JPH03192227A - Blur detection camera - Google Patents

Abstract

PURPOSE:To eliminate parallax and to detect the blur of a subject by sending light from the subject to an area sensor as a blur detection sensor by using a pericle. CONSTITUTION:Part of luminous flux transmitted through a pericle mirror 34 is guided to a phase difference AF detection module 36 by an AF detection mirror 7 and luminous flux reflected by the pericle mirror 34, on the other hand, is entered into a pentagonal prism 38 through a condenser lens 37. One surface 38a of the pentagonal prism 38 is a half-mirror and guides part of the incident luminous flux out the pentagonal prism 38 and this luminous flux is passed through a blur detection optical system 40 and reflected by a reflecting mirror 43 to reach the blur detection sensor 44. Thus, the light from the subject is sent to the area sensor 44 as the blur detection sensor by using the pericle mirror 34 to enable a TTL type camera to perform blur detection by an image detection system.

Description

[Detailed Description of the Invention] [Industrial Application Field 1] This invention relates to a camera shake detection camera, and particularly to a TTL camera.
This invention relates to a shake detection camera using a method.

[Prior art] A conventional camera with an electronic viewfinder is, for example,
It is disclosed in Japanese Patent No. 2-35329.

However, conventional cameras with electronic viewfinders do not have shake detection devices.

-, 6" A camera with an electronic viewfinder capable of detecting shake is disclosed in, for example, Japanese Patent Application Laid-Open No. 129328/1983. According to the publication, a camera with an electronic viewfinder capable of detecting shake has an area CCD sensor for shake detection. However, conventional shake detection cameras with electronic viewfinders are external light type and not TTL type.

On the other hand, a video camera that performs blur correction using the TTL method is disclosed in, for example, Japanese Patent Laid-Open No. 61-240778. According to the publication, a video camera uses an angular velocity sensor to detect image blur and corrects the blur by moving the lens.

[Problems to be solved by the invention] Conventional cameras with electronic viewfinders and shake detection devices,
The TTL video camera was configured as described above. Therefore, since still cameras do not use the TTL system, parallax occurs. As a result, it is difficult to correct image blur. Furthermore, since TTL video cameras are not still cameras, there is no need to consider the optical path.

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a blur detection camera that is free from parallax and can also detect blur in an object.

[Means for Solving the Problems] The blur detection camera according to the present invention includes a movable blur correction optical system provided in the photographic lens, a pellicle provided in the optical path from the subject to the film, and a pellicle provided in the optical path from the subject to the film. An area sensor that captures an image, a two-dimensional display device that displays the output of the area sensor, and two temporally different areas of the area sensor.
The camera shake detection means includes a shake detection means for detecting camera shake by calculating two outputs, and an extraction means for correcting the camera shake by moving a shake correction optical system when the shake detection means determines that there is a shake.

[Operation] In this invention, since the pellicle is used to transmit the light of the subject to the area sensor that detects blur, T
Image blur is detected using the TL method.

Further, an image corresponding to the output of the area sensor is output to an electronic viewfinder, which is a two-dimensional display device.

[Examples] Examples of the present invention will be described below with reference to the drawings. 1st
The figure is a block diagram showing the main parts of a shake detection camera with an electronic viewfinder according to the present invention.

Referring to Figure 1, 5PD for photometry (Silicon
The signal from I'hotodiode) 2 is connected to the lung 1 circuit 3.
1 is human power. 11?3 light times I#ls31 +t, 5
It calculates the output of pD2 and outputs the calculation result to the main CPUI. The photographic lens 3 is interchangeable (I+J), and the aperture is driven and the focus is adjusted through an aperture drive means 4 and a focus adjustment drive means 5. The photographic lens circuit 6 stores the F value, focal length, and various parameters for blur correction of the lens that is unique to the photographic lens 3. The photographic lens circuit 6 further includes an actuator for driving an optical system for blur correction and a control circuit therefor.

The AF detection mirror 7 is a phase difference type AF detection module 14
A part of the luminous flux passing through the photographing lens 3 is guided to. The AF detection mirror 7 is retracted by the AF detection mirror driving means 8 so as not to block light onto the film during film exposure. The shutter screen 9 of the focal brain shutter includes a front wheel and a rear peak, and is driven by a shutter screen driving means 10. Switch S1 is ONL, fl? at the first stroke of the shutter button. 1 Perform light and AF fishing operations. The switch S2 is turned on by the second stroke of the shutter button. The film winding means 15 winds and rewinds the film. Switches St' and S2' are shutter buttons installed on a remote finder, which will be described later, and have the same function as switches S1 and S2, respectively. Switch SC is a switch for operating continuous shooting mode.

The shake detection optical system 40 and the electronic finder imaging optical system 41 are driven by a shake detection optical system driving means 42,
This is an optical system for guiding light to a CCD for shake detection, which will be described later. When performing shake detection, a shake detection optical system 40 is used, and when displaying an image using the above-mentioned CCD, an electronic finder imaging optical system 41 is used. The main CPU is further connected to a camera shake correction controller 18 for detecting camera shake, which will be described later. The camera shake correction controller 18 communicates with the camera shake detection and correction means 11.

The main CPU further includes an exposure correction amount manual means 2 for performing exposure correction for the exposure amount determined by the photometric value.
2 is connected to film sensitivity setting means 23 for setting film sensitivity.

Next, the electronic finder will be explained. Figure 2A ~ 2nd
FIG. D is a diagram showing how the electronic finder is used. Second
Referring to FIG. A, a removable and retractable electronic viewfinder 100 is provided on the top surface of the camera body 12. The electronic viewfinder 100 has L for viewfinder image display.
It includes a CD 63 and a shutter button 101. When electronic finder 100 is used as shown in FIG. 2A, it is used as a waist level finder. 12Bll is a diagram showing a state in which the electronic finder 100 is pulled forward. The viewfinder image can be clearly seen without looking into the viewfinder eyepiece. Therefore, you can take pictures while checking the viewfinder image from the height of the camera held above your head. FIG. 2C shows a state in which the electronic finder 100 is pulled forward. You can take self-timer shots while checking the viewfinder image.

In this case, the image on the finder image display 311 L CD 63 is upside down compared to the normal case. FIG. 2D is a diagram showing a state in which the electronic viewfinder 100 is removed from the camera body 12. Using the electronic finder extension cord 102, remote photography can be performed by pressing the release button 101 while observing the finder image on the LCD 63.

The image signal sent to the LCD 63 is imaged by the CCD imaging section 51 of the shake detection sensor 44, which will be explained later with reference to FIG. The exposure time of the CCD at this time is determined based on the exposure of the camera. Therefore, the LCD 63 displays an image that has a density similar to that of the finished photograph. If you want to use exposure compensation i when shooting with auto exposure, the LCD
The two-person exposure correction means 22 may be adjusted so that the part for which the exposure should be adjusted can be observed with good gradation while viewing the image 63. Determining the exposure when shooting with manual exposure is also possible with L.
The aperture and shutter speed may be adjusted so that the portion to which the exposure is mainly desired can be observed with good gradation while viewing the CD 63. Furthermore, even if you want to intentionally overexpose or amplify the exposure, if you decide on the exposure while observing the LCD 63, you will be able to obtain a photograph with the desired exposure.

Fig. ff13A is a diagram showing an optical system of a shake detection camera with an electronic finder according to the present invention. Referring to FIG. 3A, the photographing lens 3 includes a shake correction lens 32 and a shake correction lens drive device i for moving the shake correction lens 32.
33. Note that the details of the correction lens driving bag W133 are well known, so a description thereof will be omitted. A mirror 34 is fixed inside the camera body 12, and part of the light beam passing through the photographic lens 3 is reflected toward the finder optical system, and the rest is transmitted toward the shutter curtain 35. A part of the light beam transmitted through the pellicle mirror 34 is
The IQ output mirror 7 directs the phase m AF detection module 36 . During film exposure, the Al4ft output mirror 7 is retracted by an AFI output mirror drive means 8 (not shown) to a position where it does not block the light flux reaching the film during exposure.

The light beam reflected by the pellicle mirror 34 passes through the condenser lens 37 and enters the pentaprism 38. An 11$1 light 5PD2 is provided on the three eyepiece lenses side of the pentaprism '38. One member 38a of the pentaprism 38 is a half mirror, and takes out a part of the incident light flux to the outside of the pentaprism 38. The extracted light beam passes through the blur detection optical system 40, is reflected by the reflecting mirror 43, and reaches the blur detection sensor 44. The shake detection optical system 40 and the electronic finder imaging optical system 41 are driven by a shake detection optical system driving means 42, and are selectively moved into and out of the optical path. The shake detection optical system 40 and the electronic finder imaging optical system 41 each re-image the image on the finder focal point 1Ij45 onto the shake detection sensor 44. The shake detection sensor 44 is a CCD area sensor. Note that the pentaprism 38 is not limited to a glass block type, but may be a hollow pentamirror with an air layer inside.

As described above, in this invention, the pellicle mirror 34
Since the light of the limb photograph is transmitted to the area sensor 44 which serves as a shake detection sensor, blur detection is performed using the image detection method in a TTL camera. Furthermore, a pentaprism 3 is included in the optical path up to the shake detection sensor 44.
8 is used, there is no need to provide a separate optical path to the shake detection sensor 44.

Figure m38 is a diagram showing details around the blur detection optical system driving means 42. Referring to FIG. 3B, a lens holder 200% 201, in which a lens 40a of a blur detection optical system 40 and a lens 41a of an electronic viewfinder imaging optical system 41 are fixed, is fixed to a rotating shaft 205, and a motor 21 is connected to the rotating shaft 205 via a gear.
Connected to 0. The shaft of the motor 210 is interlocked with the disc-shaped chopper 202 via a gear, and the encoder spring 20
The rotation of No. 2 is read by the photocoupler 203. As the encoder spring 202 rotates, the photocoupler 20
Every time the light between the two arms of No. 3 is interrupted, an on-off TD occurs. The controller 204 counts this pulse signal. As a result, the amount of rotation of the rotating shaft is monitored.

According to the monitored signal, by turning on or off the power to the motor 210, the shake detection optical system 40.41
The respective lenses 40 a s 41 a are rotated. In this manner, the blur detection optical system driving means 42 switches the optical path between the blur detection optical system 40 and the electronic finder imaging optical system 41.

As described above, in the present invention, the optical system for projecting the subject image onto the blur detection sensor 44 is switched by the blur detection optical system driving means 42.

As a result, the pixel area used for blur detection and the AP area selected for focusing are switched.

FIGS. 4A to 4C are diagrams showing finder screens. 4A and 4B is a transparent blur detection area, and because of this area, an optical path to the sensor 44 is secured and the amount of light to the sensor 44 is increased. The area indicated by C is the matte portion of the finder screen, on which light is scattered, an image is focused on it, and the focus is detected in the conventional manner.

The area indicated by a is also used for displaying the finder display frame of the shake detection area. b in FIG. 4A is the field of view range of the electronic finder, and this part is also transparent, so that the amount of light to the imaging optical system 41 for the electronic finder increases. FIG. 4C is a diagram showing a case where the entire surface of the finder screen is transparent.

A portion a near the center of the finder screen C is imaged on the imaging surface by the blur detection optical system 40.

This image is used by the shake detection sensor 44.

Furthermore, when the electronic finder imaging optical system 41 is used,
Almost the entire area of the finder screen C is imaged on the imaging surface, and the image is used as an image for the electronic finder.

Note that, referring to FIG. 3A, the photographing lens 3 is attached to the camera body 12 via a lens mount portion 46, and is replaceable.

Note that if a normal lens without a camera shake correction lens is attached to the camera body 12, a warning will be displayed on the finder screen when shake is detected. A warning about camera shake may be displayed on the electronic viewfinder with the words "Be careful of camera shake".

FIG. 5 shows another embodiment of the optical system according to the invention. In this case, the pentaprism 38 is omitted compared to the optical system shown in FIG. 3A.

Since there is no pentaprism 38, finder screen, or condenser lens 37, the cost can be reduced accordingly, and it is also lighter and more compact. Shake detection sensor 44'
A sensor imaging lens 45' is provided in front of the sensor. The small-area relay lens 40' and the large-area relay lens 41' are used in a switched manner, thereby switching the light flux guided to the area sensor 44'. The optical system according to this embodiment is provided with a reflecting mirror 43' and an optical system drive actuator 42'.

6A and 6B are diagrams showing how the finder screen is used when the optical system according to another embodiment of the invention shown in FIG. 5 is used. These figures correspond to FIGS. 2A to 2D according to the first embodiment. Figure 6A, 6th
Referring to Figure B, the electronic finder 63 can be removed.

Next, the contents of the camera shake detection and correction means 11 will be explained with reference to FIG. The camera shake detection sensor 44 includes a CCD imaging section 51, an output amplifier 52 for amplifying the CCD output, an illuminance monitor 53 for controlling the integration time of the CCD, and a photometry circuit 5.
4. The clock generator 55 is connected to the photometry circuit 5
4 is detected and the integration time of the CCD and the gain of the output amplifier are set. In addition, the clock generator 55
are the CCD driving clock, the A/D converter 61, and the D/D converter 61.
A clock is also generated for the A converter 58, the five sensitivity variation compensation iE memories, and the dark output compensation memory 60. The output of the shake detection sensor 44 is input to the A/D converter 61 through a differential amplifier 56 and a gain control amplifier 57. The dark output correction memory 60 and the sensitivity variation correction memory 59 respectively store data for correcting CCD sensitivity variations and dark output. Dark output correction memory 60 outputs data to D/A converter 58. The output signal of the D/A conversion is input to the differential input of the differential amplifier 56. As a result, the dark output of the CCD 51 is corrected. The gain control amplifier 57 is an amplifier whose amplification level is controlled by a digital signal. The gain control amplifier 57 is controlled by the sensitivity variation correction data stored in the sensitivity variation correction memory 59, and
Compensate for sensitivity variations in D output.

The output (a''j) of the A/D converter 61 is stored in the image part memory 64, the reference part memory 65, or the reference part memory 66. generates address data necessary for each memory operation.

The arithmetic unit 68 includes a subtraction circuit 69, an absolute value circuit 7°, an addition circuit 71, and a register 72. Data in the reference part memory 65 and the reference part memory 66 are given as input.

The calculation result from the calculation unit 68 is +11 depending on the type of calculation.
The result is stored in either the flash result memory 73, the vertical contrast memory 74, or the horizontal contrast memory 75. These memories are connected to the control cPU 76 and can be accessed from the control CPU 76. Control CPU 76 also controls address generator 67 and clock generator 55. The data in the image memory 64 is D/A converted by a D/A converter 77 and input to the video signal processing circuit 62 . The switch SRV is a switch that is turned on when the LCD 63, which is an electronic viewfinder, is pulled forward.

When this switch is turned on, the video signal processing circuit 62 displays the video on the LCD 63 with the video upside down.

Next, the h' method for detecting blur and calculation of the amount of blur will be described.

First, the shake detection sequence will be explained. In this invention, a subject image is detected in a two-dimensional manner using an area sensor 44 that can detect two-dimensional image data. Image blur is detected by detecting the shift of the subject image over time using the area sensor 44.

CCD51 is I x J u! +i elemental area sensor.

The reference part memory 65 and the reference part memory 66 are each
The correlation result memory 73 is a memory of xJ words.
This memory has a capacity of 1-f H words. CCD51
Consider dividing the light-receiving surface into MXN blocks. Each block is composed of adjacent KXL pixels. The vertical contrast memory 74 and the horizontal contrast memory 75 are each ~I
The memory has a capacity of XN words. The procedure for detecting blur will be explained below, but here, l-68, J-52,
K-8, L-8, M-8, N-6, H-5.

It is assumed that the resolution of the A/D converter 61 is 8 bits, the register 72 is 14 bits, and one word of the correlation result memory 73, vertical contrast memory 74, and horizontal contrast memory 75 is each represented by 14 bits.

FIG. 8 is a schematic diagram of a portion of the light receiving section of the CCD 51. A grid of small square IMs represents a unit pixel. In the figure, the pixel on the left side is assumed to be pixel (1, 1), and the pixel on the upper right side is assumed to be pixel (68, 52). The light receiving section of the CCD 51 is divided into blocks each consisting of 8 pixels x 8 pixels, except for two pixels on the outer periphery of the light receiving section. The parts surrounded by thick lines in the figure correspond to each block. Block the lower left block (L 1)
, the upper right block is block (8, 6).

The shake detection sequence is large. (1) Contrast calculation and block selection (2) Correlation: 1 arithmetic (3) Interpolation = I calculation It is divided into three parts.

Contrast calculation is a calculation performed for block selection. Among the various parts of the object imaged on the light receiving section of the CCD 51, some parts are suitable for blur detection, while others are not. In this embodiment, the contrast of the object is calculated in order to select a portion suitable for blur detection. The contrast of the object is adjusted for each block, and blur is detected using a total of 8i1AI blocks, 4 blocks with large contrast in the vertical direction j and 4 blocks with large contrast in the horizontal direction j.

(1) Contrast: Explanation of 1 calculation and block selection First, the output of the CCD 51 is stored in both the base memory 65 and the reference memory 66. Using this data, the vertical contrast and horizontal contrast of each block are calculated by 1. The calculation is in block (1, 1), (
2,1)%... (7°6), horizontal contrast of (8,6), blocks (1,1), (2,1),...
・Perform in the order of vertical contrast of (7, 6) and (8°6).

The output of pixel (i, j) of CCD 51 is A (1°j)
The horizontal contrast of block (k, 11.) is defined as the vertical contrast.

The procedure for executing this =11 calculation using the hardware shown in FIG. 7 will now be described. The contents of the reference memory 65 corresponding to the output of pixel (i+J) of the CCD 51 are set to R(1+J”).
), the contents of the reference section memory 66 are sci.

j). First, register 72 is cleared. Next, R(i, j) is applied to one side of the subtractor 69, and S N+1 is applied to the other side. j) (However, 1-8(k-1)+2, j
An address is sent from the address generator 67 so that the address is given as -8 (assumed to be -1)+2). Decreasing circuit 69
The absolute value of the subtracted data is taken by the absolute value circuit 70, added to the contents of the register 72 by the adder circuit 71, and stored in the register 72. Next, the address that should be i-N+1 is sent from the address generator 67, and a similar calculation is performed. In this way, 1-8(k-])+2~8
+2, j-8 (1-1) +3~81. When processing is performed within the range of +2, the following is stored in the register 72.

The same content A(i, j) is stored in the reference part memory 65 and the reference part memory 66, and A(i, j)
=R(i,j)-S(i,j), so the contents of register 72 are horizontal contrasts. When one block has been calculated, register 72
The contents of the block are transferred to and stored in the affected contrast memory 75 corresponding to that block.

Similarly, the horizontal contrasts of all remaining blocks are calculated and stored in the horizontal contrast memory 75.

Vertical contrast is then calculated. In the same way as when the horizontal contrast is calculated by tl, register 7 is first calculated.
2 is cleared. Next, R (1, J) is input to the input of one node j of the subtraction circuit 69, and S (i, N+1) I is input to the input of the other h.
However, i −8(k−1) +3, j”8 (1−1)+
An address is sent from the address generator 67 so that the following address is given.

The data subtracted by the six subtracting circuits has its absolute value taken by an absolute value circuit 70, is added to the contents of a register 72 by an adding circuit 71, and is stored in the register 72. Then i-N+1
The address to be obtained is sent from the address generator 67, and a similar calculation is performed. In this way, i −8
(k-1) +2 to +3~8, j-8 (sad -1) +2~
Processing is performed in the range of 8 (+2). As a result, the register 72 contains the following information:
j+l) Jsg7-1) λ-tC ring-
1 to 1 is JC! It is remembered. Here, A mouth, j)-R
mouth.

j) -5ci, j), so the contents of register 72 are vertical contrasts. 1 block worth of a! 6, the contents of the register 72 are transferred to and stored in the vertical contrast memory 74 corresponding to that block.

Similarly, the vertical contrast of all remaining blocks is calculated by 1 and stored in the vertical contrast memory 74.

The control CPU selects a block with the highest contrast from among the contrasts in the vertical and horizontal directions of each block that have been counted by 1 in the above manner. Next, from among the remaining blocks, the block with the highest contrast in a direction different from the previously selected contrast direction (for example, in the horizontal direction if the first selected direction was the vertical direction) is selected. In the same manner as above, eight blocks (vertical, gantry 4 @ each) are selected in the order of the highest contrast while changing the direction of contrast. The selected blocks are 81 to B8.

(2) Description of Correlation Calculation As for the contents of the reference section memory 65, the data on which the contrast has been calculated is left as is as reference image data. New data is successively read from the CCD 51, and each time new data is written into the reference section memory 66 as a reference image. Every time new data is written to the reference section memory 66, the reference image and the reference image are compared, and a spatial shift between the two is detected as image blur. Image blur is determined by correlation=1 calculation and interpolation calculation, which will be explained below.

The block Bk (k-1, 2, . . . 8) correlation value is defined by the following equation.

(1, m--2,-1°0.1.2) (ji, Jh represent the youngest image index in block Bk) Considering the case of Chu-mw-Q, it becomes. This value is the sum of the absolute value of the difference between the data of the same pixel in the standard image and the reference image for one block. Q
Considering cases other than −m−0, Ck (1, m
) is the data of pixel ci, j) of the standard image and 1 of the reference image
The absolute value of the difference from the data of the iIIIi element (i+(. Q, m) Ningsl (楚wa m--2,-1,0,],2).

One of the subtraction circuits 69 is connected to the reference part memory 65, and the other part of the subtraction circuit 69 is connected to the reference part memory 66.

Therefore, after clearing the register 72, the subtraction circuit 69
The human power of one side is R(i, j), and the human power of the other side is S(l Jumi,
j+m) is input manually, and an address determined by the control CPU is sent out from the address generator 67 so that data in a predetermined range of xsj is processed. Then, C←l, yn) is obtained in the register 72. This is transferred to a predetermined address in the correlation result memory 73 and stored therein. By repeating this process, C(L m): (1!, mm-2)
, -1, 0, 1, 2> are all found.

(3) Explanation of interpolation calculation When the calculated correlation values C(Q!, m) are arranged with the gong on the horizontal axis and m on the vertical axis, it becomes an array like a number.

C(-2,2) C(-LR) C(0,2)
C(1,2) C(2,2)C(-2,1) C(
-1,1) C(0,1) C(1,1) C(
2,1) C(-2,0) C(-i,0) C(0
,0) C(1,0) C(2,0)C(-2,-
1) C(-1,-1) C(0,-1) C(
+, -1) C (2, -1) C (-2, -2) C
(-1,-2) C(0,-2) C(+,-2)
If there is no deviation at all between the C(2,-2) reference image and the dillimu 1ihi image, C(0,0) will be 0, and will have a larger value as it goes to the outside of the array. Also, the reference image! ! For quasi-both images, 0 pixels on the right and mo@ on the top
If there is an elementary deviation, C(Q◎*m(1) becomes 0, and the further away from this value, the larger the value becomes. However, the image deviation is not necessarily an integer multiple of the pixel.

Furthermore, since the speed of image blur is not constant, the blur of the image due to image blur may differ between the standard image and the reference image. If the distribution of the correlation mc ((1, m) in this case is represented by equal lines, it will look like the one shown in Figure 9.Referring to Figure 9, the actual values obtained are shown in the figure. Although it is only on the grid points, isolines are drawn assuming the values between the grids.The center of the contour line is the smallest value.The center of the isoline is at the point m(
xo.

yo), and the reference image is considered to be shifted horizontally by XQ and vertically by yo with respect to the reference image. Correlation = 1
Since only the chains on the lattice points are obtained by ° calculation,
The MP points must be determined by interpolation using grid point data. The magnification of the blur detection optical system 40, the pixel size of the CCD 51, and the integration time are set so that the MP point becomes -7.5<x(1%yo<1.5), assuming the actual size of the blur. Ru.

Correlation: Interpolation after completion of 1 arithmetic = 1 arithmetic will be specifically explained. The :1 calculation of the lateral force direction will be explained. First, C (Lm)
Find the minimum value C(1,(+,mo)).

io values and C(uo−+, mo) and C(IL
(a) to (h) in Figure 10 due to the size relationship with o+m(1)
It is divided into the following cases.

(a) In the case of conditions It is determined that ()≦xg<1.

Point (-1, C(-1, rno) ) and point (0, C(0,
The straight line (i) connecting the points (1, C (1, 110
)') and the point (2,C(2,mo)).
Let the coordinates of the intersection with (ii) be xO.

(b) In the case of conditions 1<x(judges 1<O.

The points (-2,C(-2,mo)) and (-1,C(-
1,mo)) and the point (0°C(0,m
o )) and the straight line (i i) connecting the point (1, C(1, mo )) (the coordinates are xO).

(c) In the case of conditions It is determined that 0<XO<1.

The following is the same as in case (a), (d) In the case of conditions It is determined that 1≦XO<2.

Point (0,C(0,me)), <1. C(1,mo
)) and a straight line (ii) that passes through (2, C (2, town)) and whose slope is the inverse of the straight line (i) to No. 71. Let xO be the coordinate of the intersection.

(e) In the case of conditions -1≦xl)<O.

The following is the same as in case (b). (f) In the case of conditions It is determined that -2<XO<-1.

Point (-1, C(-1, mo) ) and point (0, C(10,
The straight line ci t) passing through mo)) and the point (-2°C(-2
, mo)), and the coordinates of the intersection of the straight line (it) and the straight line (i) with the opposite sign are XQ.

2-(C(0,10)-C(-1,so))
In the case of conditions (g) and (11), it is determined that the blurred child cannot be detected because the blurring exceeds the assumed maximum blurring.

Although the procedure for obtaining xo has been described above, yo can also be obtained in the same manner.

Next, a method of driving the correction lens 32 shown in FIG. 3A will be described. FIG. 11 is a diagram for explaining a method of driving the correction lens 32. Here, only the horizontal component of image blur will be considered. In FIG. 11, the horizontal axis t represents time, and the vertical axis X represents the position of the image, i -3, i-2% i
-1s... represents the integration start time of CCD 51,
i-3, i-2, 11, . . . represent integration completion times.

The integration time TI is represented by T1,at,-t. The subject is exposed to AC light lll'1. When the CCD 51 is illuminated, the integration time is changed so that the amount of exposure of the CCD 51 is constant. Therefore, the integral II! It is not constant for 1 period. Integration start time 3, t-2, (-H, . . . are equal intervals T. A solid curve 301 shows the locus of the image on the CCD 51 when there is image blur. Correction lens 32
If blur compensation 1 is not performed, the image moves on this curve. Note that since the blur correction starts from t-t (1), the curve 301 is drawn passing through the point (to, 0). The broken line 302 is the trajectory of the correction lens 32. The broken line 303 is the trajectory of the correction lens 32. 32 represents the locus of the image on the CCD 51 when correction is performed by moving the point p-2,
P-1% PO, ... is the integration period 1-, ~1-. ．．
1-, to 1-2.1□I-t... represents the average position of the image. CCD integrated into II of tI' based on
The data for t.

It is read out between i++ and t,+, and calculation processing is performed to obtain the image position aPl. The correction lens 32 is driven by the data obtained from tl,2.

In addition, in the following explanation, P-2, I'+, P Os
..., Xl5X2, Xl, ..., x, , x2'
,xr ,,,,Xo %Xl %X!・
... is the name of a point and is also used as the value of the X coordinate of that point.

To correct from time 1-1 (, 1 from tQ
．． It is necessary to find the speed of the blur between the two.

Find the speed of blurring from the recent changes in the image position of the two points, and calculate tg
-t, and the speed of the blurring between is also the same]' 1l
11. The latest image position known at the time t-tg is P, . The time from P-2 to P-1 is TS-
Since (TI4-TI-L)/2, the predicted value VxO of the blur speed from t6 to t is. Therefore, during t(1-t), the correction optical system is driven at the speed of VxO.

Next, consider the case of time txt. At this time, since the correction lens 32 was driven by VxO, X.

is moved to the position. Since the position of PO is known at this point, the predicted value VXI of the blur speed at t to t2 can be calculated as follows.

Since the speed V81 is determined based on the recent image position data, it is considered that the predicted value of the blur speed between tl) and t is more accurate than ■ and 0. Therefore, the prediction error ER, , is calculated as follows.

ERX 1- (VX 0-Vx +) TS From these values, the position of the image at t-t, X, is 1' 1l11 as follows.

X+ -VX O"TS-G" ERX H-1V
xo-G(Vxo-Vx+)] 'TSG is 1'll
? It is called the j coefficient, and in FIG. 11, the case where the f-measure coefficient G is 2 is shown. Here, the correction lens 32 needs to be driven to the newly preset X5 point, but the driving requires a time of t, ~1, /. During that time, the image moves to the position of X,', so the correction lens 32 moves to the position of
, -t,), and is driven at a speed of VXj between 1 and t2.

Next, consider the case at time t-t2. At this time, the correction lens has been moved to the positions represented by X, -X, +vX, -TS. At this point, the position of Pl is known, so 12-1. )'1nIIVX of the blurring speed between
□ can be calculated as follows.

VX2 is considered to be more accurate than vx as a predicted value of the blur speed during j 1 "" L 2. Therefore, the prediction error, IER, is calculated as follows.

ERx 2 = (Vx + −VX 2 ) ・TS
From these values, the image position X2 at 1-12 is predefined as follows.

X2 - time is required. During that time, it can be predicted that the image will move to the position X2, so the supplementary L1 lens 32
2), and is driven at a speed of VX2 from t2' to t.

Next, time 1-1. Consider the case of At this time, the capture iF and lens X211 are at a position expressed by x2 mx2+VX2 e'rs. At this point, since the position of P2 is known, the deviation speed between t4 and t4 (however, txt
, ~t-1' are ignored. ) VXa is from t2 to t3
Prediction of blur speed between i11 [1 as VX? Since it is considered that the accuracy is higher than that, the 1,111 error ER is calculated as follows.

ERa − (Vx 2 − VX 3 ) ・TS From these values, the image position X at [−t, is X s ” Xz
-G-E Rx 5-X2 + (VX 2-G
(Vx 2-VX 3) 1ψTS can be predicted as a-1. The correction lens 32 now needs to be driven to the newly predicted position X, but the driving requires a time of ~t31. Meanwhile, the statue is X,
Since it can be predicted that the correction lens 32 will move to the position
is driven to the position Xs' mxa +vxa * (t, -ts>), and t3' to t*LJ are driven at the speed of VX3.

Similarly, X4, Vx 4, XB, VX
sl... is determined, and the correction optical system is driven.

Note that in this example, the r measurement coefficient G was constant.

However, depending on the situation of blur detection, the value of 6 may be changed during the correction sequence. If the predicted Al4 difference ER does not become smaller even after the shake correction of number -1 is performed, or if the sign of the pre-111 error ER, does not invert, the pre-i'l? The I coefficient G may be increased.

Furthermore, if the sign of the measurement error ER cannot withstand and repeats inversion, the value of G may be reduced.

Referring to FIG. 11, the time TR from time t6 to tz
is the camera's exposure time. As can be seen from FIG. 11, the sequence for performing camera shake correction is repeated lvI times within this photographing exposure time TH. Therefore, the correction is not performed only once or twice, but at least several times. Therefore, the actual camera shake calculation and correction calculation time must be performed within a reasonably short period of time. Also, the integral time TI,, TI2,...Tlo...TI of the sensor that detects the tremor. It also needs to be short poem 101. Shooting exposure time TR is several tens of meters
For an order of 8 to 10100O, T1° is several meters
It is about s. If this relationship is not maintained, the effect of camera shake correction iE will not be realized.

As described above, in the present invention, blur detection is performed using the output of the area sensor that detects the string object image, and the above-described blur correction is performed using the detected amount of blur.

Next, the interface between the interchangeable lens 3 and the camera body 12 will be explained. From interchangeable lens 3, the lens hard H
The correction magnification KBLXSKBLY (ft1 corresponding to each focal length when zooming is performed): (image movement amount on the film surface)/(drive pulse) is entered manually as information.

And drive pulse xi - image movement EIXo/correction magnification KB
It is determined by LX or yi-image movement H1Yo/correction magnification KBLY. Next, the data necessary to drive the lens described above (0 direction X: positive, negative ■ Drive speed V 3 ■ Drive pulse X, ■ Drive pulse X.

■Direction Y: Stop, negative ■Drive speed vy.

■Drive pulse Y ■Drive pulse Y.

■Return to origin information is output to the lens side. Here, the origin return signal (2) is set only when the lens is returned to the origin, and the lens side returns the correction lens 32 to the origin only when this signal is present.

Note that Ll&ll on the lens side is driven by the supplementary iE lens 32 as described above based on manually generated data.

Furthermore, when undetectable data is output, the same data as the previous data is output.

Next, control of the integration time of the CCD 51 will be explained. FIG. 12A shows a circuit for controlling the integration time of the CCD 51, and includes an illuminance monitor SPD and an 1111 optical circuit.

The i'llJ optical circuit includes an integration circuit and a reset circuit. The illuminance monitor SI'D is formed near the imaging area of the CCD 51. The charging current that is the output of the illuminance monitor SPD is integrated by an integrating circuit. The output of the integrating circuit is proportional to the amount of light incident on the illuminance monitor during the integration period. The purpose of using the illuminance monitor and photometry circuit is to keep the output of the CCD 51 constant even when the subject is illuminated by an AC light source such as a fluorescent light. This is because when capturing an image of a subject under illumination with an AC light source, if the integration time of the CCD 51 is kept constant, the readout output will vary each time. This will be discussed later. In order to stabilize the readout output of the CCD 51, the integration time must be adjusted to match the illuminance of the subject. In this example, C0D51
Simultaneously with the start of integration, the photometry circuit also starts integration, and the exposure amount of the CCD 51 is monitored by the output of the integration circuit.

FIG. 12B is a modification of FIG. 12A.

Next, the operation of the integral time control circuit shown in FIGS. 12A and 12B when the CCD 51 is used as a shake occurrence sensor will be described. The reset circuit is turned on and the integration circuit is reset. In the storage section of CCD? Ii hook is cleared. When the CCD 51 starts integrating, the reset circuit is turned off and the photometry circuit starts integrating.

FIG. 13 is a diagram showing temporal changes in the output of the photometric circuit. The horizontal axis t represents the integration time, and the vertical axis l represents 1 of the Ml optical circuit.
1 Indicates the magnitude of force. When the subject is illuminated by an AC light source, the output of the photometric circuit increases in a curved manner as shown in the figure. The photometric output is compared with an appropriate reference value 1o, and the photometric output l is 1. When the integral lIr1
The path is reset and the integration of CCD'51 is aborted.

The output of CCD'5i is read and it is determined whether the exposure amount is appropriate. If the exposure amount is appropriate, the reference value 1o is used for subsequent exposures. If the exposure amount is inappropriate, for example, the exposure amount is multiplied by 10Xk.
is set as a new reference value 1o, and subsequent exposure is performed.

Next, the operation of the integral time control circuit when the CCD 51 is used as a 7U viewfinder image sensor will be described. Output signal of photometry circuit 31, exposure compensation manual f stage 22
The reference value io is set based on the signal from the film sensitivity setting stage 23 and the signal from the film sensitivity setting stage 1 23 (this content will be explained in sequence step #70 of the main CPU with reference to FIG. 14).

The camera sequence including the blur detection, supplementary iE, remote finder, etc. described above will be explained based on the flowcharts of the main CPU and the control CPU for blur correction.

First, the main CPUI sequence will be explained. Referring to FIG. 14, the main CPU is in a standby state in the loop of step #5 (hereinafter the step is omitted). That is, the loop #5 waits for the switch S1 to be turned on by the first stroke of the release button. When the switch S1 is turned on, the AF completion flag AFEF and the signal are reset (it 10), and the photometry circuit 31,
AF detection module 14, image stabilization controller 18
The power to the necessary circuits is turned on (#15).

Next, the origin return signal of the correction lens 32 is output to the photographing lens 3 ($t20), and after the timer is reset, it is started (#25), the lens data of the photographing lens 3 is input (#30), and the camera is opened. Photometry is performed (#35
).

In #40, it is determined whether the AF completion flag AFEF is 1, and if it is 1, the program jumps to #65, and the AF completion flag AFEF is set to 1.
Otherwise, the program proceeds to #45. AF in #45
Detection is performed. In #50, it is determined whether or not the image is in focus as a result of the AFIQ output in #45. If it is in focus, the program will go to #55, if it is not in focus, it will go to #6.
Branch to 0. After the photographing lens 3 is driven to the focus position in #60, the program jumps to #45.

At #55, the AF completion flag AFEF is set to 1.

In #65, AE calculation is performed based on the 1llll light value obtained in #35, the film sensitivity, and the distance information obtained as a result of the AF in #45. In #70, the exposure amount of the CCD 51 is set.

At #75, switch S for the second stroke of the shutter button
It is checked whether or not 2 is turned on. If switch S2 is ON, the program goes to #105,
If it is not ON, the process branches to #80. In #80, it is determined whether the switch S1 is ON. If switch S1 is ON, the program jumps to #25. If it is not ON, the shutter button has not been pressed at all, so #8
In step 5, check the timer to see if a certain amount of time has elapsed. If a certain period of time has elapsed, the circuits of the 8-1 optical circuit 3, AF detection module 14, image stabilization controller 18, etc. Turn off the till device (#100) and then #
The program jumps to the standby state at step 5.

If a certain period of time does not pass, the AP completion flag AFEF becomes 0.
is set (#90), and the program proceeds to #80.

At #105, since the second stroke of the shutter button has been pressed, the release signal is set to H level, and the control CPU is notified of this.

After #110, the aperture of the photographic lens 3 is stopped down to the aperture value determined by the AE calculation in #65 (#110).

The optical system is switched from the electronic J-finder imaging optical system 41 to the shake detection optical system 4 [1 (#11')), the shake detection signal is set to H level, and the shake detection sequence of the control CPU is started ( #120). The AP mirror 7 is retracted (#125), aperture photometry is performed (#130), and the shutter speed at which the AE calculation was performed is corrected (1135).

At #140, the CPU waits until the exposure permission 111 signal from the control CPU becomes H level. When the exposure permission i1 signal becomes H level, the shutter curtain 9 is run and exposure control is performed (#14'5). When the exposure is completed, the origin La return signal is output to the photographing lens 3 (#150), the correction lens 32 is returned to the origin position, and the shake detection No. 1a is set to L level, which informs the control CPU that the exposure has ended. You will be informed (#155).

The film is advanced (#16C1), and it is determined whether continuous shooting mode is selected (#165). If it is not the quick shooting mode, the program branches to #180, and if it is the continuous shooting mode, the program goes to #170, where it is determined whether the switch S2 is ON. Here, if Sui-Sochi S2 is OFF, the program in continuous shooting mode but not continuous shooting is 1180.
Branch to.

If the switch S2 is ON, continuous shooting is performed, so blur detection (ci signal is set to H level (#175)) and the program jumps to #130.

After #180, the AF mirror 7 is turned to the detection position (#
180), and the aperture of the photographic lens 3 is opened (#18'5).
), release (No. 3 is set to L level (#IQ (1)
, the control CPUI is notified of this fact. The camera shake detection optical system 40 is switched to the electronic viewfinder imaging optical system 41 (#200), and waits for S2 to turn OFF (
#205), the program jumps to #80 and prepares for the next shooting.

Next, the operation of the control CPU will be explained with reference to FIG. At #15 in FIG. 14, the power to the camera shake correction controller 18 is turned on, and the sequence is started. The flag and output signal are reset (#B5), C
Integration starts after the integration of CD51 is reset (
#810). When the integration is finished at #B15, the read data is dumped to the image memory 64 (#82
0).

The dumped data is sequentially read out and D/A converted,
An image is displayed on the LCD 63 by the video I code processing circuit 62. At #B25, after the integration of the CCD 51 is reset, integration is started, and the program jumps to #B15.

If the integration is not completed at #B15, the program branches to #B30. At #B30, the release signal from the main CPU is checked, and if it is at H level, the program proceeds to #B35, and if it is at L level, the program branches to #B15.

At #B35, the CPU waits until the shake detection signal from the main CPU becomes H level and shake detection is started.

When the shake detection signal becomes H level, integration is started after the CCD 51 is reset. (#B40). At this time, the optical system has already been switched to the shake detection optical system by the main CPU. At #B45, the completion of the integration of CCD'51 is awaited. After the CCD 51 completes the integration, the read data is dumped into the base memory 65 and reference memory 66 (#B 50). After the integration of CCD'51 is reset, integration is started (#B55).

Contrast calculation is performed i7 (#B60), and blocks used for blur detection are selected (#865).
). The completion of the integration by the CCD 51 is awaited (#B70). After the integration is completed, the read data is transferred to the reference memory 66 (#B75), and after the CCD 51 is reset for integration, the integration is started.

A correlation calculation (#B85) and an interpolation calculation (#B90) are performed to detect the amount of image blur. The lens data of the photographic lens 3 is read out (#B95), and the correction amount and direction of the correction lens 32 are calculated (#B100).

In #8105, it is determined whether or not the shake detection signal from the main CPυ is at the L level. If it is at the L level, the shake detection is ended and the program branches to #B130. #B10
If the shake detection signal is not at L level in step 5, shake detection continues, and the program advances to #B110. The exposure permission signal is set to H level (C#8110), notifying the main CPU that exposure may be started, lens correction data is output to the photographing lens 3 (#B120), and the program jumps to #870. .

At #B130, the release signal from the main CPU is L.
determine whether the level is If it is L level, the shooting has finished and the program jumps to #BIO. If it is not the L level, continuous shooting is being performed, so the program advances to #B135 and waits for the detection signal to become the H level. If the detection signal becomes H level, the program is #B40
Jump to.

FIG. 16 shows another embodiment of the control CPU. Even after the release starts, #B247 and #B275
The data is dumped to the image memory, and image information for camera shake detection is displayed. In other words, the entire photographic area is displayed on the LCD 63 until just before exposure, but once exposure begins, the optical system is switched and the screen shake detection area is displayed on the LCD.
This is done on D63.

〔Effect of the invention〕

As described above, according to the present invention, a pellicle is used in a blur detection camera to transmit light from an object to an area sensor serving as a blur detection sensor, so image blur is detected using the TTIJ method. Further, an image corresponding to the output of the area sensor is output to an electronic viewfinder, which is a two-dimensional display device. As a result, since image blur is detected using the TTL method, there is no parallax, unlike the external light method. Furthermore, since the camera shake is detected using an image, it is possible to provide a shake detection camera that can also detect the shake of a subject.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a main CPU that forms the central part of a shake detection camera with an electronic viewfinder according to the present invention, and FIGS. 2A to 2D are diagrams for explaining how the electronic viewfinder is attached to the camera body. FIG. 3A is a diagram showing the optical system of the shake detection camera with an electronic viewfinder according to the present invention, and FIG. 3B is a diagram showing details around the vibration detection optical system driving means, and FIGS. The figure shows the screen of an electronic finder, FIG. 5 shows a capsule-shaped example of the optical system shown in FIG. 3A, and FIGS. 6A and 6B show the embodiment shown in FIG. 5. FIG. 7 is a block diagram of camera shake detection and correction means, and FIG. 8 is a diagram showing how the electronic viewfinder is attached to the camera body.
FIG. 9 is a diagram for explaining the correlation value for determining the amount of image blur, and m 1 ('
) is a diagram for explaining the interpolation calculation method.
Figure 1 is a diagram for explaining the method of driving the correction lens, Figures 12A and 12B are circuit diagrams showing a circuit that controls the integration time of the CCD, and Figure 13 is a diagram for explaining the time of the photometry circuit output. 14 to 16 are flowcharts for explaining the operations of the main CPU and control CPU of the shake detection camera with electronic finder according to the present invention. 1 is the main CPU% 2 is the SPD, 3 is the photographic lens, 4 is the 1-stage aperture drive, 5 is the focal point, SJM3 driving means, 6 is the photographic lens circuit, 7 is the AF detection mirror, 31 is the photometry circuit, 3
2 is a correction lens, 40 is a shake detection optical system, and 41 is an electronic/
In the finder imaging optical system, 42 is an optical system switching stage f. In the drawings, the same reference numerals indicate the same or corresponding parts. Figure 2A Figure 2C Figure 4A Figure 4B Figure 4C

Claims (1)

[Scope of Claims] A movable image stabilization optical system provided in a photographic lens, a pellicle provided in an optical path from a subject to a film, an area sensor that captures an image of the subject, and the area sensor. a two-dimensional display device that displays the output of the area sensor; a shake detection unit that detects camera shake by calculating two temporally different outputs of the area sensor; and when the shake detection unit determines that there is camera shake; A camera shake detection camera comprising: a correction means for correcting the camera shake by moving the shake correction optical system.