JPH01297612A - Automatic focusing device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

Detailed Description of the Invention [Industrial Application Field] The present invention relates to an automatic focusing system that repeatedly adjusts focus by detecting the defocus state of a photographic lens by receiving object light that has passed through the photographic lens, for example. This relates to improvements in equipment.

[Conventional technology]

Conventionally, in automatic focus adjustment cameras, in the case of a subject with large movement, the amount of defocus changes during focus detection and lens driving, and the amount of defocus to be corrected and the amount of detected defocus may differ significantly. As a result, there is a problem that the subject is out of focus when the lens drive ends.

In order to solve this problem, considering the amount of detected defocus in each cycle, the amount of lens drive, and the time interval of each cycle, the function of the image plane position and time caused by the movement of the subject is expressed as a linear function and a quadratic function. The applicant has developed a focus detection method that approximates the lens drive amount and corrects the lens drive amount in Japanese Patent Application No.
63728.

[Problems to be Solved by the Invention] The present invention relates to further improvements to the focus adjustment using the above prediction method, and aims to prevent erroneous lens drive based on focus detection errors and lens drive errors.

Incorrect lens driving due to the above error will be explained below.

FIG. 2 is a diagram for explaining the lens drive correction method shown in the above-mentioned patent application. The automatic focus adjustment device operates in a servo AF mode that follows the movement of the subject, and the horizontal axis in the figure represents time t1, and the vertical axis represents the image plane position X of the subject.

The curve x (t) represented by a solid line means the image plane position at time t of a subject approaching the camera in the optical axis direction when the photographing lens is at infinity. I shown with a broken line! (1) means the photographing lens position at time t, and when x (t) and 2 (1) match, the image is in focus. [tl, t+'] indicates the focus detection operation time, and [tl', t++I] indicates the lens drive operation time. In addition, as in the example shown in the figure, the image plane position is a quadratic function (a t" + b t +
By assuming that the current and past three image plane positions (tl
, XI) (t21 X2) (t31 Time required for drive, release time lag: time from issuing the release command to starting exposure).

Next, erroneous prediction due to focus detection error and lens drive error will be explained using FIG. 9.

FIG. 9 shows the relationship between the image plane position and time, and the solid line indicates the position of the image plane that actually moves.

That is, (tl, XI) (t21 X2) (t3.X
3) Quadratic function x = a t” + b t + c
This can be approximated to (1).

On the other hand, the image plane position recognized by the camera is calculated from the detected defocus amount and lens drive amount, but since there is an error in this defocus amount and lens drive amount, Errors also occur in the image plane position recognized by the camera.

The difference between the image plane position recognized by the camera and the actual image plane position is δ1. δ2. When δ3, (t1+xl')02,
The quadratic function passing through X2' ) (t31 X3' ) is x=a't2+b't+c'
(2), and a prediction error of δ2 occurs between the lens position x4' found using this function and the image plane position X4 at t4.

This prediction error δ2 becomes an error amount several times the focus detection error or lens drive error.

In this way, the focus detection error and lens drive error, which did not pose a problem with the conventional focus adjustment device, are amplified several times by the prediction error of the predictive AF, resulting in an unacceptable amount.

[Means for solving problems]

The present invention aims to solve the above problems,
By correcting the coefficient of the higher-order term among the several-order functions used for prediction calculations as a function of the latest defocus amount, the influence of errors occurring in the focus detection system or lens drive system is reduced, and the prediction accuracy is improved. The aim was to improve the

〔Example〕

Hereinafter, the present invention will be explained in detail based on illustrated embodiments.

First, the focus detection method used in the present invention will be explained using FIG. A field lens 2 is placed on the same optical axis as the photographing lens 1 whose focus is to be detected.

Two secondary imaging lenses 3 and 4 are arranged at symmetrical positions with respect to the optical axis behind it. Furthermore, sensor rows 5 and 6 are arranged behind it. A diaphragm 7 is provided near the secondary imaging lens 3°4, and the field lens 2 is connected to the photographing lens l.
The exit pupil of is substantially imaged on the pupil plane of two secondary imaging lenses 3.4. As a result, the light flux incident on the secondary imaging lens 3.4 exits from areas of equal area that do not overlap with each other and correspond to the respective secondary imaging lenses 3 and 4 on the exit pupil plane of the photographing lens 1. It becomes what is given. The aerial image 8 formed near the field lens 2 is transferred to the secondary imaging lens 3,
When the image is re-formed on the surface of the sensor arrays 5 and 6 by the sensor array 4, the 2 images on the sensor arrays 5 and 6 are refocused based on the displacement of the aerial image 8 in the optical axis direction.
The statue will change its position. Therefore, by detecting the displacement of the relative position (the amount of image shift), the focal state of the photographic lens 1 can be known.

For this purpose, the amount of image shift can be detected by performing a correlation calculation on the photoelectric conversion outputs (image signals) of the two images formed on the sensor arrays 5 and 6. No. 58-142306, JP-A-59-107
No. 313, JP-A-60-101513, etc. are disclosed. The difference between the focal position of the photographing lens 1 and the film surface, that is, the defocus amount, is calculated from the image shift amount obtained in this manner.

FIG. 1 is a block diagram schematically showing a camera having the present embodiment apparatus equipped with the focus detection method as described above. In the figure, PRS is a camera control circuit, for example, an internal C
PU, ROM, RAM, EEPROM.

It is an l-chip microcomputer with A/D conversion functions, etc., and controls various camera operations such as automatic exposure function, automatic focus function, film winding and rewinding according to the sequence program stored in the ROM, and also has a communication function. The signals So, Sr.

5CLK is used to communicate with peripheral circuits and the lens FLNS to control the operation of each circuit and lens FLNS.

SO is a lens driving data signal output from the control circuit PR3, SI is a data signal input to the control circuit PRS, and 5CLK is a synchronization signal for the data signals So and SL.

LCM is a lens communication circuit, and when the camera is in operation, it supplies the lens power supply VL to the lens side, and the control circuit PR3
to H level (meaning high level) signal CLCM
The input serves as a buffer for communication between the camera and lens. Therefore, the control circuit PR3 outputs an H level signal CL.
When CM is input and a predetermined data signal SO is sent in synchronization with the synchronization signal 5CLK, each of the signals 5CLK and So is buffered as a buffer signal LCK via the camera-lens interface.

Output DCL to lens FLNS. At the same time, a signal DLC, which is information such as the current position, is sent from the lens FLNS.
The buffer signal is output to the control circuit PR3 as a data signal SI in synchronization with the signal 5CLK. SDR is a line sensor S consisting of the sensor arrays 5, 6, etc. for focus detection.
This is a drive circuit for NS, and the signal C3 from the control circuit PR3.
Selected when DR is at H level, each signal So, 5CLK
controlled by

CK is a line sensor driving clock φ1. This is a clock for generating φ2, and INTEND is a signal that notifies the control circuit PR3 that the accumulation operation has ended. The output signal O8 from the line sensor SNS is the clock φ1.
It is a time-series image signal synchronized with φ2, and the drive circuit SDR
After being amplified by the internal amplifier circuit, it is output as a signal AO3 to the control circuit PR5.

The control circuit PR5 inputs the signal AO8, which is the image signal, from an analog input terminal, performs A/D conversion using an internal A/D conversion function in synchronization with the clock CK, and sequentially stores the signal AO8 at predetermined addresses in the RAM. Same (signal 5AGC which is the output signal of the line sensor SNS is the A of the line sensor SNS)
This is the output of the GC control sensor (not shown), and the drive circuit S
The signal is input to the DR and used for storage control of the line sensor SNS.

In-lens control circuit LPR in synchronization with buffer signal LCK
The signal DCL input to 3 is from the camera to the lens FLN.
This is the data of the command for S, and the lens F for the command.
The operation of the LNS is predetermined.

The in-lens control circuit LPR8 analyzes the command according to a predetermined procedure, performs focus adjustment and aperture control operations, and outputs the signal D.
Various parameters (open F number, focal length, coefficient of defocus amount vs. extension amount, etc.) are output from the LC.

This embodiment shows an example of a single lens that is fully extended, and when a focus adjustment command is sent from the camera, the focus adjustment motor L is moved according to the direction of the drive amount sent at the same time.
MTR is driven by signals LMF and LMR to move the photographic lens LNS in the optical axis direction to perform focus adjustment. The driving amount of the photographing lens LNS is monitored by the pulse signal 5ENC of the encoder circuit ENC, and when the predetermined movement is completed, the signals LMF and LMR are set to L level (meaning low level) and the focus adjustment motor LMTR is activated. Brake.

When an aperture control command is sent from the camera, a stepping motor DMTR, which is known for driving an aperture, is driven in accordance with the number of aperture stages sent at the same time.

Note that since the step motor can be oven controlled, an encoder is not required to monitor its operation.

The general operation of the parts related to the present invention of the camera configured as described above will be explained according to the flowchart of FIG.

When the power switch (not shown) is turned on, the control circuit PR3
Power supply to the control circuit PR3 starts, and the control circuit PR3 starts executing the sequence program stored in the ROM.

First, in step 2, the contents of a counter C0UNT for counting the number of times the AF operation has been performed is set to "0", and the process proceeds to step 3. Here, the state of the switch SWI, which is turned on by the first stroke operation of the release button, is detected, and if the switch SW1 is off, the process advances to step 4 and the release button is pressed.
All flags for various controls set in the AM are cleared, and the process proceeds to step 5, where the contents of the counter C0UNT are set to "0".

The operations in steps 3 to 5 are repeated until the switch SWl is turned on or the power switch is turned off, and when the switch SWI is turned on, step 6 is performed.
transition to. Step 6 is the "rAF control" subroutine, in which AF control operations are executed. During this, when the second stroke of the release button is operated and the stitch switch SW2 is turned on, the interrupt processing function starts the release operation, controls the exposure operation, that is, the aperture, and shutter speed, and after the operation is completed. Then, the photographing operation for the film frame is completed by charging the shutter, feeding the film, etc.

Next, the "rAF control" subroutine carried out in step 6 will be explained using FIG. 4.

First, in step 102, the state of the flag PRMV is detected. The flag PRMV is a flag related to lens control as will be described later, but since all the flags are cleared in step 4 at the initial stage, the flag PRMV is also cleared, and the process proceeds to step 105.

In step 105, the current time of the built-in timer TIMER of the computer PR3 is input to the memory TMB, and the contents of the memory TMB are input to the memory TMA. Therefore, the current AF start time is stored in the memory TMB, and the TM
The previous AF start time is input to A, and the memory TMA
The subtraction of TMB and TMB contents, TMB-TMA, becomes the previous ranging operation time interval.

In step 106, the contents of memory TM2 are transferred to memory T.
Enter M. In step 107, the memory TMB,
The contents of TMA are subtracted and input into memory TM2. Therefore, the distance measuring operation time interval from the previous time to the present time is inputted into the memory TM'2, and the distance measuring operation time interval from the previous time to the previous time is inputted to the memory TM.

The next step 108 is an "image signal input" subroutine, and by executing this subroutine, the A/D conversion value of the image signal (signal AO3) obtained by the line sensor SNS is stored at a predetermined address in the RAM. Ru. Step 109
Then, the contents of the counter C0UNT, which counts the number of distance measurements, is incremented by "+1". In step 110, a "focus detection" subroutine is executed. In this subroutine, the focus of the photographing lens LNS is detected from the image signal data stored in the RAM, and the defocus 1lDF is calculated. In step 111, if the defocus ff1DF detected in the "focus detection" subroutine is larger than a certain value, it is determined that the main subject is out of the distance measurement frame, and step 1
12, the contents of the counter C0UNT are set to "0" and returned to the initial state.

On the other hand, if it is determined in step 111 that the constantly moving subject is within the distance measurement frame, the process proceeds to step 113, and the defocus fiDF detected in step 110 is stored at a predetermined address (memory).
Store in DF3. Also, the contents of each memory DF, to DF3 are updated, and DF contains the previous defocus amount and DF2.
stores the previous defocus amount, and DF3 stores the current defocus amount. Step 114 is a "lens drive amount calculation" subroutine, the details of which will be described later. Step 11
5, the lens drive amount D obtained in step 114 is
Store L at a predetermined address (memory) DL2. Also, input the contents of memory DL2 to memory DL, and
Let the contents of L be the previous lens drive amount, and the contents of DL2 be the current lens drive amount. In the next step 117, the flag PRMV indicating lens drive execution is set to 1j, and then the process proceeds to step 118 to return to the l'-AF control subroutine.

In the flow shown in FIG. 4, when the "rAF control" subroutine of step 6 is called again, the state of the flag PRMV is detected in step 102 as described above.

When the lens is driven in the above-mentioned "rAF control" subroutine, the flag PRMV is set to "1" in step 117, so the process proceeds to step 103 here. In step 103, lens communication is performed to detect the current driving status of the photographic lens LNS, and at the same time as it is detected that a signal indicating that the predetermined driving instructed in step 115 has been completed is input from the lens FLNS side, step 104 The process proceeds to step 5, sets the flag PRMV to "0", and executes the flow from step 5 onwards. If a signal indicating that the lens FLNS is still being driven is being outputted, the process moves to step 119 and returns to the ``rAF control'' subroutine. Therefore, in the "rAF control" subroutine, a new focus detection operation and lens control are performed only when the photographic lens LNS is not driven.

Next, the operation in the "lens drive amount calculation" subroutine performed in step 114 will be explained using FIG. 5.

When this subroutine is called, in step 202, it communicates with the lens FLNS and sends two data "SJ
, rPTHJ. Here, rSJ is the photographing lens L
It is a coefficient of the NS-specific "defocus amount" versus the "adjustment amount of the focusing lens". For example, in the case of a single lens of the entire extension type as in this embodiment, the entire photographing lens LNS is a focusing lens. Therefore, l'-3=1j, but in the case of a zoom lens, "S" depends on the zoom position.
changes. In addition, "PTHJ is an encoder E that is linked to the movement of the focusing lens (in this example, the photographic lens LNS).
This is the amount by which the focusing lens is extended per pulse generated in the NC.

In step 203, it is determined whether the in-focus position for the next photographing can be predicted (predictive AF) from the contents of the counter C0UNT that counts the number of distance measurements, and if prediction is not possible, the process moves to step 204. Here, the amount of extension of the focusing lens is converted into the number of pulses generated by the encoder ENC using the defocus amount DF, rsj, and rPHTJ, so-called lens drive: IIFP, FP=DFXS/PTH... - Calculate using the formula (3) and return to the "lens drive amount calculation" subroutine.

On the other hand, in step 203, if the data necessary for prediction is accumulated and predictive AF is possible (in this embodiment, the past two
If the data related to the previous distance measurement is stored, it is determined that predictive AF is possible.) The process advances to step 205.

Note that details of the predictive AF method in this embodiment will be described later. In step 205, the relationship between the image plane position and time is approximated to a quadratic function.

In other words, considering the relationship between image plane position X and time t as shown in the following equation,
The coefficients a, b, and c of this equation are the current and past two defocus amounts DF 1. DF 2. It is determined from DF3 and lens drive amounts DL, DL2.

X=α・a・t2+bt+C・・・・・・・・・(4) α
-1d-(lDj'31+1DI-2++ ,,,
,,,,,,,,,,,,,,,,,,,,,,,,,,,
,,,,,,(4)'c=DF 3...
......(7) (d is a constant) And the coefficient α obtained from the above formulas (5) to (7)
The next lens drive amount DL is calculated from a, b, and c using the following formula.

DL=S (a-aTL"+bTL+DF 3)-(8
) Here, TL is the release time lag (LETL) and A
The sum of F time lags (AFTL), that is, TL=LET
L+AFTL (9). Furthermore, A
The F time lag AFTL is set to 1M2, TM, or (TM, +TM 2/2).

Further, each data of the above equations (4)' to (9) is stored in each of the above memories DLl+DL2. DF, ~DF3. TM
,.

1M2 data is used.

In step 206, the lens drive fi determined by equation (8)
The number of pulses PF of the encoder ENC corresponding to DL is calculated from the following formula.

PF=D, L/PTH (10) After the above operation is executed, the subroutine "Lens drive amount calculation" is returned.

Next, prediction A using the above equation (8) in this example
The method F will be explained using FIG.

Time of past distance measurement 1. ．． 12 and current time t3
Defocus amount DF , DF2. Time 1 from DF3 and lens drive fiDLl+DL2. ．． 12°t3
The image plane position Xl+X2+x3 is found. still,
The lens drive fiDL here is the defocus fiDF
In other words, it is a value converted to the image plane position.

Considering the image plane position of the lens at t3 in the figure as the origin, t1=-TM2-TM1. x 1=-DL
2-DL, +DF 1...(11)t2=
・TM 2+ X 2 =・DL 2 +D
F 2 ... (12) t3=0.
X3=DF3 (13) The image plane position X and time t are approximated by the following quadratic function.

x = a t” + b t + c ・・・・・・
...(14) Next, the above boundary condition equations (11) to (13)
) into the above equation (14), we get the following.

DF3=C・・・・・・・・・・・・(15)-D
L 2 +DF 2 = aX (-TM 2)”+b
X (-TM 1) +c・・・・・・(16) DL 2 DL H+DF 1= a (TM 2
TM 1) 2+b (-TM 2-TM,) +c
・(17) From the above formulas (15) to (17), a,
b and c are determined as shown below.

c=DF 3 ・・・・・・・・・・・・・・・(1
9) Here, time TL (=
The image plane position X4 after LETL+AFTL) can be found from the following equation.

x 4 = a 6 TL"+b-TL+c-・-(20
) Therefore, the required lens drive amount DL is as follows.

DL3=DL=x4-x3+DF3=aT
L"+bTL+DF3 curve -<21') The above is a prediction formula for ideal image plane movement when there are no errors occurring in the focus detection system or lens drive system. However, in an actual camera, the focus Detection errors and lens drive errors occur, and as described above, the image plane position is often predicted incorrectly.

Therefore, in the present invention, the following correction is performed in order to minimize such prediction errors.

In the above prediction formula (21), the term that is most affected by the focus detection error and lens drive error and generates a large prediction error is the quadratic term, and by correcting the coefficient of this term, It is possible to reduce the amount of prediction error.

Generally, when the detection defocus 1 is small, such as when the subject is moving slowly or at a long distance, the movement of the image plane is linear, and a prediction formula corresponding to this is useful for suppressing prediction errors. It is desirable that it be a linear function. In addition, even if the actual image plane movement on the plate is quadratic in such a case, the prediction error resulting from predicting it as a linear function is relatively small because the original amount of defocus is small. (This is not a problem.On the other hand, when the detected defocus amount is large, such as when the subject is moving at high speed or at a short distance, the movement of the image plane is generally curved, and the corresponding The prediction formula should be a quadratic function.

Furthermore, even if the current image plane movement is actually linear, because the movement is fast, the lens will be driven to near the close focusing distance where the image plane position will change rapidly during the next distance measurement period. Therefore, in order to reduce the prediction error, a quadratic prediction formula is required.

In order to obtain a prediction formula that satisfies these conditions, (1
A correction term that changes continuously or stepwise may be added to the quadratic term in equation 4) so that it becomes 1 when the detected defocus 1 is large and becomes 0 when it is zero. For example, α shown in formula (4)' is such, X=αa t" + b t + c...
...(4) α:= 1 6・(lDFj DF21
1 ・・・・・・・・・(4)'Detection defocus I
If DF31+IDF21 is small, α takes a value of 0, and if large, α takes a value close to l. Note that d is a constant, and may be set to a value of about 10 to 50 in the case of a negative-eye reflex camera.

Figure 7 shows l DF 31 + l when d=30.
FIG. 3 is a diagram showing the relationship between DF 21 and α.

Figure 6(a) shows the movement of the photographing lens LNS when it is assumed that the function of image plane position X and time t can be approximated to a quadratic function, and Figure 6(b) shows the movement of the image plane The movement of the photographing lens LNS is shown on the assumption that the function of the position X and the time t can be approximated to a linear function.

According to the image plane movement prediction method of the present invention expressed by the above equation (4), the operations expressed in FIGS. 6(a) and 6(b) are performed depending on the magnitude of l DEF 31 + I DEF 21. changes continuously, making it possible to always obtain a state with little prediction error. As a result, we provide a camera equipped with an automatic focus adjustment device that can drive the photographic lens LNS to an ideal position taking into account tracking delays due to AF time lag and release time lag, and also stabilizes the movement of the photographic lens LNS. It becomes possible to do so.

As mentioned above, in the prediction calculation at step (205) in this embodiment, the lens drive amount is calculated using equation (8) based on equations (4) and (4)', so that (
The operation b) is continuously switched, and as described above, suitable automatic focus adjustment is possible.

In the above embodiment, α is determined based on the formula (4)', but as the formula (4)', α=: 1 6-
(lDr3CF!l+1DFIl) or α=1-d-
It may also be used as IDF3I.

Furthermore, as equation (4), x=αat2+βt+c,
α=1d-(lDFmDF21)β=l-e″(l
On l + l Dn l ) (e is a constant)
You can also use it as

[Effects of the Invention] As explained above, the present invention includes a focus detection circuit that determines the amount of defocus of a photographic lens, and a lens drive circuit that drives the lens based on the output of the focus detection circuit. In an automatic focus adjustment device that repeatedly performs a focus amount detection operation and a lens drive operation based on the detection result, the image of the subject after a predetermined period of time is determined based on the defocus amount determined by the focus detection circuit multiple times in the past. An arithmetic circuit that calculates the surface position using a several-order function is provided, and the lens is driven so that the image plane position of the object and the lens image plane position match after a predetermined time, and the highest-order coefficient in the above-mentioned several-order function is used. is corrected as a function of the latest detected defocus amount, so it has the following effects.

1. In the servo AF mode that follows the movement of the subject, it is possible to drive the lens with high stability even when the detected defocus amount is small, that is, when the image plane movement speed is slow, but it is high when the image plane movement speed is fast. It is now possible to maintain tracking performance.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of the main parts of a camera having an automatic focus adjustment device according to the present invention, and FIG.
Figure 3 is a flowchart of the main parts of the camera shown in Figure 1, and Figure 4 is the rAF control shown in Figure 3.
The flowchart showing the subroutine, Figure 5, is also "
Flowchart showing the lens drive amount calculation subroutine, FIGS. 6(a) and 6(b) are diagrams explaining the effect of correction in the embodiment apparatus of the present invention, and FIG. 7 shows the relationship between the defocus amount and the correction value α. 8 is a diagram explaining the focus detection method in this embodiment, FIG. 9 is an explanatory diagram explaining the drawbacks of the predictive AF method, and FIG. 10 is an explanation explaining the operation of the predictive AF method of the present invention. It is a diagram. PH1...Computer LPR3...Control circuit SNS...Sensor device patent applicant Canon Corporation

Claims (1)

[Claims] A focus detection circuit that determines the amount of defocus of the photographic lens;
In the past, several automatic focusing devices have been developed that include a lens drive circuit that drives a lens based on the output of the focus detection circuit, and repeatedly perform an operation of detecting the amount of defocus by the focus detection circuit and a lens drive operation based on the detection result. An arithmetic circuit is provided that calculates the image plane position of the subject after a predetermined time using a several-order function based on the amount of defocus obtained by the focus detection circuit in the second focus detection circuit. An automatic focusing device, characterized in that the lens is driven to match the positions, and the highest order coefficient in the multi-order function is corrected as a function of the latest detected defocus amount.