JP7305332B2 - LENS CONTROL DEVICE AND CONTROL METHOD THEREOF - Google Patents

Description

The present invention relates to image plane position prediction.

Conventionally, in order to focus on a moving subject, the image plane position at a time later than a certain time is predicted based on the focus detection result at a certain time, and the focus lens is adjusted based on the predicted image plane position. Techniques for controlling drive have been proposed.

In Patent Document 1, by using a plurality of data of the image plane position and the corresponding focus detection time, the coefficient that best fits the prediction function is obtained by the least squares method (also called collective least squares method), so that the image plane It is stated that techniques for predicting position are known.

Japanese Unexamined Patent Publication No. 2001-21794

By predicting the image plane position using the above-described lumped least squares method and controlling the driving of the focus lens based on the image plane position, it is possible to focus on a moving subject.

However, when using the lumped least squares method to predict the image plane position of an object at a time later than a certain time, there is a problem that it is necessary to increase the amount of calculation in order to obtain a more stable prediction result. . This is because it is necessary to perform calculations using more image plane position and corresponding focus detection time data.

SUMMARY OF THE INVENTION Accordingly, it is an object of the present invention to provide a lens control device and a control method thereof that can appropriately focus on a moving subject while reducing the processing load.

focus detection means for obtaining a focus detection result; estimation means for estimating information corresponding to an image plane position of an object based on the focus detection result based on a sequential least squares method; a first predicting means for predicting information corresponding to the future image plane position of the subject based on information corresponding to the image plane position; and information corresponding to the future image plane position of the subject based on the lumped least squares method. and controlling driving of the focus lens based on information corresponding to the future image plane position of the subject predicted by the first prediction means or the second prediction means. and drive control means, wherein the drive control means predicts the future image plane position of the subject predicted by the second prediction means when the shooting distance to the subject is less than the first distance. is configured to drive the focus lens based on information corresponding to

According to the present invention, it is possible to appropriately focus on a moving subject while suppressing the processing load.

1 is a block diagram showing the configuration of a digital single-lens reflex camera equipped with Example 1; FIG. 5 is a flowchart showing an operation example of photographing processing in Example 1. FIG. 5 is a flow chart showing focus adjustment processing in Example 1. FIG. 10 is a flowchart showing focus adjustment processing in Example 2. FIG. 10 is a flowchart showing focus adjustment processing in Example 3. FIG. 14 is a flow chart showing focus adjustment processing in Example 4. FIG. FIG. 4 is a diagram for explaining Kalman filter calculation in Example 1; It is a figure explaining the collective least-squares method.

[Example 1]
Examples to which the present invention is applied will be described below. In this embodiment, as an example, an example in which the present invention is applied to a digital single-lens reflex camera will be described.

<Structure of Imaging Device>
FIG. 1 is a block diagram showing the configuration of a digital single-lens reflex camera.

The lens driving circuit 202 has, for example, a DC motor or a stepping motor, and adjusts the focus position by changing the position of the focus lens of the photographing lens 201 under drive control of the microcomputer 224 . That is, the microcomputer 224 performs lens control.

A lens communication circuit 203 communicates with a microcomputer (not shown) inside the photographing lens 201 . The content of communication is controlled by the microcomputer 224 to obtain the state of the photographing lens 201 .

A diaphragm driving circuit 205 drives the diaphragm 204 . The amount to be driven is calculated by the microcomputer 224 to change the optical aperture value.

The main mirror 206 is normally arranged so as to reflect the light beam to guide it to the viewfinder (not shown). Evacuate from inside. That is, the luminous flux incident from the photographing lens 201 is switched between the finder side and the imaging device side. The central portion of the main mirror 206 is a half-mirror so that part of the light can be transmitted therethrough, and transmits part of the light flux so as to be incident on a focus detection sensor for focus detection.

The sub-mirror 207 reflects the light beam transmitted from the main mirror 206 and guides it to a focus detection sensor (located in the focus detection circuit 210) for focus detection.

The pentaprism 208 guides the light flux reflected by the main mirror 206 to a finder section (not shown). The finder section also has a focus plate, an eyepiece lens (not shown), and the like.

The photometry circuit 209 converts the color and brightness of the subject image formed on the focusing screen (not shown) into electrical signals by a photometry sensor (located in the photometry circuit 209) having a color filter. .

A light beam transmitted through the central portion of the main mirror 206 and reflected by the sub-mirror 207 reaches a focus detection sensor arranged inside the focus detection circuit 210 for photoelectric conversion. The defocus amount, which is the result of focus detection, is obtained by calculating the output of the focus detection sensor. The microcomputer 224 evaluates the calculation result and instructs the lens drive circuit 202 to drive the focus lens.

A shutter drive circuit 212 drives the focal plane shutter 211 . The shutter opening time is controlled by the microcomputer 224 .

A CCD, a CMOS sensor, or the like is used as the imaging element 213, and converts the subject image formed by the imaging lens 201 into an electric signal.

A clamp circuit 214 and an AGC circuit 215 perform basic analog signal processing before A/D conversion, and a microcomputer 224 changes the clamp level and AGC reference level.

An A/D converter 216 converts the analog output signal of the imaging element 213 into a digital signal.

The video signal processing circuit 217 is realized by a logic device such as a gate array, performs filter processing, color conversion processing, gamma processing, and compression processing such as JPEG on digitized image data. Output.

The video signal processing circuit 217 can output information such as exposure information and white balance of the signal of the image sensor 213 to the microcomputer 224 as necessary. Based on the information, the microcomputer 224 gives instructions for white balance and gain adjustment. In the case of a continuous shooting operation, the unprocessed image data is temporarily stored in the buffer memory 223, the unprocessed image data is read out through the memory controller 220, and the video signal processing circuit 217 performs image processing and compression processing. to perform continuous shooting. The number of continuous image shots depends on the size of the buffer memory.

The memory controller 220 stores the unprocessed digital image data input from the video signal processing circuit 217 in the buffer memory, and stores the processed digital image data in the memory 221 . Conversely, it outputs image data from the buffer memory 223 and the memory 221 to the video signal processing circuit 217 . Memory 221 may be removable. The memory controller 220 can output images stored in the memory 221 via an external interface 222 connectable to a computer or the like.

The operation member 225 informs the microcomputer 224 of its state, and the microcomputer 224 controls each part according to the change of the operation member.

A switch 1 (hereinafter referred to as SW1) and a switch 2 (hereinafter referred to as SW2) are switches that are turned on and off by operating the release button, and each is one of the input switches of the operation member 225 . A state in which only SW1 is on is a state in which the release button is half-pressed, and in this state autofocus operation and photometry operation are performed.

When both SW1 and SW2 are on, the release button is fully pressed, and the release button is on for recording an image. Shooting is performed in this state. A continuous shooting operation is performed while SW1 and SW2 are kept ON. Switches (not shown) such as an ISO setting button, an image size setting button, an image quality setting button, and an information display button are connected to the operation member 225, and the states of the switches are detected.

The liquid crystal drive circuit 228 drives the external liquid crystal display member 229 and the in-finder liquid crystal display member 230 in accordance with the display content command from the microcomputer 224 . Further, a backlight such as an LED (not shown) is arranged on the in-finder liquid crystal display member 230 , and the LED is also driven by the liquid crystal drive circuit 228 . The microcomputer 224 confirms the memory capacity through the memory controller 220 based on the predicted image size data corresponding to the ISO sensitivity, image size, and image quality set before shooting, and then determines the remaining number of shootable images. can be calculated. It can also be displayed on the external liquid crystal display member 229 and the in-finder liquid crystal display member 230 as required.

The non-volatile memory 231 (EEPROM) can store data even when the camera is powered off. For example, a plurality of image plane position information corresponding to defocus amounts detected at different times are stored together with corresponding time information. The image plane position and corresponding time information may be stored in a volatile memory (not shown) so that the data is erased when the power is turned off.

The power supply unit 232 supplies necessary power to each IC and drive system.

The microcomputer 224 of this embodiment is an example of the state calculation means, the distance calculation means, the error calculation means, the speed calculation means, the defocus amount detection means, and the determination means in the claims of the present application.

<Kalman filter calculation>
Here, the Kalman filter calculation will be described. Time-series data y(k) at time k is given by the following equation. Note that the time-series data is also referred to as an observed value. In the following description, time k−1, time k, and time k+1 all correspond to times when time-series data is obtained.
y(k)= ^XT (k)A(k)+ω(k)
A(k+1)=L(k)A(k)+m(k)ν(k)

X(k) and m(k) are n-dimensional column vectors, A(k) is an n-dimensional column vector (state vector), ω(k) is observation noise with mean 0 and variance σ _ω ² , L (k) is an n×n matrix, and ν is system noise with an average value of 0 and a variance of σ _ν ² .

The Kalman filter calculation is for obtaining the state vector A(k), and the calculation step is divided into a prediction step and a filtering step. First, the prediction step estimates the state in advance, and the filtering step estimates the state using the observation results.

In the prediction step, the prior state estimation vector A′(k) (n-dimensional column vector) and the prior error covariance matrix P′(k) (n×n matrix) are obtained by the following equations.
A'(k)=L(k-1)A(k-1)
P′(k)=L(k−1)P(k−1) ^LT (k−1)+
σ _ν ² (k−1)m(k−1)m ^T (k−1)

As shown in the above equation, the prior state estimation vector A'(k) is the state vector at time k by the state vector (k-1) obtained at time k-1 and an arbitrary L(k) It is an estimate. Also, the prior error covariance matrix P'(k) is an equation for estimating the error between the state vector A(k) and the prior state estimation vector A'(k) at time k.

In the filtering step, the state estimation vector A(k) (n-dimensional column vector) is obtained by the following formula based on the detected time-series data y(k). Also, the posterior error covariance matrix P(k) (n×n matrix) is obtained by the following equation.
A(k)=A'(k)+g(k)(y(k) ^-XT (k)A'(k))
P(k)=(Ig(k) ^XT (k))P'(k)

As shown in the above equation, A(k) is the difference between the actual detection result y(k) and the predicted detection result X ^T( K)A′(K) is multiplied by the Kalman gain g(k) and added to A'(k). Note that the matrix I is an n×n unit matrix.

Also, the Kalman gain g(k) is obtained by the following equation.

As shown in the above equation, the larger the observation noise σ ² _ω (k), the smaller g(k). Also, the larger the prior error covariance matrix P'(k), the smaller g(k). That is, when it is considered that there is a high possibility that an error has occurred in the detected y(k) or XT ⁽ K)A'(K), g(k) is become smaller. This makes the calculated A(k) less susceptible to errors.

The initial value A(0) of the state vector and the initial value P(0) of the error covariance matrix are given by the following equations.

<Kalman filter calculation in this embodiment>
The Kalman filter calculation in this embodiment will be described. y(k) is the detection result of the image plane position at time k. Then, in the Kalman filter calculation of this embodiment, the image plane position and the image plane moving speed at time k are estimated from the state vector A(k) as information about the state of the object. Further, by obtaining the state vector A(k+1) based on the state vector A(k), the image plane position and the image plane moving speed at the time k+1 are estimated as the information regarding the state of the object.

Note that the image plane position in this embodiment is the position of the back focal point corresponding to the photographing lens 201 (also referred to as the image plane position of the photographing lens or the lens image plane position). Further, the image plane position corresponding to the subject is the position of the back focus when the photographing lens 201 is at a position where the subject is front-focused. In other words, the image plane position corresponding to the subject is obtained by adding the result of the focus detection (defocus amount in this embodiment) to the position of the back focus at the time when focus detection is performed on the subject. is the position of the back focal point calculated by In this embodiment, an example in which the image plane position is used as the information corresponding to the image plane position will be described, but information other than the image plane position may be used as the information corresponding to the image plane position. For example, since the image plane position corresponds to the position of the photographing lens 201, the position of the photographing lens 201 corresponding to the image plane position may be used in place of the image plane position of this embodiment. In this case, the lens position corresponding to the image plane position corresponding to the object is the focus lens position calculated as follows. That is, it is the position of the focus lens calculated by adding the result of the focus detection (the defocus amount in this embodiment) to the focus lens position at the time when focus detection is performed on the object.

Here, a model formula for predicting the motion of the subject using information about the state of the subject (the image plane position and the image plane moving speed estimated by the Kalman filter calculation) will be described. An example is shown in FIG. Consider the prediction of the image plane position of the subject by a linear expression (two-dimensional) as indicated by the dashed line in FIG. The image plane position at time k is assumed to be a model that can be predicted from the average image plane moving velocity v at time k and the image plane position _yA at time 0. At this time, the column vector A is defined as the image plane position (intercept) _yA at time 0 and the average image plane movement velocity (inclination) v at time k. Also define the column vector X as 1 such that the times k and y _A are constant. The variance σ _ω ² may be entered based on the variance of the detection results. In the initial value A(0), the initial value of y _A may be entered based on, for example, the first detected image plane position y ₀ . Also, the initial value v of the average image plane moving speed may be set to 0. An appropriate value may be entered for the initial value P(0). The matrix L, the column vector m, and the variance σ _ν ² may be set according to the properties of the model, that is, the properties such as the movement of the object to be photographed, and may be time-invariant.

Note that the image plane moving speed is the speed at which the image plane position moves, and is a speed corresponding to the moving speed of the object. In this embodiment, the image plane moving speed is used for explanation, but the speed is not limited to this as long as it corresponds to the image plane moving speed. For example, it may be the moving speed of the focus lens position corresponding to the moving speed of the image plane position.

In this embodiment, an example in which the model formula is a linear formula (two-dimensional) is given as an example only, but the model formula may have any number of dimensions according to the assumed movement of the subject. A column vector A may be defined together.

We have now defined the matrices, vectors and variances required for the prediction step. After that, by repeating the filtering step using the detection result of the image plane position and the prediction step for the next time, a model formula for predicting the movement of the subject by the Kalman filter calculation can be obtained.

According to the Kalman filter calculation, as described above, since calculation is performed in consideration of errors, it is possible to predict the image plane position with high accuracy even in a situation where errors are likely to occur in focus detection results.

<Calculation by collective least squares method>
Here, with reference to FIG. 8, the prediction of the image plane position based on the collective least squares method disclosed in Patent Document 1 will be described in more detail.

The image plane position at time _xk is _yk , which is represented by ● in the figure. Here, when a plurality of ● in the figure are represented by an (n−1)-order multidimensional model for time t, y(t) is the image plane position, T(t) is the time parameter, and the coefficients of the model formula is a column vector C(t), the following formula gives a curve like the one-dot chain line in the figure.
y(t)=C(t) ^T T(t)

Column vector C(t) and column vector T(t) are as follows.

As a method for obtaining the column vector C described above, a collective least squares method can be used. A matrix given by the (n-1)th order polynomial expression for the model formula, N for the number of past image plane position measurement results, and the times at which focus detection results corresponding to past image plane positions including time k were obtained Let Z(k) be a column vector Y(k) given from a past image plane position including time k. At this time, the column vector C(k) to be obtained is given by the following equation.
C(k)=[Z(k) ^T Z(k)] ⁻¹ Z(k) ^T Y(k)

Matrix Z(k) and column vector Y(k) are given by the following equations respectively.

By predicting the image plane position using the above-described lumped least squares method and controlling the driving of the focus lens based on the image plane position, it is possible to focus on a moving subject.

However, when using the collective least squares method to predict the image plane position of an object at a time later than a certain time, there is a problem that it is necessary to increase the amount of calculation in order to obtain a more stable prediction result. . The amount of calculation for obtaining the column vector C is at least O(N·n ² ), and in order to obtain a stable column vector C, the number N of past measurement results must be increased. The amount of computation increases in proportion to N.

On the other hand, the computational complexity of the Kalman filter with n-dimensional parameters is O(n ³ ). When the model formula has a low dimension, the number of computations of the Kalman filter can be made sufficiently smaller than that of the lumped least squares method of the same dimension. Also, when there are variations in the observation results, the batch least squares method tends to take a ^large number N of past observation results used for calculation. If N is made large, it becomes much larger than the computational complexity of the Kalman filter. Therefore, by using the Kalman filter in an appropriate scene as in the present embodiment, the amount of calculation can be made smaller than the amount of calculation of the collective least-squares method.

<Operation example of shooting processing>
Next, an operation example of photographing processing according to the first embodiment of the present invention will be described with reference to the flowchart of FIG.

In general, cameras have a mode in which the lens is driven in relation to the subject's image plane at a certain time (one-shot shooting mode), and a mode in which the lens is driven while predicting the subject's image plane at a time later than the current time ( There are two types of predictive shooting mode). Example 1 shows the motion when the camera is set to the predictive shooting mode.

At step S401, the microcomputer 224 determines the state of SW1. If SW1 is on, the process proceeds to step S402. When SW1 is turned off, the microcomputer 224 controls to terminate the predictive photographing mode.

At step S402, the microcomputer 224 controls to perform focus adjustment processing. Details will be described later with reference to FIG.

At step S403, the microcomputer 224 determines the state of SW2. If SW2 is off, the process returns to step S401. If SW2 is on, the process proceeds to step S404.

In step S404, the microcomputer 224 raises the main mirror 206 and operates the focal plane shutter 211 to perform photographing. After that, the process returns to step S401.

<Focus adjustment processing>
Next, an example of the operation of focus adjustment processing in step S402 will be described with reference to the flowchart of FIG.

At step S501, the microcomputer 224 drives the focus detection circuit 210 to obtain the defocus amount. Further, the image plane position, which is the position of the imaging lens at which the subject is in focus, is acquired from the defocus amount and the current imaging lens position.

In step S502, the microcomputer 224 controls storage processing. In the storage process, the image plane position and the detection time of the defocus amount obtained in step S501 are stored in the memory.

At step S503, the microcomputer 224 performs the aforementioned Kalman filter calculation. The Kalman filter is a kind of sequential identification method and does not require multiple pieces of time-series data for calculation unlike the lumped least squares method. Here, from the state vector A(k), the image plane position and the image plane moving speed at the time (k) are estimated as information about the state of the object.

At step S504, microcomputer 224 evaluates whether the number of data stored in memory is less than the first number. If the number of data is less than the first number, go to step S508. If the number of data is greater than or equal to the first number, go to step S505.

At step S505, the microcomputer 224 evaluates the degree of identification of the Kalman filter performed at step S503. The degree of identification is evaluated using the posterior error covariance matrix P described above. This is because by using the same matrix P, it is possible to determine whether the result of the state estimation vector A obtained by the Kalman filter calculation has converged. If the degree of identification, ie P, is less than the first value, the process proceeds to step S507. This is because there is a possibility that the result of the state estimation vector A obtained by the Kalman filter calculation has not converged, that is, an error has occurred in the image plane position obtained by the Kalman filter calculation. If the degree of identification, that is, P is greater than or equal to the first value, the process proceeds to step S506. By switching the prediction method based on the degree of identification in this way, it is possible to select a method that reduces the amount of calculation when the Kalman filter calculation has sufficiently converged.

In step S506, the microcomputer 224 performs focus prediction by the collective least squares method. Specifically, using a plurality of time-series data of the image plane positions and their detection times stored in the memory in step S502, a model formula of the image plane position of the subject is obtained by the lumped least squares method. The image plane position of the subject at time (k+1) later than time (k) of is obtained.

In step S507, the microcomputer 224 performs focus prediction using the Kalman filter. By obtaining the state vector A(k) in step S503, the model formula of the image plane position of the subject can be identified, and the future image plane position is predicted based on the result. More specifically, the image plane position (corresponding to the intercept) and the image plane moving speed (corresponding to the tilt) at the time (k) estimated as information about the state of the subject by obtaining the state vector A(k). and the model formula, the image plane position at time k+Δd is calculated. Here, the time k+Δd is the time corresponding to the shooting time. For example, time k+Δd corresponds to the time of S404.

In step S508, the microcomputer 224 controls to drive the focus lens based on the focus detection result. The photographing lens is driven based on the image plane position of the object obtained in step S501.

In step S509, the microcomputer 224 drives the lens based on the focus prediction result. Since the future image plane position is obtained in step S506 or step S507, the focus lens is driven based on the result.

<effect>
As described above, in the first embodiment, driving of the focus lens is controlled based on the image plane position of the object calculated using the Kalman filter calculation. As a result, it is possible to appropriately focus on a moving subject while suppressing the processing load as compared with the case where the Kalman filter calculation is not used.

In addition, in the first embodiment, by switching between the case of using the Kalman filter calculation and the case of not using the Kalman filter calculation according to conditions, it is possible to appropriately focus on a moving subject while suppressing the processing load.

[Example 2]
In the first embodiment, an example in which the Kalman filter calculation is used when the degree of identification by the Kalman filter calculation is equal to or higher than the threshold has been described.

In a second embodiment, in addition to the degree of identification by the Kalman filter calculation, an example will be described in which the movement of the subject is considered. Note that the description of common points with the first embodiment will be omitted as much as possible, and the description will focus on the points of difference.

Since the configuration of the digital single-lens reflex camera (FIG. 1) and the operation of the photographing process (FIG. 2) are the same as those of the first embodiment, description thereof will be omitted.

<Focus adjustment processing>
The operation of focus adjustment processing in this embodiment will be described with reference to the flow chart of FIG.

Since step S601 is the same as step S501, description thereof is omitted.

In step S602, the microcomputer 224 performs subject distance calculation. Here, the subject distance is the shooting distance from the camera to the subject. Using the lens communication circuit 203, the subject distance obtained from the current imaging position of the photographing lens is obtained from the photographing lens 201. FIG.

Steps S603 and S604 are the same as steps S502 and S503, respectively, so description thereof will be omitted.

At step S605, the microcomputer 224 calculates the image plane moving speed. The image plane moving speed may be obtained by the Kalman filter calculation in step S604. In this embodiment, the model formula is two-dimensional (linear formula) with respect to time k. Since the column vector A is defined to represent the image plane position (intercept) at time 0 and the average image plane movement speed (inclination) at time k, by obtaining A(k), the image plane movement You can get speed. Alternatively, it may be determined by other known methods.

Since steps S606 and S607 are the same as steps S504 and S505, respectively, the description thereof is omitted.

At step S608, the microcomputer 224 evaluates the subject distance. If the object distance obtained in step S602 is greater than or equal to the first distance, the process proceeds to step S609, and if less than the first distance, the process proceeds to step S610. Assuming that the model formula used for the Kalman filter calculation is, for example, a linear (two-dimensional) model formula, when the object distance is short, the image plane position of the object may not match the model formula. Therefore, when the subject distance is less than the first distance, the process proceeds to step S610 to obtain the future image plane position by the lumped least squares method.

At step S609, the microcomputer 224 evaluates the image plane moving speed. If the image plane moving speed obtained in step S604 is less than the first speed, the process proceeds to step S611, and if it is the first speed or more, the process proceeds to step S610. This also depends on the model formula used for the Kalman filter calculation, but it is because the image plane position of the subject does not match the model formula when the image plane moving speed is high.

Steps S610 to S613 are the same as steps S506 to S509, respectively, so description thereof will be omitted.

<effect>
As described above, in the second embodiment, the Kalman filter calculation is employed according to the movement of the subject to obtain the image plane position. More specifically, whether or not to adopt the Kalman filter calculation is switched according to the image plane moving speed corresponding to the moving speed of the subject. It is possible to appropriately focus on a moving subject while suppressing the processing load.

[Example 3]
In the first embodiment, an example in which the Kalman filter calculation is used when the degree of identification by the Kalman filter calculation is equal to or higher than the threshold has been described. In addition, in the second embodiment, an example in which the Kalman filter calculation is used in consideration of the movement of the subject in addition to the first embodiment has been described.

In a third embodiment, an example using Kalman filter calculation in consideration of variations in focus detection results will be described. Note that the description of common points with the first embodiment will be omitted as much as possible, and the description will focus on the points of difference.

Since the configuration of the digital single-lens reflex camera (FIG. 1) and the operation of the photographing process (FIG. 2) are the same as those of the first embodiment, description thereof will be omitted.

<Focus adjustment processing>
The operation of focus adjustment processing in this embodiment will be described with reference to the flow chart of FIG.

Steps S701 to S703 are the same as steps S501 to S503, respectively, so description thereof will be omitted.

In step S704, the microcomputer 224 performs variation calculation. Focus detection results may generally have some variation. For example, the longer the subject distance, the greater the variation in focus detection results. In step S702, the microcomputer 224 uses the image plane position and its detection time stored in the memory to obtain the variation (dispersion) of the image plane position of the object.

Since steps S705 and S706 are the same as steps S504 and S505, respectively, the description thereof is omitted.

At step S707, the microcomputer 224 evaluates variations. If the image plane position variation obtained in step S704 is equal to or greater than the second value, the process proceeds to step S709 to perform focus prediction using the Kalman filter calculation. Since the Kalman filter calculation takes into account the errors in the detection results, by performing focus prediction using the Kalman filter calculation, it is possible to obtain stable calculation results even in situations where the focus detection results vary. be. As a result, it is possible to calculate the image plane position with higher accuracy. If the image plane position variation obtained in step S704 is less than the second value, the process proceeds to step S708.

Steps S708 to S711 are the same as steps S506 to S509, respectively, so description thereof will be omitted.

<effect>
As described above, in the third embodiment, Kalman filter calculation is used according to variations in focus detection results. More specifically, the Kalman filter calculation is used when the variation in the focus detection result is equal to or greater than the threshold. As a result, it is possible to appropriately focus on a moving subject while suppressing the processing load.

[Example 4]
When using the Kalman filter calculation, if the matrix L and the column vector m are fixed, the image plane position of the object may deviate from the model formula depending on the movement of the object. Specifically, when the object approaches, the object distance becomes shorter and the image plane movement speed of the object becomes faster. image plane position tends to deviate from the model formula. The amount of change in state estimation vector A can be changed by matrix L, column vector m, and variance σ _ν ² . Therefore, in a fourth embodiment, an example will be described in which the Kalman filter calculation is performed with higher accuracy by changing the setting parameters used for the Kalman filter calculation according to various conditions.

First, Example 4 to which the present invention is applied will be described. Note that the description of common points with the first embodiment will be omitted as much as possible, and the description will focus on the points of difference.

Since the configuration of the digital single-lens reflex camera (FIG. 1) and the operation of the photographing process (FIG. 2) are the same as those of the first embodiment, description thereof will be omitted.

<Focus adjustment processing>
Next, an operation example of focus adjustment processing according to the fourth embodiment of the present invention will be described with reference to the flowchart of FIG.

Since step S1301 is the same as step S501, description thereof is omitted.

In step S1302, the microcomputer 224 calculates the image plane moving speed.

In step S1303, the microcomputer 224 performs subject distance calculation. Using the lens communication circuit 203, the subject distance obtained from the current imaging position of the photographing lens is obtained from the photographing lens 201. FIG.

In step S1304, the microcomputer 224 sets the matrix L and the column vector m according to the focal length, the image plane moving speed obtained in step S1302, and the object distance obtained in step S1303. In this way, by changing the matrix L and the column vector m based on the image plane moving speed and the object distance, the future image plane position can be obtained without the object distance of the object deviating from the model formula.

At step S1305, the microcomputer 224 evaluates the column vector m obtained at step S1304. Evaluate if column vector m is less than a first threshold. If the column vector m is less than the first threshold, proceed to step S1306; otherwise, proceed to step S1307.

At step S1306, the microcomputer 224 sets the column vector m to a default value. When the amount of change indicated by the column vector m is small, the default value is used to define the minimum value of the amount of change.

In step S1307, the microcomputer 224 performs Kalman filter calculation. Since step S1307 is the same as step S503, description thereof is omitted.

In step S1308, the microcomputer 224 performs focus prediction using the Kalman filter. Since step S1308 is the same as step S507, description thereof is omitted.

In step S1309, the microcomputer 224 controls driving of the focus lens based on the focus prediction result. Since the future image plane position is obtained in step S1308, drive control of the focus lens is performed based on the result.

<effect>
According to the embodiment described above, by changing the matrix L and the column vector m according to various conditions, even under conditions where the image plane position of the subject may deviate from the model formula, the future The image plane position can be obtained.

<Other Examples>
Although the Kalman filter operation is used in the above-described embodiment, a recursive least squares method (RLS method), which is another kind of recursive identification method, may be used. In this case, σ ² _ω (k)=1. That is, the iterative least squares method is a special case in which fewer parameters are set than the Kalman filter calculation of this embodiment. Therefore, from the viewpoint of the amount of calculation, the same effect as the Kalman filter can be obtained.

As for model formulas, as described above, model formulas other than those introduced in this embodiment can be used.

In the above-described embodiment, the image plane position and the image plane moving speed at time k are calculated from the state vector A(k) as information related to the state of the subject. If there is, it is not limited to this. For example, the configuration may be such that the image plane position and the subject moving speed at time k are calculated from the state vector A(k).

Also, in the fourth embodiment, it is explained that the matrix L and the column vector m are set according to various conditions, but at least one of the matrix L, the column vector m or the variance σ _ν ² may be set.

The matrix L, the column vector m, and the variance σ _ν ² may be set by the user. This makes it possible to adjust the calculation result in accordance with the characteristics of the shooting scene recognized by the user.

Further, the present invention supplies a program that implements one or more functions of the above-described embodiments to a system or apparatus via a network or a storage medium, and one or more processors in the computer of the system or apparatus reads and executes the program. It can also be realized by processing to Furthermore, it can be implemented by a circuit (eg, an ASIC) that implements one or more functions.

Although preferred embodiments of the present invention have been described above, the present invention is not limited to these embodiments, and various modifications and changes are possible within the scope of the gist.

224 microcomputer

Claims (8)

focus detection means for obtaining focus detection results;
estimating means for estimating information corresponding to the image plane position of the object based on the focus detection result based on the iterative least-squares method;
a first prediction means for predicting information corresponding to the future image plane position of the subject based on the information corresponding to the image plane position of the subject estimated by the estimation means;
a second prediction means for predicting information corresponding to the future image plane position of the subject based on the lumped least squares method;
drive control means for controlling drive of the focus lens based on information corresponding to the future image plane position of the subject predicted by the first prediction means or the second prediction means;
The drive control means adjusts the focus lens based on the information corresponding to the future image plane position of the subject predicted by the second prediction means when the shooting distance to the subject is less than the first distance. A lens control device characterized by driving. Provided with a distance calculation means for calculating the shooting distance to the subject,
The drive control means corresponds to the future image plane position of the subject predicted by the first prediction means when the photographing distance to the subject calculated by the distance calculation means is equal to or greater than a first distance. 2. The lens control device according to claim 1, wherein the focus lens is driven based on the information obtained. a second prediction means for predicting information corresponding to the future image plane position of the subject based on the lumped least squares method;
having speed calculation means for calculating an image plane moving speed that is a speed corresponding to the moving speed of the image plane position;
The drive control means corresponds to the future image plane position of the object predicted by the second prediction means when the image plane moving speed calculated by the speed calculation means is equal to or higher than a first speed. 2. The lens control device according to claim 1, wherein the focus lens is driven based on the information. The drive control means corresponds to the future image plane position of the subject predicted by the first prediction means when the image plane moving speed calculated by the speed calculation means is less than a first speed. 4. The lens control device according to claim 3, wherein the focus lens is driven based on the information. Having setting means for setting parameters in the iterative least squares method,
5. The lens control device according to any one of claims 1 to 4 , wherein the setting means sets the parameter according to the movement of the subject. 6. The lens control device according to claim 5, wherein said setting means sets said parameter based on at least one of an image plane moving speed and a photographing distance to an object. 7. The lens control device according to claim 1, wherein the iterative least squares method is a Kalman filter. A control method for a lens control device,
a focus detection step of obtaining a focus detection result; an estimation step of sequentially estimating information corresponding to an image plane position of a subject based on the focus detection result based on the least squares method;
a first prediction for predicting information corresponding to the future image plane position of the subject based on the information corresponding to the image plane position of the subject estimated in the estimating step, or the future of the subject based on a collective least squares method; a prediction step of predicting information corresponding to the future image plane position of the subject by a second prediction of predicting information corresponding to the image plane position of
a drive control step of controlling drive of the focus lens based on information corresponding to the future image plane position of the subject predicted by the first prediction or the second prediction;
In the drive control step, when the shooting distance to the subject is less than the first distance, the focus lens is driven based on information corresponding to the future image plane position of the subject predicted by the second prediction. A control method for a lens control device, characterized by:

Images