JPH09281382A - Moving object judging device and focus adjusting device equipped with the moving object judging device - Google Patents

Abstract

The present invention relates to a moving body determination device that is mainly mounted on a camera or other optical equipment and that determines whether or not a subject is a moving body, and a focus adjustment device that includes the device. An object of the present invention is to accurately determine a moving object regardless of the detection variation of the “distance to the object” and to accurately determine a low-speed object. An image plane position detection unit that detects an image plane position of a subject image by a photographing optical system, and a detection time of a “predetermined number of image plane positions” detected by the image plane position detection unit 1 The first correlation means 2 for calculating the correlation coefficient and the magnitude of the correlation coefficient calculated by the first correlation means 2 are
It is configured to include a moving body determination unit 3 that determines “the subject is a moving body” when the value is larger than a predetermined threshold value.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is mainly mounted on a camera or other optical equipment, and a moving body determination device for determining whether or not a subject is a moving body, and a focus adjustment for switching an operation mode of focus adjustment according to the moving body determination. With the device, in particular
The present invention relates to a moving body determination device that calculates a correlation coefficient between an image plane position (or “distance to a subject” and the like) and time, and determines a moving body based on the correlation coefficient, and a focus adjustment device including the same.

[0002]

2. Description of the Related Art Conventionally, an optical device such as a camera is equipped with a focus adjusting device for automatically adjusting the focus of a photographing optical system. In these focus adjustment devices, known focus detection methods (phase difference detection method, outside light active method,
Defocus amount and "distance to subject" are detected according to the external light passive method, infrared distance measuring method, etc.).

The focus adjusting device adjusts the focus by moving the photographing optical system in correspondence with the defocus amount or the "distance to the subject". In general,
In such focus adjustment, the following two operation modes are prepared. That is, in the single AF mode,
When the focus is determined during the focus adjustment period (for example, the period when the release button is half pressed), the focus adjustment is stopped (focus lock). On the camera side, the imaging operation is usually started after waiting for this focus lock state. Such a single AF mode is mainly used for focus priority photographing.

In the continuous AF mode, the focus adjustment of the photographing optical system is continuously repeated during the focus adjustment period. In this state, the imaging operation is promptly started according to the release button operation of the photographer. Such a continuous AF mode is mainly used for photographing with priority given to a photo opportunity.

By the way, when the subject is a moving object, a period from the full depression of the release button to the image pickup operation (hereinafter, this period is referred to as "release time lag").
In addition, since the subject moves excessively, a focus shift occurs. Particularly, in the phase difference detection method, since the photographing light flux is taken in through the sub-mirror, it is impossible to obtain focus information and perform focus adjustment during the mirror-up period.

Therefore, as a kind of the above focus adjusting device, in the continuous AF mode or the single AF mode,
It is known to extend the trend of focus information before release and predict the movement of a subject between release time lags. By performing focus adjustment based on this prediction, it is possible to correct the focus shift during image pickup. Further, in general, the focus information is detected one by one through arithmetic processing and the like, so that it is data that is detected discretely and with a delay. In particular, in the phase difference detection method, the accumulation time of the focus detection CCD requires about 10 μsec to several 100 msec, and the arithmetic processing requires about several 10 msec. Therefore, the defocus amount in the phase difference detection method is at least 30 msec.
The data is detected discretely and with a delay at the above time intervals.

Since such a data delay causes a dead time in control theory, it is easy to cause insufficient or excessive (overshoot) in the follow-up operation of focus adjustment, and sufficient follow-up performance is obtained. Will be difficult. Therefore, as a kind of the above-described focus adjustment device, there is known a device that predicts the current focus information by extending the trend of past focus information in the continuous AF mode. By continuously performing the focus adjustment in accordance with this predicted value, it is possible to correct the shortage and the excessive movement that occur in the control operation of the focus adjustment, and obtain sufficient tracking performance.

By the way, for a stationary subject,
When focus adjustment based on these predictions (generally referred to as “prediction driving” below) is performed, prediction is made in the wrong direction based on the variation in the focus information, so focus adjustment is performed in the misguided direction. .

Therefore, a device for determining whether or not a subject is a moving body (hereinafter referred to as a "moving body determination device") is installed in the focus adjusting device for performing the above-described predictive driving, and the determination result of the moving body determination device is displayed. Accordingly, the suitability of the predictive drive has been determined. FIG. 13 is a flowchart showing the operation of this type of moving body determination device. First, the defocus amount D (n) detected using the known phase difference detection method is based on the conversion unique to the photographing optical system, and the remaining drive amount DP (n) of the photographing optical system.
Is converted to

Next, using the remaining drive amount DP (n) of this time, the remaining drive amount DP (n-1) of the previous time, and the moving amount M (n) of the lens position L during the period, P ( n) = DP (n) + M (n) −DP (n−1) (1) is calculated to obtain the movement amount P (n) of the image plane position F (FIG. 1).
3S41). The image plane position F here is a value representing an image plane position as an example shown in FIG. 14, and in general,
This corresponds to the image distance.

The movement amount P (n) of the image plane position F and the central storage time t (n), t when the defocus amount is detected
Using (n-1), S (n) = P (n) / {t (n) -t (n-1)} (2) is calculated, and the image plane velocity S (n) is calculated. (Fig. 13 S4
2). In this state, if the magnitude of the current image surface velocity S (n) and the magnitude of the previous image surface velocity S (n−1) continuously exceed the predetermined threshold value (FIGS. 13 S43 and S44), “ The subject is a moving body "and a moving body flag is set (S45 in FIG. 13).

On the other hand, one of the current image surface velocity S (n) and the previous image surface velocity S (n-1) is
If it falls below a predetermined threshold value (S43, S44 in FIG. 13), it is determined that the subject is a stationary body and the moving body flag is reset (S46 in FIG. 13). As described above, the moving body determination of the subject is performed based on the magnitude of the image plane velocity S (n).

Further, the present applicant has filed Japanese Patent Application Laid-Open No. 4-13301.
Japanese Patent No. 6 discloses a configuration for obtaining the image plane velocity at appropriate time intervals. For example, as shown in FIG. 15, the remaining drive amount DP (n) of this time, the remaining drive amount DP (n−2) of two generations ago, and the movement amount M2 (n) of the lens position L in the period are shown. Then, P2 (n) = DP (n) + M2 (n) −DP (n−2) (3) is calculated to obtain the movement amount P2 (n) of the image plane position F.

The moving amount P2 (n) of the image plane position F and the central storage time t (n), t when the defocus amount is detected.
By using (n-2), S2 (n) = P2 (n) / {t (n) -t (n-2)} (4) is calculated, and the image surface velocity across two generations is calculated. Find S2 (n). Based on the magnitude of this image plane velocity S2 (n), the moving object determination of the subject is performed.

In such a conventional example, the influence of the detection variation on the image plane velocity is reduced by roughly calculating the image plane velocity at appropriate time intervals (generally, time intervals of 300 msec or less). You can Therefore, the accuracy of the moving body determination is improved. Furthermore, in the focus detection method that detects the distance to the subject, such as the known external light active method and external light passive method, the distance U
(N), the previous distance U (n−1), and the detection time interval Δ
Using t and, V (n) = {U (n) −U (n−1)} / Δt (5) is calculated to obtain the subject velocity V (n). It is also possible to determine the moving body of the subject based on the magnitude of the subject speed V (n).

[0016]

By the way, in the conventional example, when detecting the defocus amount, various variations occur due to detection errors and the like.

Since the image plane velocity S (n) according to the equation (2) is calculated using the defocus amount including such variations, the image plane velocity S (n) of the stationary subject is not always zero. I won't. Therefore, in the conventional example shown in FIG. 13, when the image plane velocity S (n) calculated due to the variation exceeds the predetermined threshold value twice in succession even if the subject is a still subject, “the subject is a moving body”. There is a problem that it is erroneously determined to be "Yes".

In particular, when the brightness or contrast of the subject is low, the variation width of the defocus amount is extremely wide. Therefore, in such a subject situation, there is a problem that the frequency of erroneous motion determination is significantly increased. Such an erroneous determination can be avoided to some extent by uniformly raising the threshold value for performing the moving body determination. However, there is a problem that by increasing the threshold value for moving body determination, it becomes impossible to sufficiently determine the moving body for a subject moving at a low speed.

Further, in the conventional example disclosed in Japanese Patent Laid-Open No. 4-133016, it is possible to reduce the influence of variations included in the image plane velocity by calculating the image plane velocity at appropriate time intervals. However, the thinned defocus amount is not used at all in determining the moving body, and significant information is not used sufficiently. Therefore, there is a problem that the accuracy of the moving body determination is insufficient.

Previously, the present applicant has filed Japanese Patent Application No. 7-22394.
In the No. 6 application, an automatic focus adjustment device was applied for performing regression analysis on the image plane position in a predetermined sample section and obtaining the image plane velocity from the derived inclination β of the prediction function Qr (t). In such regression analysis, significant information in the sample section is sufficiently used, so that the “influence of detection variation” included in the image plane velocity is eliminated as much as possible. Therefore, by using the image plane velocity derived in this way, it is possible to further improve the accuracy of moving object determination as compared with the conventional example.

However, the derivation of the prediction function Qr (t) based on the regression analysis is always possible even when the defocus amount varies irregularly. Therefore, FIG.
As shown in (c), the prediction function Qr (t) having a certain degree of slope β is calculated even in a situation where the defocus amount varies for a stationary subject. Such a gradient β of the prediction function Qr (t) is meaninglessly generated due to the bias of the data in the sample section, and is unrelated to the movement of the subject.

Particularly, as the brightness or contrast of the subject decreases, the variation width of the defocus amount increases, so the slope β of the prediction function Qr (t) is calculated independently of the movement of the subject. Therefore, in such a subject situation, the moving object determination based on the inclination β of the prediction function Qr (t) has a problem that the frequency of erroneous determination increases.

Further, in these conventional examples, since the moving body determination is performed based on the threshold value determination of the image plane velocity, the image plane velocity below the threshold value is uniformly determined as the stationary body. As a result, the image surface speed of the “moving object at low speed” or the “moving object at a long distance” becomes small, and there is a problem that it is erroneously determined as a stationary object. The above-mentioned problem is not a problem that occurs only in the moving body determination using the phase difference detection method, but is a problem that generally occurs in the moving body determination using another focus detection method.

For example, in the focus detection method for detecting the distance to the object such as the external light active method, the external light passive method or the infrared distance measuring method, there is a variation included in the distance U (n) to the stationary object. In addition, the subject velocity V (n) calculated by the equation (5) does not always become zero, and there is a problem that the moving body determination is erroneous. Therefore, it is a common object of the inventions described in claims 1 to 8 to accurately determine a moving object in order to solve the above-mentioned problems.

In particular, according to the first aspect of the invention, there is provided a moving object judging device capable of accurately judging a moving object of a subject and accurately detecting a slow moving image surface regardless of variations in image plane position. The purpose is to provide. Claim 2
An object of the present invention is to provide a moving body determination apparatus that can accurately determine a moving body of a subject and can accurately detect a low-speed subject movement regardless of variation in distance measurement to the subject.

An object of the invention described in claim 3 is to provide a moving object judging device capable of stably carrying out the moving object judgment in addition to the object of claim 1 or claim 2.
According to the inventions of claims 4 and 5, together with the object of claim 1 or claim 2, while performing the moving body determination of the subject in a stable manner, the moving body determination is performed on the subject or the image plane position that is nonlinearly moving. It is an object of the present invention to provide a moving body determination device that can be accurately and reliably performed.

According to the invention described in claim 6, in addition to the object of claim 4 or claim 5, it is an object of the invention to provide a moving object judging device capable of quickly judging a moving object for an object to be accelerated or decelerated. And In the invention described in claim 7, in addition to the object of claim 1, it is an object of the invention to provide a particularly suitable moving object determination device when applied to a device for performing focus detection of a phase difference detection method.

In the invention described in claim 8, claims 1 to 7 are provided.
It is an object of the present invention to provide a focus adjustment device including the moving body determination device according to any one of 1.

[0029]

FIG. 1 is a principle block diagram for explaining the invention described in claim 1. In FIG.

The moving object determining apparatus according to the first aspect of the present invention comprises an image plane position detecting means 1 for detecting the image plane position of a subject image by the photographing optical system, and a "predetermined number of samples" detected by the image plane position detecting means 1. For the “image plane position”, the first correlation means 2 for calculating the correlation coefficient with the detection time, and the magnitude of the correlation coefficient calculated by the first correlation means 2 are larger than a predetermined threshold value. It is configured to include a moving body determination unit 3 that determines that “the subject is a moving body” at times.

FIG. 2 is a principle block diagram for explaining the invention described in claim 2. The moving object determination apparatus according to claim 2, wherein the object distance measuring unit 4 detects a distance to the object,
For the "distance to a subject of a predetermined number of samples" detected by the subject distance measuring unit 4, a second correlation unit 5 that calculates a correlation coefficient with the detection time, and a phase calculated by the second correlation unit 5 When the magnitude of the number of relations is larger than a predetermined threshold value, the moving body determination unit 3 that determines that the subject is a moving body is configured.

According to a third aspect of the present invention, there is provided the moving body determining apparatus according to the first or second aspect, wherein the moving body determining means 3 adds a hysteresis characteristic to a threshold value or a correlation coefficient to obtain a phase relationship. It is characterized in that a number threshold determination is performed. The moving body determining apparatus according to claim 4 is the moving body determining apparatus according to claim 1 or 2, wherein the moving body determining unit 3 determines whether or not the subject is a moving body according to the result of matching threshold determinations over a predetermined number of times. It is characterized by determining whether or not.

According to a fifth aspect of the present invention, there is provided the moving body determination apparatus according to the first or second aspect, wherein the moving body determination means 3 determines that the subject is a moving body in accordance with the result of matching threshold determinations over a predetermined time. It is characterized by determining whether or not there is. The moving body determination device according to claim 6 is the moving body determination device according to claim 4 or 5, wherein in the moving body determination means 3, the threshold value to be compared with the correlation coefficient depends on the generation of the correlation coefficient. Are different.

FIG. 3 is a principle block diagram for explaining the invention described in claim 7. According to a seventh aspect of the present invention, in the moving body determination apparatus according to the first aspect, the image plane position detection means 1 spatially determines a set of optical images obtained by dividing the transmitted light flux of the photographing optical system. Defocus amount detecting means 6 for detecting a phase difference between the subject image and the image pickup surface based on the detected phase difference.
, A lens position detecting means 7 for detecting the lens position of the photographing optical system, the defocus amount and the lens position, and
Image plane position calculating means 8 for calculating the image plane position of the subject image, and the first correlating means 2 comprises the image plane position calculating means 8
For the "image plane position of a predetermined number of samples" calculated by
The defocus amount detecting means 6 is characterized by calculating a correlation coefficient with "a central time of accumulation of a set of optical images".

FIG. 4 is a principle block diagram for explaining the invention described in claim 8. The focus adjustment device according to claim 8 comprises the moving object determination device 11 according to any one of claims 1 to 7, and extends the movement of the image plane position or the “distance to the subject” to obtain the subject. Predicting means 12 for predicting the movement of the subject, predicting driving means 13 for predictively driving the photographing optical system to the in-focus position in accordance with the "movement of the subject" predicted by the predicting means 12, and "moving the subject as a moving body" by the moving body determining means 3. If it is determined that the prediction driving is performed, the prediction driving unit 13 performs the prediction driving, and in other cases, the prediction driving unit 13 prohibits the prediction driving.

(Operation) In the moving object judging apparatus according to the first aspect, the image plane position detecting means 1 detects the image plane position of the subject image by the photographing optical system. The first correlation means 2 is
A correlation coefficient is calculated for the “image plane position of a predetermined number of samples” and its detection time.

For example, the correlation coefficient ρ is calculated as follows based on the image plane position F (k) for n times and the detection time t (k). First, using the detection times t (k) and the image plane position F (k) for n times,

[Equation 1]

[Equation 2] Is calculated, and the average value tave at the detection time and the average value Fave at the image plane position are obtained.

Next, using these values,

(Equation 3)

(Equation 4)

(Equation 5) Were calculated, obtaining a dispersion S1 ² of the detection time, the variance S2 ² position of the image plane, and both the covariance S12.

Using these variances, ρ = S12 / (S1 · S2) (11) is calculated, and the phase relationship between the detection time t (k) and the image plane position F (k) is calculated. Find the number ρ. The correlation coefficient ρ thus derived is a scale for measuring the linear relationship between the detection time and the image plane position, and takes a value from -1 to 1.

That is, as shown in FIG.
When the image plane position is concentrated around the primary diagram having a positive inclination, the value of the correlation coefficient ρ approaches “1”. Also, FIG.
As shown in (b), when the image plane position is concentrated around the primary diagram having a negative slope, the value of the correlation coefficient ρ becomes “−”.
Approaches 1 ”. Furthermore, as shown in FIG.
If the image plane positions are distributed away from the primary diagram or concentrated around the primary diagram with a slope of "0", the value of the correlation coefficient ρ approaches "0".

Therefore, when the magnitude | ρ | of the correlation coefficient is close to "1", it is estimated that the image plane position is displaced according to the detection time, so that "the subject is a moving object" is accurately determined. I can judge. On the other hand, when the magnitude of the correlation coefficient | ρ | is close to “0”, it means that the image plane position has a substantially constant value, or that the image plane position fluctuates regardless of the detection time. In both cases, it can be accurately determined that the subject is a stationary body.

Therefore, the moving body determining means 3 makes a threshold determination of the magnitude | ρ | of the correlation coefficient to accurately determine "whether or not the subject is a moving body". When the variance S2 ² of the image plane positions is “0”, the denominator of the equation (11) is zero, so that the correlation coefficient ρ cannot be calculated. However, the image plane position variance S2 ² is "0".
The case where the image plane position takes a constant value in the sample section is an extremely rare case in consideration of the detection variation of the image plane position. Therefore, such a case may be overlooked.

In this case, since the image plane position has a constant value, it may be definitely determined that "the subject is a stationary body". In the moving body determination device according to the second aspect, the distance to the subject is detected by the subject distance measuring unit 4. The second correlation means 5 calculates a correlation coefficient for a predetermined sample number of “distance to subject” and the detection time.

The correlation coefficient thus derived is a scale for measuring the linear relationship between the detection time and the "distance to the subject", and takes a value from -1 to 1. That is, when the “distance to the subject” is concentrated on the first-order time function having a positive slope, the value of the correlation coefficient approaches “1”. Further, when the “distance to the subject” is concentrated on a first-order time function having a negative slope, the value of the correlation coefficient approaches “−1”.

Further, if the "distance to the subject" is distributed away from the linear time function or concentrated around a constant value, the value of the correlation coefficient ρ approaches "0". Therefore, when the magnitude of the correlation coefficient | ρ | is close to “1”, it is estimated that the “distance to the subject” is changing every moment according to the detection time, and “the subject is a moving body”.
Can be accurately determined.

On the other hand, when the magnitude | ρ | of the correlation coefficient is close to "0", the "distance to the subject" takes a substantially constant value, or the "distance to the subject" is irrespective of the detection time. Is a case in which the subject is stationary, and in both cases, it can be accurately determined that the subject is a stationary body. Therefore, the moving body determination unit 3 accurately determines "whether or not the subject is a moving body" by performing a threshold determination on the magnitude of the correlation coefficient.

When the variance of the "distance to the subject" is "0", the denominator side of the correlation coefficient is zero, so that the correlation coefficient cannot be calculated. However,
The case where the variance of the "distance to the subject" is "zero" is when the "distance to the subject" is constant in the sample section, which is extremely rare considering the detection variation of the "distance to the subject". That is the case. Therefore, such a case may be overlooked.

Further, in such a case, the "distance to the subject" has a constant value, and therefore, "the subject is a stationary body" may be definitely determined. In the moving body determination device according to the third aspect, the moving body determination means 3 adds hysteresis to the threshold determination. Therefore, after the correlation coefficient increases and it is once determined that the subject is a moving body, the determination that the subject is a moving body continues until the correlation coefficient decreases by a predetermined width.

After the correlation coefficient decreases and it is once determined that the subject is a stationary body, the determination that the subject is a stationary body is made until the correlation coefficient increases by a predetermined width. continue. Therefore, the result of the moving object determination does not frequently change due to a slight change in the correlation coefficient, and a stable moving object determination can be realized.

In the moving object judging apparatus according to the fourth aspect, the moving object judgment is performed according to the result of coincidence of the threshold judgments over a predetermined number of times. Therefore, the result of the moving body determination does not change due to the sudden change of the correlation coefficient, and the accuracy of the moving body determination can be further improved. Such an effect can also be achieved by increasing the number of samples of the image plane position or the “distance to the subject”.
The configuration described in (1) has the advantage that the storage capacity for storing the sample data can be reduced because only the history of the correlation coefficient or the determination result needs to be accumulated a predetermined number of times.

If the number of samples of the image plane position or the "distance to the subject" is simply increased, the locus of the image plane position or the "distance to the subject" will be "terms other than the first order of time".
May appear and the linearity may deteriorate. In such a state, even though there is a causal relationship between the image plane position (or “distance to the subject”) and the detection time, the correlation coefficient is calculated to be low, and thus the moving body determination is erroneous, There is a problem that the normal moving body determination is delayed.

However, in the configuration according to the fourth aspect, since the sample section is divided into a predetermined number of times (the sample sections may be divided so as to overlap) and the correlation coefficient is detected, each short sample is detected. Within the section, the “linear term other than the first-order of time” does not appear largely on the locus of the image plane position or the “distance to the subject”, and a good linear approximation is maintained.

Therefore, even for a subject whose image plane position or "distance to the subject" changes non-linearly, the correlation coefficient is calculated by dividing the correlation coefficient a predetermined number of times.
The causal relationship between the image plane position (or “distance to the subject”) and the detection time can be reliably detected, and accurate moving body determination can be performed. In the moving body determination device according to the fifth aspect, the moving body determination is performed according to the result of coincidence of the threshold determinations over a predetermined time.

Therefore, the result of the moving body determination does not change due to the sudden change of the correlation coefficient, and the accuracy of the moving body determination can be further improved. Such an effect can also be achieved by increasing the number of samples of the image plane position or the “distance to the subject”, but in the configuration according to claim 5, only the history of the correlation coefficient or the determination result is determined for a predetermined time. Since it is sufficient to store only the amount, there is an advantage that a large storage capacity is not required.

If the number of samples at the image plane position (or "distance to subject") is simply increased, "terms other than the first order of time" appear on the locus of the image plane position or "distance to subject". Therefore, the linearity may be reduced. In such a state, even though there is a causal relationship between the image plane position (or “distance to the subject”) and the detection time, the correlation coefficient is calculated to be low, and thus the moving body determination is erroneous, There is a problem that the normal moving body determination is delayed.

However, in the structure according to the fifth aspect, since a plurality of sample sections are set within a predetermined time, in each short sample section, the image plane position or the "distance to the subject" is indicated by "time". "Terms other than the first-order of" do not appear significantly, and linearity is not lost. Therefore, even for a subject in which the image plane position or the “distance to the subject” changes non-linearly, the image plane position (or “distance to the subject”) and the detection time can be determined by performing the threshold determination for a predetermined time. It is possible to reliably detect the causal relationship between them and accurately determine the moving body.

According to a sixth aspect of the present invention, there is provided a moving body determining apparatus, wherein when the threshold determination is performed a predetermined number of times or a predetermined time, the threshold is changed according to the generation of the correlation coefficient. In the moving body determination device according to claim 7,
The configuration when the phase difference detection method is applied is shown.

First, the defocus amount detection means 6 detects a spatial phase difference between a pair of optical images obtained by dividing and forming a transmitted light flux of the photographing optical system, and based on the phase difference, "the subject image and the image are picked up. The amount of defocus corresponding to "the distance from the surface" is detected. On the other hand, the lens position detecting means 7 detects the lens position of the photographing optical system. Next, the image plane position calculating unit 8 synthesizes the defocus amount and the lens position detected in this way to obtain the image plane position of the subject image.

The first correlation means 2 is the image plane position calculation means 8
For the "image plane position of a predetermined number of samples" calculated by
The focus adjusting device according to claim 8 which calculates a correlation coefficient with the "accumulation center time of a set of optical images" in the defocus amount detecting means 6, according to any one of claims 1 to 7. According to the moving body determination by the moving body determination device 11, the propriety of predictive driving is determined.

[0060]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 6 is a diagram showing a first embodiment (corresponding to claims 1, 3, 7, and 8).
In FIG. 6, a photographing optical system 22 is attached to the camera 21, and a quick return mirror 23 and a sub mirror 24 are sequentially arranged along the optical axis of the photographing optical system 22.

A focus detection unit 25 of the TTL phase difference detection system is arranged in the reflection direction of the sub mirror 24, and an output of the focus detection unit 25 is connected to a microprocessor 26. The PWM (pulse width modulation) output of the microprocessor 26 is connected to the motor 28 via the motor drive circuit 27, and the power of the motor 28 is transmitted to the lens drive mechanism 29 on the photographing optical system 22 side.

An encoder 30 is arranged on the drive shaft of the motor 28 to detect the amount of rotation, and the pulse output of the encoder 30 is connected to the microprocessor 26. A lens information storage unit 22a that stores individual lens information is arranged inside the photographing optical system 22, and the data output of the lens information storage unit 22a is connected to the microprocessor 26 via the lens mount unit. .

Regarding the correspondence relationship between the inventions described in claims 1 and 3 and the first embodiment, the image plane position detecting means 1 includes the "image plane position" in the focus detecting section 25, the encoder 30 and the microprocessor 26. The first correlation means 2 corresponds to the "correlation coefficient calculation function" in the microprocessor 26, and the moving body determination means 3 corresponds to the "threshold determination function" in the microprocessor 26. To do.

Regarding the correspondence relationship between the invention according to claim 7 and the first embodiment, the defocus amount detecting means 6 corresponds to the "defocus amount calculating function" in the focus detecting section 25 and the microprocessor 26. The lens position detecting means 7 corresponds to the encoder 30, and the image plane position calculating means 8 corresponds to the "image plane position calculating function" in the microprocessor 26.

Regarding the correspondence between the invention according to claim 8 and the first embodiment, the prediction means 12 corresponds to the "regression analysis function" in the microprocessor 26, and the prediction drive means 13 is the motor drive circuit. 27, motor 28, lens drive mechanism 29, encoder 30 and microprocessor 2
6 corresponds to the "driving amount calculation function", and the selection means 14 corresponds to the "operation mode switching function" in the microprocessor 26.

FIG. 7 is a flow chart showing the operation of the first embodiment. The operation of the first embodiment will be described below with reference to these drawings. First, the light flux transmitted through the photographing optical system 22 is
The light enters the focus detection unit 25 via the sub mirror 24.
In the focus detection unit 25, this incident light flux is divided into images,
On the light receiving surface of the focus detection CCD inside the focus detection unit 25,
Form a set of light images.

When the release button (not shown) is pressed halfway in this state, the focus detection CC
Photoelectric charge accumulation of D is started (S1 in FIG. 7). The microprocessor 26 detects the lens position L (k) at the photocharge accumulation center time t (k) by sequentially counting the pulse outputs of the encoder 30. Next, the microprocessor 26 takes in the transfer output of the focus detection CCD and forms "a set of image patterns of light images" on the internal memory.

Here, the microprocessor 26 detects the phase difference in space between the pair of optical images and calculates the defocus amount D (k) based on the phase difference (FIG. 7S).
2). Next, the microprocessor 26 calculates DP (k) = DP (k) = based on the defocus amount D (k) and the “conversion coefficient LD between the defocus amount and the remaining drive amount” obtained from the lens information storage unit 22a. LD · D (k) ... (12) is calculated to obtain the remaining drive amount DP (k) of the motor 28 (S3 in FIG. 7).

The microprocessor 26 calculates F '(k) = L (k) + DP (k) (13) based on the remaining drive amount DP (k) and the lens position L (k). Then, the focus position F ′ (k) is calculated (S4 in FIG. 7).
The in-focus position F ′ (k) represents a target position where the photographing optical system 22 should be extended at the central storage time t (k) in order to obtain the in-focus state.

As shown in FIG. 14, this focusing position F '
By shifting the value of (k) by an appropriate offset C, F (k) = F ′ (k) + C (14) The image plane position F (k) is obtained. Here, the offset C is a constant interval determined by the measurement origin Co of the lens position L (k), and in the following statistical calculation, (for example,
It can be ignored (as C = 0).

Next, the microprocessor 26 uses the past n
Regarding the image plane position F (k) for each batch (preferably about 10 batches) and the central storage time t (k), the above equation (6)-
Equation (10) is calculated. Using the variance value calculated in this way, the microprocessor 26 calculates β = S12 / S1 ² (15) α = Fave −β · tave (16) and the image plane position F The undetermined coefficient of the prediction function Qr (t) of (k) is obtained (FIG. 7 S5).

Qr (t) = α + β · t (17) Further, the microprocessor 26 calculates ρ = S12 / (S1 · S2) by using the above calculated value, and the image plane position F ( k) and the correlation center ρ between the central storage time t (k) are obtained (S6 in FIG. 7).

At this point, if the prediction drive is not executed (S7 in FIG. 7), the microprocessor 26
Compares the magnitude | ρ | of the correlation coefficient with the threshold ρth1 (S8 in FIG. 7). For example, the threshold ρth1 is set to about “0.8”. Here, the magnitude of the correlation coefficient | ρ |
When it exceeds th1, according to the central storage time t (k),
Since it is estimated that the image plane position F (k) is displaced,
It can be determined that “the subject is a moving body”. Therefore, the microprocessor 26 starts the prediction driving shown in FIG. 8 based on the prediction function Qr (t) shown in the expression (17) (S9 in FIG. 7).

On the other hand, the magnitude of the correlation coefficient | ρ |
In the case of 1 or less, the image plane position F (k) has a substantially constant value, or the image plane position fluctuates regardless of the central storage time t (k). Can also be determined as “the subject is a stationary body”. Therefore, the microprocessor 26 (17)
The image plane velocity β in the equation is set to zero (FIG. 7S10), and the normal lens drive is continued without performing the predictive drive (FIG. 7S1).
1).

At this point, if the prediction drive is in progress (FIG. 7S7), the microprocessor 26
The magnitude | ρ | of the correlation coefficient is compared with the threshold ρth2 (Fig. 7).
S12). This threshold ρth2 is set to a value less than the above-mentioned threshold ρth1 in order to provide a hysteresis characteristic for threshold determination. Here, the magnitude of the correlation coefficient | ρ |
If it exceeds, it is estimated that the image plane position F (k) is displaced according to the central storage time t (k), and therefore it can be determined that “the subject is still a moving body”.
Therefore, the microprocessor 26 continues the predictive drive shown in FIG. 8 based on the predictive function Qr (t) shown in Expression (17) (S13 in FIG. 7).

On the other hand, the magnitude of correlation coefficient | ρ |
In the case of 2 or less, either the image plane position F (k) has a substantially constant value or the image plane position fluctuates regardless of the central storage time t (k). Can also be determined as “the subject has been changed to a stationary body”. Therefore, the microprocessor 26
The image plane velocity β in equation (17) is set to zero (S10 in FIG. 7),
The normal lens drive is started without performing the predictive drive (S11 in FIG. 7).

By the operation described above, in the first embodiment, the correlation coefficient ρ is calculated for the image plane position F (k) and the central storage time t (k), and the moving object determination based on the correlation coefficient ρ is performed. Is done. Therefore, there is no risk of erroneously determining that the stationary subject is a moving body due to variations in detection of the image plane position F (k), and the moving body can be accurately determined.

Further, since the magnitude of the correlation coefficient | ρ | takes a value from 0 to 1 regardless of the speed of the image plane velocity, the "object having a high image plane velocity" and the "image plane velocity" The moving body determination based on the same threshold value can be performed for both the “subject with slow motion”. In addition, since the moving body determination is performed based on the magnitude of the correlation coefficient ρ, it is possible to accurately determine that the subject is a moving body even for a subject having a low image plane velocity.

Furthermore, since the hysteresis characteristic is added to the threshold value determination, the result of the moving object determination does not frequently change due to a slight change in the correlation coefficient ρ, and stable moving object determination can be realized. Next, another embodiment will be described. FIG. 9 shows a second embodiment (claims 1, 3,).
4 (corresponding to Nos. 4, 7, and 8).

Regarding the configuration of the second embodiment,
Except for the difference in the internal function of the microprocessor 26, the configuration is the same as that of the first embodiment shown in FIG. 6, and therefore the description thereof is omitted here. The operational features of the second embodiment will be described below. First, when the release button (not shown) is half-pressed, the photoelectric charge accumulation in the focus detection CCD is started in the focus detection unit 25 (S1 in FIG. 9).

The microprocessor 26 uses the encoder 3
The lens position L (k) at the central time t (k) of accumulation of photocharges is detected via 0. Next, the microprocessor 26 calculates the defocus amount D (k) by a conventional method based on the transfer output of the focus detection CCD (FIG. 9S).
2). Next, the microprocessor 26 multiplies the defocus amount D (k) by the "conversion coefficient LD of the defocus amount and the remaining drive amount" to calculate the remaining drive amount DP (k) of the motor 28.
Is calculated (S3 in FIG. 9).

The microprocessor 26 determines the remaining drive amount DP
(K) and the lens position L (k) are combined to obtain a focus position F ′ (k) (S4 in FIG. 9). This focusing position F '
Using (k) as the image plane position F (k), the following statistical calculation is performed. First, the microprocessor 26 performs the prediction function Qr (t) on the image plane position F (k) and the accumulation center time t (k) for the past n times, similarly to the first embodiment.
And the correlation coefficient ρ is obtained (S5 and S6 in FIG. 9).

At this point, when the predictive driving is not executed (S7 in FIG. 9), the microprocessor 26
Compares the magnitude | ρ | of the correlation coefficient with the threshold value ρth1 (S8 in FIG. 9). Here, when the magnitude of the correlation coefficient | ρ | is less than or equal to the threshold ρth1, the image plane position F (k) has a substantially constant value, or the image plane position F (k) is independent of the accumulation center time t (k). This is a case in which the position is changing, and in both cases, it can be determined that the subject is a stationary body. Therefore, the microprocessor 26 sets the image plane velocity β to zero (FIG. 9S9), and then continues the normal lens driving without performing the predictive driving (FIG. 9S10).

On the other hand, the magnitude of the correlation coefficient | ρ |
If it exceeds 1, it is estimated that the image plane position F (k) is displaced in accordance with the central storage time t (k), and therefore it can be determined that the “object is a moving body”. Therefore,
After initializing the internal counter COUNT to zero (S11 in FIG. 9), the microprocessor 26 makes a threshold determination on the history of the image plane velocity β, and carefully considers whether or not the image plane is continuously moving in the same direction. The judgment is made (S12 in FIG. 9).

When it is determined that the subject is a moving object by such a history test, the predictive drive based on the predictive function Qr (t) is started (S13 in FIG. 9). If the history test cannot determine that the subject is a moving object, the normal lens drive is continued (S10 in FIG. 9). Further, at this time point, when the predictive driving is being executed (S9 in FIG. 9), the microprocessor 26 compares the magnitude | ρ | of the correlation coefficient with the threshold ρth2 (S14 in FIG. 9). This threshold ρth2 is set to a value less than the above-mentioned threshold ρth1 in order to provide a hysteresis characteristic for threshold determination. Here, when the magnitude of the correlation coefficient | ρ | exceeds the threshold ρth2, the central storage time is set. Image plane position F according to t (k)
Since it is estimated that (k) is displaced, it can be determined that “the subject is still a moving body”. Therefore, the microprocessor 26 initializes the internal counter COUNT to zero (FIG. 9S15), and then uses the latest prediction function Q.
Predictive drive is continued based on r (t) (FIG. 9S1).
6).

On the other hand, the magnitude of the correlation coefficient | ρ |
In the case of 2 or less, either the image plane position F (k) has a substantially constant value or the image plane position fluctuates regardless of the central storage time t (k). Can also be determined as “the subject has been changed to a stationary body”. After confirming that such a judgment result is continuously obtained a predetermined number of times m times (for example, about 2 to 10 times) (FIGS. 9S17 and S18), the microprocessor 2
6 sets the image plane velocity β to zero (FIG. 9S19), interrupts the predictive drive, and starts normal lens drive (FIG. 9S2).
0).

With the operation described above, the same effects as those of the first embodiment can be obtained in the second embodiment. Next, another embodiment will be described. FIG.
It is a flowchart which shows operation | movement of 3rd Embodiment (corresponding to Claim 1, 3, 4, 5, 6, 7, 8).

Regarding the configuration of the third embodiment,
Except for the difference in the internal function of the microprocessor 26, the configuration is the same as that of the first embodiment shown in FIG. 6, and therefore the description thereof is omitted here. The operational features of the third embodiment will be described below. First, when a release button (not shown) is pressed halfway, photocharge accumulation of the focus detection CCD is started in the focus detection unit 25 (FIG. 10S1).

The microprocessor 26 uses the encoder 3
The lens position L (k) at the central time t (k) of accumulation of photocharges is detected via 0. Next, the microprocessor 26 calculates the defocus amount D (k) based on the transfer output of the focus detection CCD (FIG. 10S2).

Next, the microprocessor 26 multiplies the defocus amount D (k) by the "conversion coefficient LD between the defocus amount and the remaining drive amount" to obtain the remaining drive amount DP of the motor 28.
(K) is calculated (S3 in FIG. 10). Microprocessor 2
Reference numeral 6 synthesizes the remaining drive amount DP (k) and the lens position L (k) to obtain a focus position F ′ (k) (FIG. 10S4).

This focusing position F '(k) is set to the image plane position F.
As (k), the following statistical calculation is performed. First, the microprocessor 26 determines the image plane position F (k) for the past n times.
With respect to the accumulation central time t (k), the prediction function Qr (t) and the correlation coefficient ρ are obtained in the same manner as in the first embodiment (FIGS. 10S5 and S6).

At this time, if the prediction drive is not executed (S7 in FIG. 10), the microprocessor 26
Performs threshold determination on the correlation coefficient of the past m times (for example, about 2 to 10 times) (FIG. 10S8). Where past m
When the magnitude of the correlation coefficient for each batch exceeds the individual threshold value, it can be reliably determined that the “subject is a moving body”. Therefore, the microprocessor 26 starts the prediction drive based on the prediction function Qr (t) (FIG. 10S).
9).

On the other hand, if at least one of the magnitudes of the correlation coefficients for the past m times is less than or equal to the threshold value, it can be cautiously determined that "the subject is a stationary object". Therefore, the microprocessor 26 continues the normal lens drive without performing the predictive drive (FIG. 10S10). Further, at this point, when the predictive driving is being executed (FIG. 10S7), the microprocessor 26 compares the magnitude | ρ | of the correlation coefficient with the threshold ρth2 (FIG. 10S11).

Here, the magnitude of the correlation coefficient | ρ |
When it exceeds th2, according to the central storage time t (k),
Since it is estimated that the image plane position F (k) is continuously displaced, it can be determined that “the subject is still a moving body”. Therefore, the microprocessor 26 continues the prediction drive based on the latest prediction function Qr (t) (FIG. 10S12).

On the other hand, the magnitude of the correlation coefficient | ρ |
In the case of 2 or less, the threshold value determination is performed on the correlation coefficient for the past m times (FIG. 10S13). Here, when the magnitudes of the correlation coefficients for the past m times are less than the respective threshold values, it can be reliably determined that “the subject is a stationary body”. Therefore, the microprocessor 26 sets the image plane velocity β to zero (FIG. 10S14) and suspends the prediction drive based on the prediction function Qr (t) (FIG. 10S15).

On the other hand, in the case where at least one of the magnitudes of the correlation coefficient for the past m times exceeds the threshold value, another subject may cross the shooting angle of view for a moment. Therefore, the microprocessor 26 continues the prediction drive while maintaining the previous prediction function Qr (t). With the operation described above, the third embodiment is
The same effect as that of the embodiment can be obtained.

Further, as an effect peculiar to the third embodiment, since the moving object determination is performed based on a plurality of "threshold determinations of the correlation coefficient ρ", a sudden variation of the correlation coefficient ρ causes There is no influence on the result of motion determination,
The accuracy of motion determination can be further improved. Also, since the sample interval is divided into multiple sections and the correlation coefficient is detected individually, the individual correlation coefficient can be detected with sufficient linear approximation even for an object whose image plane velocity changes momentarily. can do.

Therefore, it is possible to accurately and surely determine a moving object even for a subject whose image plane position changes non-linearly. Furthermore, when the threshold value determination is performed a predetermined number of times, the threshold value is changed according to the generation of the correlation coefficient ρ, so that the moving object determination can be freely performed according to the history of the correlation coefficient ρ.

For example, when the threshold value is gradually lowered as the generation of the correlation coefficient ρ becomes newer, it is possible to quickly detect a subject that has started to move rapidly as a moving body. On the contrary, when the threshold value is gradually increased as the generation of the correlation coefficient ρ becomes newer, a subject that stops rapidly can be quickly detected as a stationary body. Next, another embodiment will be described.

FIG. 11 shows a fourth embodiment (claims 2 and 5).
FIG. In FIG. 11, a photographing optical system 42 is arranged on the front surface of the camera body 41, and a release 43 is arranged on the upper surface (warship section). Release 4
The output of 3 is connected to the microprocessor 44, and the output of the microprocessor 44 is connected to the light emitting element 46 via the drive circuit 45.

Light emitted from the light emitting element 46 is emitted from the camera body 41.
The light is projected to the subject side through a light transmitting lens 47 arranged in front of the subject. The reflected light from the subject is received by the light receiving element 49 via the light transmitting lens 47 and the light receiving lens 48 which is arranged at a predetermined base line distance. The output signal of the light receiving element 49 is sent to the microprocessor 44 via the light receiving circuit 50.
Is input to

The PWM output of the microprocessor 44 is connected to the feeding mechanism 51 of the photographing optical system 42. Regarding the correspondence relationship between the invention according to claims 2 and 5 and the fourth embodiment, the object distance measuring means 4 includes a light emitting element 46, a light transmitting lens 47, a light receiving lens 48, and a light receiving element 49.
And the second correlating means 5 corresponds to the "correlation coefficient calculation function" in the microprocessor 44, and the moving body determining means 3 corresponds to the microprocessor 44. It corresponds to the "threshold judgment function" in.

FIG. 12 is a flow chart showing the operation of the fourth embodiment. The operation of the fourth embodiment will be described below with reference to these drawings. First, when the release 43 is half-pressed (FIG. 12S1), the microprocessor 44 turns on the light emitting element 46. The emitted light that has been projected is reflected by the subject and returns to the camera body 41. The light receiving lens 48 collects this reflected light and forms a bright spot on the light receiving surface of the light receiving element 49.

The microprocessor 44 includes a light receiving element 49.
The light receiving position of the bright spot is taken from (S2 in FIG. 12). The microprocessor 44 performs triangulation calculation using the light receiving position and the base line length to calculate the distance U (k) to the subject (FIGS. 12S3 and S4). Microprocessor 44
Moves the photographing optical system 42 to the in-focus position based on the distance U (k) to the subject (S5 in FIG. 12).

Such an operation is repeated until the number of samples of the distance U (k) exceeds n (S6 in FIG. 12). Here, when the number of samples of the distance U (k) exceeds n, the past n times (for example, several tens of times) of the distance U (k) and its detection time t (k) are

(Equation 6)

(Equation 7) And the average value Uave of the distance and the average value t of the detection time.
ask for ave.

Next, using these values,

(Equation 8)

[Equation 9]

[Equation 10]And the variance S1 of the detection time t (k)^Two When,
Variance S2 of distance U (k) ^Two And the covariance S12 of both
Ask.

Using these variances, ρ = S12 / (S1 · S2) (11) is calculated, and the correlation coefficient ρ between the detection time t (k) and the distance U (k) is calculated. Is obtained (S12 in FIG. 12). Here, when the magnitude | ρ | of the correlation coefficient is equal to or smaller than the threshold ρth (FIG. 12S8),
Whether the distance U (k) is a substantially constant value or the detection time t (k)
Since it is estimated that the distance U (k) fluctuates irrespective of, it can be determined that “the subject is a stationary body”.

Therefore, the microprocessor 44 resets the correlation flag and the moving object flag (FIG. 12S).
9, S10), so that the focus is locked when focusing.
Switch to single AF mode. On the other hand, when the magnitude of the correlation coefficient | ρ | exceeds the threshold ρth (see FIG. 12S
8) The microprocessor 44 determines whether the correlation flag is in the set state.

If the correlation flag is reset, the correlation flag is set and the microprocessor 4
The timer T set inside 4 is initialized to start time counting (FIG. 12 S13). At this point, there may be a case where the correlation coefficient ρ suddenly rises. Therefore, the moving body flag is reset and not immediately determined without performing the moving body determination.
10) Maintain the single AF mode.

On the other hand, when the magnitude | ρ | of the correlation coefficient exceeds the threshold value ρth over a predetermined time Tth (S1 in FIG. 12).
2, S14), since it can be estimated that the subject is displaced according to the detection time t (k), it can be determined that "the subject is a moving body". Therefore, the microprocessor 4
4 sets the moving object flag (FIG. 12S15) and then switches to the continuous AF mode in order to continue the focus adjustment following the moving object (FIG. 12S16).

By the operation described above, in the fourth embodiment, the correlation coefficient ρ is calculated for the distance U (k) and the detection time t (k), and the moving body determination is performed based on the correlation coefficient ρ. . Therefore, there is no risk of erroneously determining that the stationary subject is a moving object due to variations in the detection of the distance U (k), and the moving object can be accurately determined.

Furthermore, since the magnitude of the correlation coefficient | ρ | takes a value from 0 to 1 regardless of the speed of the object speed, the "object with a high object speed" and the object with a "low object speed" are For both "and", the moving object determination based on the same threshold can be performed. Further, since the moving body determination is performed based on the magnitude of the correlation coefficient ρ, it is possible to accurately determine that the subject is a moving body even for a subject having a low subject velocity.

Furthermore, the "correlation coefficient ρ over the predetermined time T
Since the moving body determination is performed based on the “threshold determination of”, the result of the moving body determination is not influenced by the sudden change of the correlation coefficient ρ, and the accuracy of the moving body determination can be further improved. Further, since the correlation coefficient is individually detected by dividing the sample section into a plurality of sections, the individual correlation coefficient is detected by a sufficiently linear approximation even for a subject whose subject velocity changes every moment. be able to.

Therefore, it is possible to accurately and surely determine the moving object even for a subject whose position changes non-linearly. It should be noted that in the above-described embodiment, the magnitude of the correlation coefficient ρ is determined to deterministically determine whether or not the object is a moving object, but the present invention is not limited to that, and is gentle before and after the threshold value. By providing such an inclination (membership function), the “degree to which a subject can be determined to be a moving body” may be fuzzyly obtained.

Further, in the above-described embodiment, the absolute value | ρ | of the correlation coefficient is determined by the threshold value, but the invention is not limited to this, and the magnitude of the correlation coefficient ρ may be determined.
For example, the square value ρ ² (that is, the contribution rate) of the correlation coefficient may be used as the threshold value. In such a configuration, since ρ ² = S12 ² / (S1 ² · S2 ² ) (23) may be calculated, it is necessary to calculate the square root (that is, standard deviation) of the variances S1 ² and S2 ^2. Therefore, there is an advantage that the calculation processing can be reduced.

[0116]

As described above, in the moving body determining apparatus according to the first aspect, the correlation coefficient is calculated for the image plane position and the detection time, and the moving body determination is performed based on the correlation coefficient. That is, when the magnitude of the correlation coefficient | ρ | is close to “1”, it is estimated that the image plane position is displaced according to the detection time, and therefore it is definitely determined that “the subject is a moving body”. You can

On the other hand, when the magnitude | ρ | of the correlation coefficient is close to "0", the image plane position has a substantially constant value, or the image plane position fluctuates regardless of the detection time. This is a case, and in both cases, it is possible to reliably determine that the subject is a stationary body. Therefore, there is no risk of erroneously determining that the stationary subject is a moving object due to variations in the detection of the image plane position, and the accuracy of moving object determination can be significantly improved.

Further, since the moving object determination is performed by efficiently using the data of the image plane position in the sample section, the moving object determination is performed as compared with the conventional case where the data is thinned to perform the moving object determination. The reliability of can be further improved.
Further, the magnitude of the correlation coefficient | ρ | takes a value from 0 to 1 regardless of the speed of the image plane speed, so that "a subject with a high image plane speed" and "a subject with a slow image plane speed" For both and, the moving object determination based on the same threshold can be performed.

Further, in the past, since the moving object determination was made based on the magnitude of the image plane velocity, the subject having a small image plane velocity, such as "a subject moving at a low speed" or "a subject moving at a distant place", is uniform. It was determined to be stationary. However, in the present invention, since the moving body determination is performed based on the magnitude of the correlation coefficient, it is possible to correctly determine such a subject having a small image plane velocity as “moving body”.

In the moving body judging apparatus according to claim 2,
The correlation coefficient is calculated for the “distance to the subject” and the detection time, and the moving body determination is performed based on the correlation coefficient.
That is, when the magnitude of the correlation coefficient | ρ | is close to “1”, it is estimated that the “distance to the subject” is displaced according to the detection time, and therefore the “subject is a moving body” is surely performed. You can judge.

On the other hand, when the magnitude of the correlation coefficient | ρ | is close to “0”, the “distance to the subject” takes a substantially constant value, or the “distance to the subject” is irrespective of the detection time. Is fluctuating, and in both cases, it can be reliably determined that the subject is a stationary body. Therefore, there is no possibility of erroneously determining that the “still subject is a moving body” due to variations in the detection of the “distance to the subject” and the like, and the accuracy of moving body determination can be significantly improved.

Further, since the moving object determination is performed by efficiently using the data of the "distance to the subject" in the sample section, the moving object determination is performed as compared with the case where the moving object determination is performed by thinning out the data. The reliability can be further enhanced.
Furthermore, since the magnitude of the correlation coefficient | ρ | takes a value from 0 to 1 regardless of the subject speed, the same is true for both the “subject with fast subject speed” and the “subject with slow subject speed”. A moving object determination based on a threshold can be implemented.

Further, in the past, since the moving object was determined based on the speed of the object, the object having a low object speed was uniformly determined as a stationary object. However, in the present invention, since the moving body determination is performed based on the magnitude of the correlation coefficient, it is possible to correctly determine such a subject having a low subject velocity as “moving body”.

In the moving body judging apparatus according to the third aspect, since the hysteresis characteristic is added to the threshold judgment, the result of the moving body judgment does not frequently change due to the slight variation of the correlation coefficient, and the stable moving body judgment is realized. can do. Therefore, even in the case where another subject enters the focus detection area due to camera shake or the like and the correlation coefficient fluctuates for a moment, the result of the moving object determination does not suddenly change.
Therefore, it is possible to stabilize the focus adjustment operation and the like.

In the moving body determination apparatus according to claim 4 or 5, the moving body determination is performed based on a plurality of threshold value determinations. Therefore, due to sudden changes in the correlation coefficient,
The accuracy of the moving body determination can be further improved without affecting the result of the moving body determination. Further, since the correlation coefficient is individually detected by dividing the sample section into a plurality of sections, a good linearity is obtained for each short sample section for a subject whose image plane velocity or subject velocity changes momentarily. After the approximation is performed, each correlation coefficient is accurately detected.

Therefore, even for a subject whose image plane position or subject distance changes non-linearly, the causal relationship with the detection time can be reliably determined. As a result, even for non-linearly moving image planes or subjects,
Accurate and reliable motion determination can be performed. Claim 6
In the moving body determination device according to the above, when performing the threshold determination for a predetermined number of times or for a predetermined period of time, the threshold is changed according to the generation of the correlation coefficient. Can be done.

For example, when the threshold value is gradually lowered as the generation of the correlation coefficient becomes newer, it is possible to quickly detect a subject that has started to move rapidly as a moving body. On the contrary, when the threshold value is gradually increased as the generation of the correlation coefficient becomes newer, it is possible to quickly detect a subject that has stopped rapidly as a stationary object.

According to the seventh aspect of the present invention, in the moving object determination apparatus, the defocus amount and the lens position are combined to obtain the image plane position of the subject image, and the image plane position and the "accumulation center time when the defocus amount is detected. , The correlation coefficient is calculated. Therefore, when the focus detection device of the phase difference detection method is provided in parallel, it is not necessary to newly provide the image plane position detection means, and the structure of the moving body determination device can be realized easily and at low cost.

According to an eighth aspect of the focus adjusting apparatus, the moving body determining apparatus 11 according to any one of the first to seventh aspects.
It comprises. Therefore, the appropriateness of the predictive drive can be accurately determined based on the accurate and reliable determination of the moving body, so that the focus adjustment suitable for the subject situation can be appropriately performed to significantly improve the performance and reliability of the focus adjustment. You can

[Brief description of drawings]

FIG. 1 is a block diagram illustrating the principle of the invention described in claim 1.

FIG. 2 is a principle block diagram for explaining the invention described in claim 2;

FIG. 3 is a principle block diagram for explaining the invention described in claim 7;

FIG. 4 is a principle block diagram for explaining the invention described in claim 8;

FIG. 5 is a diagram showing a change in image plane position.

FIG. 6 is a diagram showing a first embodiment (corresponding to claims 1, 3, 7, and 8).

FIG. 7 is a flowchart showing the operation of the first embodiment.

FIG. 8 is an explanatory diagram showing prediction driving.

FIG. 9 is a flowchart showing the operation of the second embodiment (corresponding to claims 1, 3, 4, 7, and 8).

FIG. 10 shows a third embodiment (claims 1, 3, 4, 5).
6 (corresponding to No. 6, 7, and 8).

FIG. 11 is a fourth embodiment (corresponding to claims 2 and 5).
FIG.

FIG. 12 is a flowchart showing the operation of the fourth embodiment.

FIG. 13 is a flowchart showing conventional moving body determination.

FIG. 14 is a diagram illustrating an example of use of terms related to focus adjustment.

FIG. 15 is a diagram showing a locus of an image plane position F.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Image plane position detection means 2 1st correlation means 3 Moving body determination means 4 Subject distance measuring means 5 2nd correlation means 6 Defocus amount detection means 7 Lens position detection means 8 Image plane position calculation means 11 Moving body determination device 12 Prediction Means 13 Prediction driving means 14 Selection means 21 Camera 22 Imaging optical system 22a Lens information storage 23 Quick return mirror 24 Submirror 25 Focus detection 26 Microprocessor 27 Motor drive circuit 28 Motor 29 Lens drive mechanism 30 Encoder 41 Camera body 42 Imaging optics System 43 Release 44 Microprocessor 45 Drive circuit 46 Light emitting element 47 Light transmitting lens 48 Light receiving lens 49 Light receiving element 50 Light receiving circuit 51 Feeding mechanism

Claims (8)

[Claims] 1. An image plane position detecting means for detecting an image plane position of a subject image by a photographing optical system, and a detection time of "a predetermined number of image plane positions" detected by the image plane position detecting means. First correlator for calculating the correlation coefficient, and when the magnitude of the correlation coefficient calculated by the first correlator is larger than a predetermined threshold value, it is determined that the subject is a moving body. A moving body determination device comprising: moving body determination means. 2. A subject distance measuring means for detecting a distance to a subject, and a correlation coefficient with a detection time for a “distance to a predetermined number of samples of a subject” detected by the subject distance measuring means. The second correlation means and the moving body determination means for determining that the subject is a moving body when the magnitude of the correlation coefficient calculated by the second correlation means is larger than a predetermined threshold value. A moving object determination device characterized by the above. 3. The moving body determination device according to claim 1, wherein the moving body determination unit adds a hysteresis characteristic to the threshold value or the correlation coefficient to determine the threshold value of the correlation coefficient. A moving object determination device characterized by the above. 4. The moving body determination device according to claim 1, wherein the moving body determination unit determines whether or not the subject is a moving body according to a result of matching threshold determinations performed a predetermined number of times. Characteristic moving object determination device. 5. The moving body determination device according to claim 1, wherein the moving body determination means determines whether or not the subject is a moving body according to a result of matching threshold determinations over a predetermined time. Characteristic moving object determination device. 6. The moving body determination device according to claim 4, wherein in the moving body determination means, a threshold value to be compared with the correlation coefficient differs depending on the generation of the correlation coefficient. And a moving body determination device. 7. The moving object determination device according to claim 1, wherein the image plane position detection unit detects a phase difference in space between a pair of optical images obtained by dividing and forming a transmitted light flux of the photographing optical system. Then, based on the phase difference, a defocus amount detecting means for detecting a defocus amount corresponding to "a distance between the subject image and the imaging surface", a lens position detecting means for detecting a lens position of the photographing optical system, Image surface position calculating means for calculating the image surface position of the subject image by combining the defocus amount and the lens position, wherein the first correlating means is configured by the image surface position calculating means. A moving object determination device characterized by calculating a correlation coefficient with respect to the calculated "image plane position of a predetermined number of samples" and "accumulation center time of a set of optical images" in the defocus amount detection means. 8. The moving body determination device according to claim 1, wherein the movement of the subject is predicted by extending the movement of the image plane position or the “distance to the subject”. A predicting unit, a predicting driving unit that predictively drives the photographing optical system to a focus position according to the “movement of the subject” predicted by the predicting unit, and the moving body determining unit determines that the “subject is a moving body”. And a selecting unit that executes the predictive driving by the predictive driving unit and prohibits the predictive driving by the predictive driving unit in other cases.