JPH0736055B2 - Automatic focus adjustment device - Google Patents

Description

TECHNICAL FIELD OF THE INVENTION The present invention relates to an automatic focusing device for an optical device such as a camera.

(Background of the Invention) As an automatic focusing device for a camera, a defocus amount corresponding to a deviation amount between a planned focal plane of an objective lens (imaging optical system) and an actual image plane of a subject is calculated at a certain time interval. However, it is known that the objective lens is driven according to the defocus signal to perform automatic focus adjustment. However, when the objective lens is driven in accordance with the defocus signal that is discretely output as described above, if the subject is moving in the optical axis direction of the objective lens during driving of the objective lens, the lens drive is Since it is performed independently of the movement of the subject, there is a drawback that proper focus adjustment is not performed.

(Object of the Invention) It is an object of the present invention to solve these drawbacks and to obtain an automatic focusing apparatus that predicts the movement of a subject even when the subject is moving and follows the focus.

(Example) Below, the content of this invention is demonstrated based on a specific example. FIG. 1 is a block diagram showing a configuration in which the present invention is applied to a camera. As shown in FIG. 2, the focus detection means 104 for detecting the focus adjustment state of the taking lens 101 in FIG. 2 is composed of focus detection optical means 201, photoelectric conversion means 202, and focus detection calculation means 203. Photoelectric conversion means 202 is a photographing lens
It receives a subject image formed by 101 and focus detecting optical means 201, and generates an image output according to the image.
This means 202 has a charge storage type photoelectric conversion section, the charge storage time can be determined by devising the device so that the photoelectric conversion section itself can have such a function, and the image output is determined by the focus detection calculation section 203. It may be determined by processing. In any case, the focus detection unit 104 transmits the timing of starting and ending the charge accumulation to the control unit 109.

From the end of charge accumulation to the start of the next charge accumulation, charge transfer and calculation of the defocus amount are performed. The defocus amount is calculated by a known method, and this is also the control means 109.
Be transmitted to. The time counting means 108 receives the charge accumulation start and end timings or other timings from the control means 109, counts the time interval of each timing, and the control means 109 stores the result in the memory 110.

In FIG. 3, the horizontal axis represents time t and the vertical axis represents the movement position x of the lens 101. Solid lines 301 and 301 'represent the movement locus of the lens 101, and a thick solid line 302 represents the object in the imaging lens 101.
Corresponds to the locus when the object moves in the optical axis direction. If the object is stationary, it becomes parallel to the time axis, and when the object moves, it has a certain inclination as shown in the figure. If the movement loci 301, 301 'of the lens 101 substantially coincide with the object movement locus 302 representing the movement of the image plane position based on the movement of the object, the imaging lens 101 is in focus. Lens movement locus 301
Is called converging drive from the point Q _{0, which is} greatly deviated from the object movement locus 302, to the point Q _{1 at} which the first substantially coincides, and as shown in the lens locus 301 ′, the object movement locus 302 from the point Q ₁ Driving the imaging lens so as to follow is referred to as follow-up driving. Time t _n-1 , t _n , t _{n + 1} , t
_{n + 2} indicates a time point when the charge accumulation of the photoelectric conversion means 202 is started, time t ′ _n−1 , t ′ _n , t ′ _{n + 1} indicates an end point of the charge accumulation, and t ″ _n is The time point when the convergence drive is changed to the follow-up drive is shown, T _n-1 and T _n represent charge accumulation time, and T _n ^C _-1 and T _n ^C
Is the calculation time, during which the focus detection calculation control means 203 calculates the defocus amount based on the data from the photoelectric conversion means 202, and based on this defocus amount and the information stored in the memory 110, The estimated defocus amounts X _n-1 and X _n, which will be described in detail, and the correction amounts δ _n-1 (τ) and δ _n (τ) that correct the estimated defocus amount as the target object moves are calculated. . This estimated defocus amount is at time t _n-1 ,
The defocus measurement number means 106 is set to t _n , t _{n + 1} , and t _{n + 2} . T _'n is from time t _"n time T' represents the time to _{n, M n-1, M} n is the lens during each time T _n-1, T _n 101
Represents the amount of movement including the direction, for example, M _n is time t _n
Is a quantity that represents the distance between the position of the lens 101 and the position of the lens 101 at time t ′ _n including the direction. Therefore, when the lens 101 moves in one direction and in the opposite direction, M
The signs of _n-1 and M _n are reversed. M _n ^C _-1 and M _n ^C are T _n ^C _-1 and T _n , _{respectively.}
^The amount of movement of the lens 101 during ^C is shown including the direction.

In the present embodiment, the position of the image plane at a predetermined timing is obtained, and based on this, the amount of movement of the image plane due to the movement of the object is corrected and the lens 101 is driven. Although a method of calculating the amount related to the movement of the target object (subject) will be described later, the correction amount of the movement of the image plane due to the movement of the object is represented by δ (τ) as a function of time τ.

Further, the estimated defocus amount is estimated at the time of the calculation end as described above, but at this time, since the time has already passed from the time of photoelectric conversion, the time of photoelectric conversion (for example, the predetermined timing will be described later). As described above, the amount of movement of the image plane position due to the movement of the object in the elapsed time is the corrected amount with respect to the image plane position at the center of the charge accumulation time) (the calculation method will be described later).

In this way, the two estimated defocus amount itself is the predicted amount.
By adding (τ), it is possible to predict the image plane position after δ has elapsed from the end of the calculation.

In FIG. 3, at the time t _n-1 when the estimated defocus amount X _n-1 is set in the defocus amount counting means 106, the estimated defocus amount X _n is calculated from the position Q ₀ of the lens 101 at that time.
_Let Q ₂ be a point separated by _n-1 . This point Q ₂ is the predicted image plane position at the calculation end timing t _n-1 . When a correction straight line 310 (shown by a broken line) is drawn from this point Q ₂ substantially parallel to the object movement locus 302, the horizontal line from the point Q ₂ and each time point t _n-1 + τ with this correction straight line 310 are drawn. Represents the correction amount δ _n-1 (τ).

The same applies to the correction amount δ _n (τ), and the time t _n
Lens position Q ₃ at and estimated defocus amount at this time
The point Q ₄ is determined as the predicted image plane position at the timing of time t _n from X _n, and the object movement trajectory 30 is determined from this point Q _4.
A correction line 311 is drawn almost parallel to 2.

Referring again to FIG. 1, the control means 109 controls the object movement correction pulse having a frequency corresponding to δ _n-1 (τ) so that the correction amount δ _n-1 (τ) at time t _n-1 + τ will be obtained with the passage of time. Is sent to the defocus amount counting means 106, and the value of the estimated defocus amount is corrected to track the movement of the subject.

In this way, the defocus amount counting means 106 holds the estimated defocus amount at each moment in consideration of the movement of the subject, and the drive signal generating means 105 responds to the contents of the defocus amount counting means 106, Signals are generated to cause the driving motor of the lens 101 to rotate normally, reversely, and stop, depending on whether the content is a positive value, a negative value, or zero.

The drive means 102 including the drive motor receives the signal and drives the lens 101. The movement amount signal generating means 103 generates a pulse as the lens 101 moves. As a method therefor, an independent means may be provided as described in JP-A-58-88710, or when the drive motor is a pulse motor, a drive pulse may be used instead.

The signal pulse relating to the lens movement amount is a movement amount counting means.
Counted by 107, the count value is the start of charge accumulation t _n-1
And read at the timing of the end t'n _-1 and then reset to zero. In this way, the movement amount counting means 107 causes the lens movement amounts M _n-1 and M _n ^C at each time interval.
_-1 is detected and stored in the memory 110.

The driving of the lens by the drive signal generating means 105 is performed in the direction to make the content of the defocus amount counting means 106 zero based on the content of the defocus amount counting means 106. The estimated defocus amount in the defocus amount counting means 106 is up-down or down-counted toward zero by a signal pulse relating to the lens movement amount from the movement amount signal generating means 103.

In this way, the estimated defocus amount held by the defocus amount counting means 106 gradually decreases as the lens moves and approaches zero, and the system approaches the in-focus state. In the case of FIG. 3, such convergence drive continues during t _n-1 → t _n . However, it is preferable that the lens moves at a uniform speed during the period corresponding to the charge accumulation time T _n-1 , but it is not always necessary that the lens moves at a constant speed during the transfer / calculation time T _n ^C _-1. For example, it may be stopped during the charge transfer time. Next, the next cycle starting from time t _n will be described. At the time t _n, the defocus amount based on the data received at the time T _n-1 has already been calculated by the focus detection means 104 during the period T _n ^C _-1.
New estimated defocus amount using the contents stored in 10
X _n and the correction amount δ _n (τ) are calculated, and at the time t _n , this new X _n is set in the defocus amount counting means 106, and at the same time, the electric charge accumulation in the photoelectric conversion means 202 is restarted. It As described above, the estimated defocus amount is corrected based on δ _n (τ) with the passage of time.

Convergence driving similar to the previous time is performed until the time t ″ _n . At time t ″ _n , the lens is in focus and the estimated defocus amount in the defocus amount counting means 106 becomes exactly equal to zero. The signal generating means 105 transmits a signal for stopping driving to the driving means 102, and the lens 101 stops. In this stopped state, no movement amount signal is generated either, so that the estimated defocus amount held by the defocus amount counting means stops at zero for a while.

However, the control means 109 generates a subject movement correction pulse at a frequency according to the correction amount δ _n (τ) when the subject is moving, and this pulse counts the defocus amount according to whether the δ _n (τ) is positive or negative. The means 106 counts up or down. Therefore, the estimated defocus amount, which is the content of the means 106, is not zero, whereby the lens drive signal is generated from the drive signal generation means 105 and the lens moves,
A lens movement amount signal is generated and the defocus amount counting means 10
The estimated defocus amount of 6 becomes zero again. After a while, if the in-focus state collapses as the subject moves,
Subject movement correction pulse is generated again, the same thing is continued with until the charge accumulation end time t _'n less. Of course time t
_It is good to continue until _{n + 1} . During this period, the lens is driven along the correction amount δ _n (τ) due to the movement of the subject as shown in FIG. Even during this follow-up drive, the lens moves along the straight line 311 indicated by the broken line representing the correction amount δ _n (τ), so minute drive and stop may be included many times, but see the average. If it is moving at a uniform speed, based on the prediction calculation,
Even when the subject is moving, the lens is properly focused.

Next, a method of calculating the estimated defocus amount X _(o) and the correction amount δ (τ) will be described in detail with reference to FIG. FIG. 4 is an enlarged view of a part of FIG. When the lens movement speed is constant during the accumulation time, such as the period of T ^n-1 , it becomes a little simpler, but in FIG. 4, the most complicated case, that is, the lens driving speed changes during the accumulation time, and It tries to analyze the case in which unavoidable error is included. Time t estimated defocus amount at time t _n from the data to _n X _n and the correction amount reflecting the movement varies in time of the object δ _n (τ) is very shown. Here, if the estimated defocus amount X _n is not completely correct, the tip of X _n should be on the thick solid line 302, but in general, an error Δ is included and becomes as shown in the figure.

This error Δ is generally relatively large when the imaging lens 101 is far away from the in-focus position, but becomes smaller as it approaches the in-focus position. In the state shown in FIG. 4, since the lens 101 is very close to the in-focus position, the above-mentioned error Δ is so large that the lens 101 can be regarded as being substantially in focus at the point Q ₁ of the lens movement locus 301. It is a small value. The trajectory of movement of the lens 101 is shown by a solid line 301 and is a dashed line 3 of the estimated trajectory determined from the estimated defocus amount X _n and the correction amount δ _n (τ).
After advancing toward 10 (convergence drive), the point Q ₁ is met, and then it follows the estimated trajectory (following drive).

Further, in FIG. 4, m _n , m ' _n , m _{n + 1} are time points t _n , t' _n , t
The position of the lens 101 at _{n + 1} is shown. Also, the tip of X _n Q ₄
And the midpoint of the charge accumulation time on the thick solid line 302 Straight line l connecting the accumulation end time t as' the point of intersection with the time axis of the _n m 'shown a size including a direction to the _n X' _n
And the size including the direction from the coordinates x _n to m _{n + 1 of} this midpoint is Z _{n + 1} as shown in the figure. As described above, in this embodiment, the image plane position (x _n ) at a predetermined timing (t _n + T _n / 2)
Is expressed by using the amount (Z _{n + 1} ) measured from the lens position (m _{n + 1} ) at another predetermined timing (end of calculation). That is, Z _{n + 1} = x _n −m _{n + 1} may be obtained as a quantity measured from the lens position at another timing.

When the situation in the charge storage time is as described above, if the defocus amount calculated by the focus detection unit 104 from the image output at this time is P _{n + 1} , the following relationship is established.

The coordinate x _n of the image plane position of the midpoint is Than In addition, as the amount due to the movement of the lens (imaging optical system)
Since M _n and M _n ^C are given as M _n = m _n ′ −m _n and M _n ^C = m _{n + 1} −m _n ′, using them, the image plane at the predetermined timing can be obtained. The position Z _{n + 1} is organized as follows using the above equation with Z _{n + 1} = x _n −m _{n + 1} .

In addition, P _{n + 1} is T ′ _n , T _n
X _n are each obtained as a previous calculation result as an output of the M _n ^C, M _n is the moving amount counting means 107 as the output of the time counting means 108, which are stored in the memory 110. To summarize the concept of the processing of the prediction calculation means, first, based on the defocus amount calculated by the focus detection means as described above and the contents stored in the memory, a predetermined timing (for example, the midpoint of the charge accumulation time) ), The amount representing the position of the image plane is obtained. Next, the accumulation time T _n and the subsequent calculation time T _n
^In the repetition cycle with ^C , the image plane position at a time point (future) after the final predetermined timing is predicted using a plurality of quantities representing the image plane position at the predetermined timing. Then, the driving means drives the lens with respect to the predicted image plane position.

In this embodiment, the image plane moving speed corresponding to the inclination k of the object moving locus is obtained based on the plurality of image plane positions relating to the predetermined timing and the predetermined time interval therebetween, and the image plane accompanying the movement of the object during the predetermined time τ is obtained. The movement correction amount δ _{n + 1} (τ) is calculated to predict the future image plane position. This point will be described below.

Various methods are possible for calculating the correction amount δ _{n + 1} (τ) due to the movement of the subject. For example, this may be obtained by connecting the positions of the previous estimated defocus amount and the present estimated defocus amount.

However, in this case, there is a drawback that the error Δ for each time is accumulated. FIG. 5 is the same as FIG. 3 and shows only the parts necessary for the explanation. The correction amount δ
To obtain _{n + 1} (τ), charge accumulation time T of the past several times
Coordinates x _n , x _n-1 , x on the thick solid line in the center of _n-1 , T _n
It is better to use _n-2 ...

That is, it is preferable to use the midpoint of the charge accumulation time as the predetermined timing. In this case, the predetermined time interval is represented by dT _{n as} shown in FIG. 5, and the time from the midpoint of the charge accumulation time to the end of calculation is represented by ΔT _{n as} shown in FIG. Therefore, the image plane moving velocity corresponding to the inclination K _{n + 1} of the object moving trajectory is obtained by dividing the difference between the image plane positions x _n and x _n-1 by a predetermined time interval dT _n between them. Required by.

The image plane position Z _{n + 1} at the predetermined timing thus obtained and the image plane moving speed K _{n + 1} are used to predict the image plane position at a (future) time point after the predetermined timing. You can do it.

For example, the estimated defocus amount, which is the predicted amount of the image plane position at the end of the calculation, is given by the following equation X _{n + 1} , and the correction amount of the image plane movement after τ time has elapsed is the following equation δ _{n + 1}
It is given by (τ).

Becomes

If the subject is still or cannot be determined, it is assumed that the subject is still and the estimated defocus amount X _{n + 1} and the correction amount δ _{n + 1} (τ) at time t _{n + 1} are as follows. Is determined as.

The slope k may be obtained by weighted averaging the slopes k _n , k _n-1 , k _n-2, ... of the past several times, or the coordinates x _n , x _n-1 of the image plane position three or more times in the past. , X _n−2, ... A curve of an appropriate degree that is most suitable, for example, a quadratic curve may be obtained and the next slope k _{n + 1} may be interpolated.

Also, this way the k _{n + 1} determined>1>α> 0 becomes constant may be caused to follow a slope of .alpha.k _{n + 1} using.

From the control means 109, specifically, pulses are given to the defocus amount counting means 106 at a time interval determined from the inclination k _{n + 1} so that the lens is driven to follow along the broken line 311 in FIG.
The larger k _{n + 1} | is, the smaller the time interval is, and the pulse is sent at that time. However, it is better to send the pulse somewhat ahead so that the movement of the lens does not follow the broken line 311 in FIG.

Further, in the above description, the case where the whole lens group is fixed and moved as the lens 101 for which a simple false focus is detected is taken into consideration, but even when only one group is moved and focused, an equivalent single lens is considered as the whole group and The movement of the lens is treated as being equivalent to the movement of the lens 101.

Further, the blocks 105, 106, 107, 108, 109, 110 in FIG. 1 and the block 203 in FIG. 2 may be wholly or partially configured by a microcomputer.

Next, the overall temporal flow will be described using a flowchart. After the step operation is started (START) in FIG. 6A, the two types of interrupts described later are disabled in step and the parameter Vint is set to zero. In step, X _P corresponding to the estimated defocus amount X _O and the pulse interval T _P that determines the correction amount δ ₀ (τ) are determined. Initially, focus is detected without moving the lens, so _XP = 0, T _P = 0. However, when the value of T _{P is} a value other than zero, it indicates the pulse interval at which the control means 109 sends the object movement correction pulse to the defocus amount counting means 106, but T _P = 0 means that no pulse is generated. It shall be. Further, the defocus amount counting means 106 has T _P > 0 and T _P <
It is assumed that the counting contents are up-counted and down-counted by the pulse received corresponding to 0, respectively. In step, the content of the movement amount counting means 107 is set to zero, and the counting is started. Sets the value of X _P to the counter of the defocus amount counting means 106 in the step. When X _P ≠ 0, the drive signal generation means 105 moves the lens 101 forward or backward via the drive means 102 according to the polarity of X _P , that is, the sign.

Now if X _P = 0, so the lens 101 is stationary. Control means 109 for generating subject movement correction pulse in step
The value of T _P is set in the counter of. This counter is down-counted at a predetermined speed, and when the content of the counter becomes zero, one pulse is sent from the control means 109 to the defocus amount counting means 106, and the content of the counter is again T _P
The value of is set. In this way, a pulse is generated at each time interval T _P. In addition, the defocus amount counting means can also instruct whether the up-count is the down-count according to the polarity of T _P.
It is transmitted to 106. In this way, the value of T _P is repeatedly set in the counter, but the first pulse is shorter than T _{P at an} appropriate time so that the lens movement does not follow the straight line of δ _n (τ). It may be generated. For example, only the first value set on the counter is 0.5T _P. Here, the value of T _P is given by T _P = q / k _n with respect to k _n of Equation 5,
q is a constant determined by the discount rate of the counter, the amount of movement of the lens 101 by the movement amount signal generating means 103, and the like.

Next, in step, the counter of the time counting means 108 is set to zero and counting is started, and at the same time, charge accumulation is started in step. Then, in a step, the speed change interrupt and the charge accumulation end interrupt are enabled. A speed change interrupt is generated at the time t ″ _n at the time of T _n in FIG. 3 for the system. Therefore, the speed change interrupt is generated, for example, when the content of the defocus amount counting means 106 changes from non-zero to zero. are generated. Accordingly fly to routine Figure 6B speed change interrupt occurs. then prohibits subsequent speed change interrupt in step, in storing the contents of the counter time count means 108 to T _'n.

Next, a parameter Vint indicating that a speed change interrupt has occurred
Is set to 1 and RETURN is performed.

After that, the charge accumulation end interrupt occurs and the process moves to FIG. 6C.
First, the speed change interrupt is prohibited in step, and the content of the counter of the time counting means 108 is stored in T _n in step. This value corresponds to the charge accumulation time. After that, this counter is set to zero. When there is a speed change interrupt, it is necessary to calculate the amount corresponding to _T'n in FIG. 3, that is, the time from the generation of the speed change interrupt to the end of charge accumulation.
This is done in steps. In the step, the content of the movement amount counting means 107 is read and stored in M _n as the movement amount of the lens 101 during this period. And the counter content is set to zero again. In step S2, data is transferred from the photoelectric conversion means 202 to the focus detection calculation means 203, and the focus detection calculation means performs calculation to obtain the defocus P _{n + 1} .
At the same time as the end of the step, the lens driving is stopped in the step, and the time T _n ^C required for the data transfer and the calculation and the lens movement amount M _n ^C are calculated in the step.

The steps are these X _n , T _n , T ′ _n , T _n ^C , M _n , M _n ^C , P
Estimated defocus amount X using the formula using _{n + 1}
Calculate _{n + 1} and the correction slope k _{n + 1} . In step, these values are converted into the actual driving pulse number X _P , and the process returns to step to start the next charge accumulation period. After that, this will be repeated.

(Effects of the Invention) As is clear from the above description, according to the present invention, the movement of the target object is predicted by the prediction calculation means, and the focus of the imaging optical system is adjusted. Suitable automatic focus adjustment is possible.

[Brief description of drawings]

1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a block diagram showing a configuration example of the focus detecting means of FIG. 1, and FIG. 3 is a diagram showing movement of an imaging lens and movement of a target object. 4 is a graph showing a part of FIG. 3 in an enlarged manner, FIG. 5 is a graph for obtaining the correction amount δ (τ), FIG. 6A,
6B and 6C are flowcharts for explaining the operation of this embodiment. 101 ... Taking lens, 102 ... Driving means, 103 ... Moving amount signal generating means, 104 ... Focus detecting means, 106 ... Defocus amount counting means

Claims (4)

[Claims] 1. An imaging optical system for forming a light image of a target object, a photoelectric conversion means for photoelectrically converting a light image passing through the imaging optical system, and the target detected during an accumulation time of the photoelectric conversion means. A focus detection unit that sequentially calculates a defocus amount corresponding to a deviation amount between an imaging plane of the imaging optical system that forms an optical image of an object and a predetermined planned focal plane, and when the target object moves However, in order to perform appropriate focus adjustment, the plurality of defocus amounts calculated by the focus detection unit, the amount due to the movement of the imaging optical system, and the predetermined time are used to determine the movement of the target object. Prediction calculating means for predicting the accompanying image plane position of the image forming optical system, and driving for driving the image forming optical system based on an amount related to the image forming plane position of the image forming optical system predicted by the predictive calculating means. And means are provided. Automatic focus adjustment device. 2. The photoelectric conversion means comprises a charge storage type photoelectric conversion element, and the prediction calculation means obtains an amount relating to an image plane position of the target object detected in a charge storage time of the photoelectric conversion means, The automatic focus adjusting device according to claim 1, wherein the predictive calculation is performed based on this. 3. The prediction calculation means predicts the image formation position in a form in which a temporal change of the target object is reflected from a relationship of a plurality of quantities relating to the image plane position. Range (2) The automatic focusing device according to the item. 4. The prediction calculation means performs the prediction calculation by using a function that best matches the image plane position of the target object three or more times in the past.
An automatic focusing device according to the item.