + 3δ ₂₃ + 4δ ₃₄ + 3δ ₄₅ + δ ₅₆ ) / 12 (55) Similarly, from the seven image signals, the phase differences δ ₁₂ , δ ₂₃ ,.
δ ₃₄ , δ ₄₅ , δ ₅₆ , δ ₆₇ may be calculated, and the phase difference δ may be calculated from the following equation.

Δ = (δ ₁₂ + 3δ ₂₃ + 4δ ₃₄ + 4δ ₄₅ + 3δ ₅₆ + δ ₆₇ ) / 16 (fourth embodiment) In the fourth embodiment of the present invention,
In accordance with the focal length of the taking lens and the time interval for taking in the image signal, the number of image signals and the calculation method that can ensure sufficient detection accuracy under the conditions are selected.

FIG. 39 shows the relationship between various measures for reducing the influence of camera shake, the image signal capture time interval, and the focal length of the photographing lens (allowable focal length) in which the detection error due to camera shake becomes the same as the allowable error. It is the figure calculated | required by simulation.

Here, the phase difference between the image signals L ₁ and R ₂ is δ ₁₂ , the phase difference between L ₃ and R ₂ is δ ₂₃ , and the phase difference between L ₃ and R ₄ is δ ₃₄ , L ₅ and R ₄ . Assuming that the phase difference is δ ₄₅ and the image signal capturing time interval is constant, no countermeasure means that the phase difference δ = δ used in the defocus calculation
₁₂ and Measures 1 those, the phase difference for use in defocus calculation [delta] =
(Δ ₁₂ + δ ₂₃ ) / 2 Measure 2 is a phase difference δ =
(Δ ₁₂ + 2 · δ ₂₃ + δ ₃₄ ) / 4 Measure 3 is the phase difference δ used for the defocus calculation.
(Δ ₁₂ + 3 · δ ₂₃ + 3δ ₃₄ + δ ₄₅ ) / 8 The measure 4 is the phase difference δ used in the defocus calculation.
(Δ ₁₂ + 2δ ₂₃ + 2δ ₃₄ + δ ₄₅ ) / 6, which is a simulation result when the time interval for taking in the image signal does not change.

In this figure, the allowable focal length increases as the image signal capturing time interval decreases, and the allowable focal length increases as the number of image signals used in one operation increases.

Here, as the number of image signals used increases, the detection accuracy also improves. However, the time required to capture the image signals becomes longer, and the time required to detect the focus becomes longer. Gets worse.

Therefore, the minimum number of image signals required to obtain the required accuracy depends on the focal length and accumulation time of the photographing lens actually used for photographing, and the drive time of the focus detecting light shield 67. By calculating the calculation method, the time required for focus detection can be reduced as much as possible.

In FIG. 39, measure 4 is effective with respect to measure 3 when the image signal capture time interval is longer than 15 msec. However, when the time interval is less than 15 msec, the effect becomes counterproductive. Use, 1
If it is shorter than 5 msec, use measure 3.

Since the configuration and the rough operation of the camera according to the fourth embodiment are the same as those of the first embodiment,
The description here is omitted, and the focus detection operation will be described.

FIG. 40 is a flowchart of the "focus detection" subroutine. When this subroutine is called, the flow advances through step (1701) to step (1702).

In step (1702), step (0)
The accumulation time TC and the gain at the time of focus detection are calculated from the subject brightness obtained in the photometry subroutine of 05) and the aperture value of the aperture 65 for focus detection. In the next step (1703), the drive time T of the focus detection light shielding plate 67 for pupil switching is set.
D is read from the data written in the ROM, and in step (1704), the image signal capturing time interval TA is calculated as the sum of TC and TD.

In step (1705), the zoom encoder of the zoom lens is read, and the focal length FA of the photographing lens is detected. In step (1706), the number NA of image signals required for one focus detection is calculated from the image signal capture time interval TA and the focal length FA of the photographing lens. Since the detailed description of this calculation method will be described later, the description here is omitted.

At steps (1707) and (1708), the focus detection diaphragm 65 and the focus detection light
7 is driven so that the state shown in FIG.

In step (1709), it is determined whether the NA calculated in step (1706) is 2 or not.
If “A = 2”, the flow proceeds to step (1713); otherwise, the flow proceeds to step (1710). In the next step (1710), it is determined whether or not “NA = 3”.
If “N = 3”, the flow proceeds to step (1714); otherwise, the flow proceeds to step (1711). In the step (1711), it is determined whether or not “NA = 4”,
If “NA = 4”, go to step (1715),
If “NA ≠ 4”, it means that “NA = 5”, so the process proceeds to step (1712).

Step (1713) is an "image signal fetching 2" subroutine for fetching two image signals, and the next step (1714) is an "image signal fetching 3" subroutine for fetching three image signals. The following step (1715) is an "image signal acquisition 4" subroutine for acquiring four image signals, and step (1712) is an "image signal acquisition 5" subroutine for acquiring five image signals. is there.

Here, steps (1712) to (1712)
Since the operation of the subroutine of 15) will be described later, the description here is omitted.

When one of the subroutines of steps (1712) to (1715) is completed, step (171)
6) and (1717), where the focus detection aperture 65 and the focus detection light shielding plate 67 are driven, respectively.
(A).

In the following step (1718), the defocus amount of the photographing lens is calculated in the "defocus amount calculation" subroutine, and this subroutine is returned in step (1719). Here, a detailed description of the “defocus amount calculation” subroutine will be described later, and the description here will be omitted.

FIGS. 41 and 42 are flow charts of the "calculate the number NA of image signals" subroutine. When this subroutine is called in step (1706) of FIG. 40, the process goes through step (1801) to step (180).
Proceed to 2).

In the step (1802), it is determined whether or not the image signal capturing time interval TA is smaller than 5 msec.
If “TA <5 msec”, the flow proceeds to step (1803); otherwise, the flow proceeds to step (1810).
In step (1810), it is determined whether or not TA is smaller than 10 msec. If “TA <10 msec”, the flow advances to step (1811); otherwise, to step (181)
Go to 8). In step (1818), TA is 1
It is determined whether or not the time is shorter than 5 msec, and “TA <15 msec”
If so, the process proceeds to step (1819); otherwise, the process proceeds to step (1824). Step (182
In 4), it is determined whether TA is less than 20 msec,
If “TA <20 msec”, the flow proceeds to step (1825); otherwise, the flow proceeds to step (1830).

Here, the upper limit value of the accumulation time TC is set to 20 mse.
c, assuming that the drive time TD of the focus detection light shielding plate 67 is 3 msec, if “TA <5 msec”, step (1803)
Proceeding to "5msec≤TA <10msec", proceed to step (1811) and proceeding to "10msec≤TA <15msec"
If there is, the process proceeds to step (1819), and “15 msec ≦ TA
If <20 msec ”, proceed to step (1825),
If “20 msec ≦ TA ≦ 23 msec”, step (18)
30).

In the step (1803), it is determined whether or not the focal length FA of the photographing lens is smaller than 11 mm.
If A <11 mm, the flow proceeds to step (1807), 2 is input to the number NA of image signals, and the flow proceeds to step (1833). If not FA <11 mm, step (1804)
Proceed to. In step (1804), it is determined whether or not FA is smaller than 83 mm. If "FA <83 mm", 3 is input to NA in step (1808), and "FA <83m" is entered.
If not "m", the flow proceeds to step (1805). In step (1805), it is determined whether or not the FA is smaller than 207 mm.
Proceed to 9), enter 4 for NA, and if not "FA <207 mm", proceed to step (1806) and enter 5 for NA.

As described above, the steps (1803) to (1)
809), if the focal length FA is “FA <11 mm”, “NA = 2”; if “11 mm ≦ FA <83 mm”, “NA = 3”; if “83 mm ≦ FA <207 mm”, “NA = 4”, and if “FA ≦ 207 mm”, “NA = 5”.

In steps (1811) to (1817), similarly, if “FA <6 mm”, “NA = 2”
If “6 mm ≦ FA <23 mm”, “NA = 3”,
If “23 mm ≦ FA <77 mm”, “NA = 4”
If “FA ≦ 77 mm”, “NA = 5”.

Similarly, in steps (1819) to (1823), if “FA <11 mm”, “NA = 3”; if “11 mm ≦ FA <26 mm”, “NA = 4”; ≦ 26 mm, “NA = 5”.

Similarly, in steps (1825) to (1829), if “FA <6 mm”, “NA = 3” and “6”
If mm ≦ FA <12 mm, “NA = 4” and “FA
If ≦ 12 mm, “NA = 5”. Step (1
830) to (1832), “NA = 4” if “FA <10 mm”, and “NA = 4” if “FA ≦ 10 mm”.
= 5 ".

As described above, the image signal capture time interval T
When the number NA of image signals used for one focus detection is determined from A and the focal length FA of the photographing lens, step (183)
In 3), this subroutine is returned.

[0215] Figure 43 is a flow chart of the "image signal read 2" subroutine performs two image signals L ₁ and R ₂ uptake. When the "image signal fetch 2" subroutine is called in step (1713) in FIG. 40, the flow advances to step (1902) via step (1901).

[0216] At step (1903), the accumulation operation image signals L _1, the following step (1903) reads the image signal L _1. In step (1904),
The focus detection light shielding plate 67 is driven to the state shown in FIG. 7C, and in the subsequent step (1905), the image signal R ₂ is accumulated, and in step (1906), the image signal R ₂ is read out. At (1907), this subroutine is returned.

FIG. 44 is a flowchart of a subroutine "image signal take-in 3", in which three image signals L ₁ , R ₂ ,
Perform L ₃ uptake. Here, this is a case where the capture time interval of the image signal at the time of one focus detection is constant, that is, the drive time TD and the accumulation time TC of the focus detection light shielding plate 67 are the same. There is no need to measure time.

When the "image signal fetch 3" subroutine is called in step (1714) of FIG. 40, the flow advances to step (2002) via step (2001).

[0219] At step (2002), the accumulation operation image signals L _1, to read the next step (2003) in the image signal L _1. In step (2004), the focus detection light shielding plate 67 is driven to bring it into the state shown in FIG.
In steps (2005) and (2006), the image signal R ₂
Is stored and read out. In step (2007), and drives the focus detecting light shielding plate 67 again, and the state of FIG. 7 (b), step (2008), when the accumulation and reading of the image signals L ₃ in (2009), a step (201
At 0), this subroutine is returned.

FIG. 45 shows a subroutine "image signal take-in 4", in which four image signals L ₁ , R ₂ , L ₃ and R ₄ are taken in. Also in this case, the image signal capturing time interval during one focus detection is the same, and the time is not measured. When the "image signal fetch 4" subroutine is called in step (1715) in FIG. 40, the flow advances to step (2102) via step (2101).

[0221] At step (2102), (2103), the accumulation and reading of the image signal L _1. In step (2104), the focus detection light shielding plate 67 is driven to the state shown in FIG. 7C, and the steps (2105) and (210)
In 6), the accumulation and readout of the image signals R _2. In step (2107), the focus detection light shielding plate 67 is driven again to bring it into the state shown in FIG.
To accumulate and read an image signal L ₃ in (2109). In step (2110), the focus detection light shielding plate 67 is driven to the state shown in FIG.
Doing accumulation and reading of the image signals R ₄ at (2112), the process returns from this subroutine at step (2113).

FIG. 46 shows a subroutine "image signal fetch 5", in which five image signals L ₁ , R ₂ , L ₃ , R ₄ and L _{5 are} input.
Import Also in this case, since the time interval for capturing the image signal during one focus detection is constant, the time is not measured. When the "image signal capture 5" subroutine is called in step (1712) of FIG.
01), and proceeds to step (2202).

[0223] Step (2202), the (2203), the accumulation and reading of the image signals L _1, step (2204), and drives the focus detecting light shielding plate 67 in the state of FIG. 7 (c), the step ( 2205), (2206)
In performs accumulation and reading of the image signals R _2. In step (2207), the focus detection light shielding plate 67 is driven to the state shown in FIG. 7B, and the subsequent steps (2208) and (220)
In 9), the accumulation and reading of the image signals L _3. In step (2210), the focus detection light shielding plate 67 is driven again to the state shown in FIG.
To accumulate and read an image signal R ₄ at 212). In step (2213), the focus detection light shielding plate 67 is driven to the state shown in FIG.
Doing accumulation and reading of the image signal L ₅ at 15) in step (2216) to return this subroutine.

FIG. 47 shows a "defocus amount calculation" subroutine in which the calculation formula is switched according to the photographing conditions in order to obtain the necessary detection accuracy.

When the "defocus amount calculation" subroutine is called in step (1718) of FIG. 40, the flow advances to step (2302) via step (2301).

In step (2302), it is determined whether or not NA is 5, and if "NA = 5", step (230) is executed.
Proceed to 3) to perform arithmetic processing using the five image signals,
If “A ≠ 5”, the flow proceeds to step (2310). In step (2310), it is determined whether or not NA is 4;
If “NA = 4”, the process proceeds to step (2311) to perform arithmetic processing using four image signals, and “NA ≠ 4”
If so, the flow shifts to step (2315). In step (2315), it is determined whether the NA is 3 or not, and “NA =
If "3", the process proceeds to step (2316) to perform arithmetic processing using three image signals. If "NA「 3 ", step (2316) to perform arithmetic processing using two image signals. 2319).

In step (2303), the phase difference δ ₁₂ between the image signals L ₁ and R ₂ is calculated by correlation calculation, and in the following step (2304), the phase difference δ ₂₃ between the image signals L ₃ and R ₂ is calculated.
In step (2305), the image signals L ₃ and R
The phase difference δ _{23 of 4} is calculated, and in step (2306),
The phase difference δ ₄₅ between the image signals L ₅ and R ₄ is calculated.

In the following step (2307), it is determined whether or not the image signal fetching time interval TA is larger than 15 msec. If "TA> 15 msec", the process proceeds to step (230).
Go to 8), otherwise go to step (2309). This is because, as shown in FIG. 39, when the capturing time interval of the image signal is longer than 15 msec, the permissible focal length of the measure 4 is larger than that of the measure 3 when the capturing time interval TA of the image signal is greater than 15 msec. Proceeds to step (2308), performs the calculation of measure 4, and if TA is 15 msec or less, proceeds to step (2309),
The arithmetic processing of the measure 3 is performed to calculate the phase difference δ used when calculating the defocus amount.

In step (2311), the phase difference δ ₁₂ between the image signals L ₁ and R ₂ is calculated, and in step (2312), the phase difference δ ₂₃ between the image signals L ₃ and R ₂ is calculated, and in step (2313) In), the phase difference δ ₃₄ between the image signals L ₃ and R ₄ is calculated. In the following step (2314), the phase difference δ used for the defocus amount calculation processing is calculated by the calculation processing of measure 2.

[0230] At step (2316), calculates the phase difference [delta] ₁₂ of the image signals L ₁ and R _2, in step (2317), calculates the phase difference [delta] ₂₃ of the image signals L ₃ and R _2. In the following step (2318), the phase difference δ used for calculating the defocus amount is calculated by the calculation processing of measure 1.

In step (2319), the phase difference δ ₁₂ between the image signals L ₁ and R ₂ is calculated, and in step (2320), δ ₁₂ is input as the phase difference δ used for calculating the defocus amount.

Steps (2308), (2309),
When the phase difference δ is calculated by any one of (2314), (2318), and (2320), step (232)
Proceeding to 1), in step (2321), when the defocus amount DF is calculated from the phase difference δ, the sensitivity K of the focus detection system, and the pixel pitch P of the CCD, in step (2322),
Return this subroutine.

The above description is based on the case where the minimum number of image signals and the calculation method capable of obtaining the necessary detection accuracy are selected in accordance with the photographing conditions (focal length of the photographing lens, image signal capturing time interval). This is an embodiment, which eliminates the need to perform an unnecessary image signal capturing operation, so that the time required for focus detection can be shortened and the power consumption can be reduced.

According to the above-described embodiment, in the focus detection device for detecting the defocus amount of the photographing lens from the phase difference between the image signals taken at different times, the correlation value indicating the degree of coincidence of the two images is calculated by the correlation operation. The range in which the amount of vertical movement of the camera is detected based on the correlation value, the amount of vertical movement is corrected, and the correlation value is calculated from the focal length of the photographing lens and the time interval for capturing the image signal. Is set, the correlation calculation is not performed for an unnecessarily wide range, and the time required for the calculation can be efficiently reduced, and the time required for focus detection can be reduced.

(Correspondence between the Invention and the Embodiment) In each of the above embodiments, the CCD 64 (70) is provided on the solid-state image pickup device of the present invention by using the focus detection diaphragm 65, the focus detection light shielding plate 67,
The motors 66 and 68 serve as the pupil shape changing means of the present invention, and the portion of the digital signal processing section 72 for executing the operation described with reference to FIGS. 49 corresponds to the range setting means for setting the movement amount detection range of the present invention.

The step (203) in FIG.
Step 4 (406), Step (608) in FIG.
To (610), step (807) in FIG. 29, step (1005) in FIG. 31, step (1210) in FIG.
To (1212), steps (1408) to (1) in FIG.
410), step (1606) in FIG. 38, steps (2308), (2309), (2314),
The part that executes the arithmetic processing of (2318) and (2320) corresponds to the image signal position changing means of the present invention.

The correspondence between the components of the embodiment and the components of the present invention has been described above. However, the present invention is not limited to the configuration of the embodiment, and the functions and features described in the claims can be obtained.
Needless to say, any configuration may be used as long as the functions of the embodiment can be achieved.

(Modification) In each of the above embodiments, the CCD is used as the solid-state imaging device.
The format of the CD may be not only an area sensor but also a line sensor. Also, as pupil shape changing means,
Although the focus detection aperture 65 is used, a physical aperture (LCD)
).

Although the focus detecting light-shielding plate 67 is moved into and out of the photographing optical path by a motor, it can be moved by using an actuator such as a plunger.

Although the camera is assumed to have a configuration having two pupil regions, three or more pupil regions may be provided. The pupil area is arranged in the horizontal direction when the camera is held (the CCD line direction is also coincident). However, the present invention is not limited to this. May be in the vertical direction when the camera is turned on.

The moving amount detection range for detecting the moving amount of the subject in the vertical direction (the direction perpendicular to the direction in which the pupil regions are arranged) is, as described with reference to FIG. Image signal capture time interval, image signal accumulation time,
Further, the drive time of the focus detection light shielding plate 67 (the drive time of the pupil shape changing means) is determined in consideration of each. However, a certain effect can be obtained by considering at least one of the information. It is possible. Especially,
As in the present embodiment, when the focus detection light-shielding plate 67 enters and retracts into the photographing optical path using a motor, the driving time is almost constant, and therefore, this information is almost unnecessary even if it is not used. No problem.

Further, in order to set the number of image signals used for detecting the amount of movement of the subject in the vertical direction (the direction perpendicular to the direction in which the pupil regions are arranged), the description has been given with reference to FIGS. 41 and 42. In this way, the focal length is determined in consideration of the focal length of the photographing lens, the image signal capture time interval, the image signal accumulation time, and the drive time of the focus detection light shielding plate 67 (the drive time of the pupil shape changing unit). However, it is possible to obtain a certain effect by considering at least one of the pieces of information.

Although the present invention has been described with respect to an example in which the present invention is applied to a digital camera, the present invention is also applicable to a video camera.

[0244]

As described above, according to the present invention,
The maximum movement amount of the image signal is predicted from the focal length of the optical system and the time interval for capturing the image signal, and further, the accumulation time of the optical image and the drive time of the pupil shape changing means. The necessary range is limited except for the range, and a correlation value indicating the degree of coincidence of the two images obtained in time series within this range is calculated by a correlation operation, and from this correlation value, the degree of coincidence of the two images is high. Since the position of the image signal is searched and the phase difference is calculated using the image signal at this position, the amount of calculation and the calculation time for calculating the relative displacement between the subject and the camera are minimized. In addition, it is possible to provide a focus detection device capable of performing optimum focus detection in consideration of the calculated relative displacement between the subject and the camera.

[Brief description of the drawings]

FIG. 1 is a diagram showing a light beam passing through each pupil region and an image forming state on a CCD surface in a focused state in a camera according to each embodiment of the present invention.

FIG. 2 is a diagram showing a light beam passing through each pupil region and an image forming state on a CCD surface in a front focus state in the camera according to each embodiment of the present invention.

FIG. 3 is a diagram showing a light beam passing through each pupil region and an image forming state on a CCD surface in a back focus state in the camera according to each embodiment of the present invention.

FIG. 4 is a perspective view showing a portion for changing a pupil region provided in the camera according to each embodiment of the present invention.

FIG. 5 is a central sectional view of FIG. 4;

FIG. 6 is a block diagram showing a main part of an electric configuration of the camera according to each embodiment of the present invention.

FIG. 7 is a diagram for explaining a time when a pupil region is changed in the camera according to each embodiment of the present invention.

FIG. 8 is a diagram for explaining a method of driving the CCD of FIG. 6;

FIG. 9 is a view for explaining a method of driving the CCD of FIG. 6;

FIG. 10 is a diagram for explaining a method of driving the CCD of FIG. 6;

FIG. 11 is a diagram illustrating a case where a phase difference is detected by a correlation operation in the camera according to each embodiment of the present invention.

FIG. 12 is a diagram illustrating a case where a phase difference is detected by a correlation operation in the camera according to each embodiment of the present invention.

FIG. 13 is a diagram for explaining a case where the camera shakes up and down during shooting.

FIG. 14 is a diagram for describing a change in a correlation amount in the camera according to each embodiment of the present invention.

FIG. 15 is a diagram for explaining a change in a correlation amount in the camera according to each embodiment of the present invention.

FIG. 16 is a diagram for explaining a method of interpolation calculation in the camera according to each embodiment of the present invention.

FIG. 17 is a flowchart illustrating shift range calculation in the camera according to each embodiment of the present invention.

FIG. 18 is a diagram for describing a camera shake detection method according to the first embodiment of the present invention.

FIG. 19 is a diagram for explaining a camera shake detection method according to the first embodiment of the present invention.

FIG. 20 is a flowchart showing a series of operations of the camera according to the first embodiment of the present invention.

FIG. 21 is a flowchart illustrating an example of an operation of the camera according to the first embodiment of the present invention at the time of focus detection.

FIG. 22 is a flowchart illustrating an example of an operation when calculating a defocus amount of the camera according to the first embodiment of the present invention.

FIG. 23 is a flowchart showing another example of the operation at the time of focus detection of the camera according to the first embodiment of the present invention.

FIG. 24 is a flowchart illustrating another example of the operation when calculating the defocus amount of the camera according to the first embodiment of the present invention.

FIG. 25 is a diagram for explaining a camera shake detection method according to the second embodiment of the present invention.

FIG. 26 is a flowchart illustrating an example of an operation of the camera according to the second embodiment of the present invention when detecting a focus;

FIG. 27 is a flowchart illustrating an example of an operation of the camera according to the second embodiment of the present invention when calculating a defocus amount.

FIG. 28 is a flowchart showing another example of the operation at the time of focus detection of the camera according to the second embodiment of the present invention.

FIG. 29 is a flowchart showing another example of the operation when calculating the defocus amount of the camera according to the second embodiment of the present invention.

FIG. 30 is a flowchart showing another example of the operation at the time of focus detection of the camera according to the second embodiment of the present invention.

FIG. 31 is a flowchart showing another example of the operation when calculating the defocus amount of the camera according to the second embodiment of the present invention.

FIG. 32 is a diagram for describing a camera shake detection method according to a third embodiment of the present invention.

FIG. 33 is a flowchart showing an example of the operation of the camera according to the third embodiment of the present invention at the time of focus detection.

FIG. 34 is a flowchart showing an example of the operation of the camera according to the third embodiment of the present invention when calculating the defocus amount.

FIG. 35 is a flowchart showing another example of the operation of the camera according to the third embodiment of the present invention at the time of focus detection.

FIG. 36 is a flowchart showing another example of the operation when calculating the defocus amount of the camera according to the third embodiment of the present invention.

FIG. 37 is a flowchart showing another example of the operation at the time of focus detection of the camera according to the third embodiment of the present invention.

FIG. 38 is a flowchart showing another example of the operation when calculating the defocus amount of the camera according to the third embodiment of the present invention.

FIG. 39 is a diagram illustrating a relationship between an image signal capturing time interval and an arithmetic processing method in the camera according to the fourth embodiment of the present invention.

FIG. 40 is a flowchart showing an operation of the camera according to the fourth embodiment at the time of focus detection.

FIG. 41 is a flowchart showing a partial operation when calculating the number of image signals of the camera according to the fourth embodiment of the present invention.

FIG. 42 is a flowchart showing a continuation of the operation in FIG. 41.

FIG. 43 is a flowchart showing details of the operation of “image signal capture 2” of FIG. 40;

FIG. 44 is a flowchart showing details of the operation of “image signal capturing 3” in FIG. 40.

FIG. 45 is a flowchart showing details of the operation of “image signal capturing 4” in FIG. 40;

FIG. 46 is a flowchart showing details of the operation of “image signal capturing 5” in FIG. 40;

FIG. 47 is a flowchart showing an operation of a camera according to a fourth embodiment of the present invention when calculating a defocus amount.

FIG. 48 is an arrangement diagram of an optical system of a general single-lens reflex camera.

49 is a diagram for explaining the principle of focus detection of the single-lens reflex camera having the configuration of FIG. 48.

[Explanation of symbols]

 64, 70 CCD that is a solid-state imaging device 65 Focus detection aperture 67 Focus detection light shield 66, 67 Motor 72 Digital signal processing unit

Claims (4)

[Claims] 1. An imaging means for photographing, an optical system for photography, and a photographing light beam passing through the optical system are divided into at least two different regions in a time-series manner and are formed on the imaging means. Pupil shape changing means, and converting an optical image formed in time series on the imaging means through the respective areas into an image signal, and detecting a phase difference of the image signal to thereby control the optical system. A focus detection device comprising: a calculation unit that calculates a focus state; a movement amount detection unit that detects a movement amount of an optical image on the imaging unit in a direction orthogonal to a direction in which the regions are arranged; Range setting means for setting a range in which the movement amount is detected by the detection means; and an image signal for changing a position of an image signal used for phase difference detection performed by the calculation means in accordance with an output of the movement amount detection means Position changing means is provided. Focus detection device. 2. The focus detection apparatus according to claim 1, wherein said range setting means changes a range in which a movement amount is detected according to a photographing condition. 3. The method according to claim 1, wherein the range setting unit is at least one of a focal length of the optical system, an accumulation time of the optical image by the imaging unit, a time interval of capturing the image signal, and a driving time of the pupil shape changing unit. 2. The focus detection device according to claim 1, wherein a range in which the movement amount is detected is set based on the two pieces of information. 4. The focus detection apparatus according to claim 1, wherein the movement amount detection means detects the movement amount based on a correlation value indicating a degree of coincidence of each optical image obtained by a correlation operation.