JP2002099013A - Shake detecting device and blur correcting optical instrument - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a shake detecting device and a shake correcting optical instrument capable of changing a composition and following up a moving subject without the sense of unpleasant or discomfort of a user, realizing high-quality photographing even shortly after a composition change is finished or while following up the subject, and accurately correcting a blur in a case the intentional shake at the time of starting of the shake detecting operation is followed by an unintentional camera shake afterward. SOLUTION: A reference value retention part 53 sequentially stores the reference value calculated by a reference value calculating part 52 and retains the earliest reference value at the time of a composition change detected by a composition change start detecting part 44. The earliest reference value is used when a composition change finish detecting part detects termination of the composition change.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a shake detecting device for detecting vibration due to hand shake and the like, and to a blur correcting optical apparatus such as an optical device such as binoculars or a photographing device such as a camera incorporating the same.

[0002]

2. Description of the Related Art A main source of vibration applied to an optical device such as binoculars or a photographing device such as a camera is a camera shake of a user. Heretofore, a blur correction optical device has been proposed as a means for correcting image vibration and blur caused by camera shake. The vibration applied to the optical device or the photographing device by the user can be roughly divided into vibration caused by unintentional movement of the user due to hand shake during rest (hereinafter referred to as a stationary state), panning and panning. Vibration caused by a movement intended by the user such as shooting (hereinafter, this state is referred to as a composition change state). Conventional blur correction optical devices are effective for correcting image vibration and blur in a stationary state.

[0003] The operation of the conventional blur correction optical apparatus will be described below with reference to FIG. FIG. 12 is a block diagram showing a basic configuration of a conventional shake correction optical apparatus including a shake detection device. The angular velocity sensor 10 is a sensor that detects a shake applied to the blur correction optical device, and usually uses a piezoelectric vibration type angular velocity sensor that detects a Coriolis force. The output of the angular velocity sensor 10 is transmitted to the reference value calculator 52. The reference value calculation unit 52 is a part that calculates a reference value of the shake from the output of the angular velocity sensor 10. The shake signal output from the angular velocity sensor 10 is transmitted to the integration unit 54 after subtracting the reference value. The integration unit 54
This is a part that integrates a shake signal expressed in units of angular velocity with respect to time and converts it into a shake angle of the blur correction optical device.

The target drive position calculating section 56 calculates target drive position information of the blur correction lens 80 by adding information such as the focal length of the lens to the information on the shake angle obtained by the integration section 54. . The drive signal calculation unit 58 is configured to drive the blur correction lens 80 in accordance with the target drive position information.
The difference from the position information of 0 is calculated, and a drive current is supplied to the coil 73.

The driving section 70 is a section for driving the blur correction lens 80, and includes a yoke 71, a magnet 72,
It is composed of a coil 73. The coil 73 is placed in a magnetic circuit formed by the yoke 71 and the magnet 72. When a current flows through the coil 73, a force is generated in the coil 73 according to Fleming's left-hand rule. The coil 73 is attached to a lens barrel 82 containing a blur correction lens 80. Shake correcting lens 80 and lens barrel 82
Has a structure that can be moved in a direction perpendicular to the optical axis I, so that the movement of the coil 73 can drive the blur correction lens 80 in a direction perpendicular to the optical axis I.

The optical position detector 79 monitors the movement of the blur correction lens 80, and includes an infrared light emitting diode (IRED) 74, a slit plate 76 having a slit 76a, a PSD (Position Sens).
active device) 77.

The light emitted from the IRED 74 first passes through the slit 76a, thereby narrowing the light beam width.
It reaches SD77. The PSD 77 outputs a signal corresponding to the position of the light on the light receiving surface. Since the slit plate 76 is attached to the lens barrel 82, the blur correction lens 80
Is the movement of the slit 76a and the movement of light on the light receiving surface of the PSD 77. Therefore, the position of the light on the light receiving surface of the PSD 77 is equivalent to the position of the blur correction lens 80. The signal output from the PSD 77 is a position signal 78
As feedback.

[0008]

Such a shake correcting optical apparatus is effective for correcting shake caused by unintentional camera shake in a stationary state. However, the usage mode of the camera or the like is not limited to the stationary state, and the user may intentionally shake the apparatus.

In particular, in recent years, autofocus (hereinafter, AF) cameras have become popular. An AF camera usually focuses on an object located at a position corresponding to a specific AF area. For example, the following procedure is generally used when shooting a still object (imaging in a still state). It is a target. (1-1) The camera is moved and the composition is changed (shake) so that the AF area matches the target subject. (1-2) By operating a half-push button or the like, AF driving including operations such as driving of a side distance and a lens is performed, and a focus is set on a target. (1-3) If necessary, change the composition while the focus is locked. (1-4) Shoot when the composition is determined.

[0010] Further, since the performance of AF has been improved, it has become possible to easily capture a moving object such as a running car. In this case, the following procedure is generally used as an imaging procedure. (2-1) Match the AF area to the target subject.
Since the subject is moving, the photographer intentionally shakes the camera in accordance with the subject. (2-2) The AF is driven by operating the half-press button or the like, and the focus is adjusted to the target. Again,
As in (2-1), the camera is shaken intentionally. (2-3A) Shooting with the camera shake (pan shot: when the subject keeps moving). (2-3B) The camera is stopped and photographing is performed with the camera held normally (when the movement of the subject stops).

In a composition change state such as panning or panning in which the user intentionally shakes the apparatus, the user tries to correct the shake caused by the movement desired by the user, and the operability and the feeling of use are very poor. There was a problem that it became something. To cope with this problem, a method for distinguishing between a composition change state and a stationary state is disclosed in, for example, Japanese Patent Laid-Open No. 05-1.
42614, JP-A-07-261234, JP-A-10-213832, and the like. In these conventional methods, by automatically discriminating the shake intended by the user and the unintended hand shake from the output of the shake detection sensor, the user is not discomforted or uncomfortable in the composition change state. Can be changed as intended.

However, it is difficult to cope with a condition in which the detection of a shake starts with an intended shake and then changes to an unintended hand shake. In a conventional method, for example, there is a method in which a DC component of a shake detection signal is cut, and the end of the intended shake is detected when the shake detection signal returns to around zero. However, if the detection signal is already large when the detection of the shake is started, it takes time for the signal to return to around 0 even if the DC component is cut, and the blur correction is not performed. There has been a problem that it is impossible or the accuracy is reduced.

In the prior art, the intended movement (shake) is completed when the signal returns to a specific value. For example, when the camera is swung right and left, the movement direction is changed. When it changes, it is determined that the intended movement has been completed, and an operation that gives the user discomfort or discomfort has occurred.

Further, in the prior art, the output is corrected as it is by using the output of the angular velocity sensor as it is, or the deflection is corrected by the data obtained by integrating to obtain the displacement angle. However, the accuracy of the shake correction was low regardless of the stationary state and the composition change state.

SUMMARY OF THE INVENTION An object of the present invention is to provide a shake detecting apparatus and a shake correcting optical apparatus capable of correcting a shake caused by a hand shake in a normally held state with a high degree of accuracy, which makes a user feel uncomfortable / uncomfortable. Without changing the composition, you can follow a moving subject without moving, and even if you shoot while following the subject, the shooting result will be improved or it will not adversely affect, and highly accurate blur correction even immediately after the composition change is completed At the start of the shake detection, with the intended shake,
An object of the present invention is to provide a shake detection device and a shake correction optical device that can correct shake with high accuracy even when a condition that changes to unintended hand shake occurs thereafter.

[0016]

The present invention solves the above-mentioned problems by the following means. In addition, in order to make it easy to understand, description is given with reference numerals corresponding to the embodiment of the present invention, but the present invention is not limited to this. That is, according to the first aspect of the present invention, a vibration detecting section (10a, 10b) for detecting a vibration and outputting a vibration detection signal (ω), and a reference value calculation for calculating a reference value (ω ₀ ) of the vibration detection signal. From the part (52), the vibration detection signal and the reference value,
A movement start detection unit (4) for detecting that a large movement has been started as compared to a substantially stationary state of the device including the vibration detection unit.
And 4) the movement start detecting unit is configured to start the movement when the vibration detection signal and / or the reference value outputs a low-frequency signal for a predetermined time or more. Is detected.

According to a second aspect of the present invention, a vibration detecting section (10a, 10b) for detecting a vibration and outputting a vibration detection signal (ω).
A reference value calculation unit (52) for calculating a reference value (ω ₀ ) of the vibration detection signal; and comparing the vibration detection signal and the reference value with a substantially stationary state of a device including the vibration detection unit. A movement start detection unit (44) for detecting that a large movement has started, wherein the movement start detection unit calculates a difference value between the vibration detection signal and the reference value, The difference value is the movement start reference time (T_S
TART) indicating that they are moving in the same direction,
And detecting the start of the movement when the difference value at the elapse of the movement start reference time (T_START) is equal to or greater than the movement start reference value (ω'START). .

According to a third aspect of the present invention, there is provided a vibration detecting section (10a, 10b) for detecting a vibration and outputting a vibration detecting signal, and a reference value calculating section (5) for calculating a reference value of the vibration detecting signal.
2) and a movement end detection unit (46) for detecting that a large movement has ended compared to the substantially stationary state of the device including the vibration detection unit, wherein the movement end detection The unit detects the end of the movement from a movement end determination value (ω ₀ fin) serving as a reference for detecting the end of the movement and the vibration detection signal (ω) and / or the reference value (ω ₀ ). Is a shake detection device.

According to a fourth aspect of the present invention, there is provided the first or second aspect.
, The movement end determination value (ω ₀ fin) serving as a reference for detecting the end of the movement and the vibration detection signal (ω) and / or the reference value (ω ₀ ). And a movement end detection unit (46) for detecting the movement end.

According to a fifth aspect of the present invention, in the shake detecting device according to the fourth aspect, the movement end detecting section (46) is configured to determine whether the vibration detection signal (ω) and / or the reference value (ω ₀ ) A state within a predetermined range (± ω ”fin) from the movement end determination value (ω ₀ fin) is a movement end reference time (T_
FINISH) detecting the end of the movement when the movement is continued.

The invention of claim 6 is the invention of claim 4 or claim 5.
Wherein a reference value storage unit (53) for sequentially storing a predetermined number of the reference values (ω ₀ ) is provided, and among the reference values stored in the reference value storage unit, And storing a provisional reference value based on a reference value past the time when the movement start detection unit (44) detects the movement.

The invention of claim 7 is the invention of claim 4 or claim 5.
Wherein a reference value storage unit (53) for sequentially storing a predetermined number of the reference values (ω ₀ ) is provided, and from among the reference values stored in the reference value storage unit,
A shake detection device characterized by storing an oldest reference value as a provisional reference value when the movement start detection section (44) detects the movement.

The invention of claim 8 is the invention of claim 6 or claim 7.
, The movement end detection unit (46) sets the provisional reference value to the movement end determination value (ω
₀ fin).

[0024] The invention of claim 9 is the invention of claims 6 to 8.
In the shake detection device according to any one of the above, the movement end detection unit (46) is configured to determine an absolute value of a difference between the provisional reference value and the vibration detection signal (ω) or the reference value (ω ₀ ). The state where the absolute value of the difference is equal to or less than the movement end threshold (ω ″ fin) is determined as the movement end reference time (T_F).
ISH) detecting the end of the movement when the movement is continued.

According to a tenth aspect of the present invention, in the shake detecting apparatus according to any one of the fourth to ninth aspects,
The reference value calculation section (52) calculates a moving average of the vibration detection signal as a reference value (ω ₀ ), and uses the calculated value when the movement is detected, rather than in the substantially stationary state. A vibration detection apparatus characterized in that the number of data of the vibration detection signal is reduced.

According to an eleventh aspect of the present invention, in the shake detecting apparatus according to any one of the fourth to ninth aspects,
The reference value calculation unit (52) includes a low-pass filter, and when the movement is detected, sets a cutoff frequency used for the low-pass filter to be higher than that in the calculation in the substantially stationary state. This is a shake detection device that is a feature.

According to a twelfth aspect of the present invention, in the shake detecting device according to any one of the third to eleventh aspects, the movement end detecting unit is configured to perform a predetermined time (T_PAN)
If the end of the movement is not detected even after the lapse of the above, the movement end determination value (ω ₀ fin) is updated.

[0028] The invention of claim 13 is the invention of claims 3 to
13. The shake detecting device according to any one of items up to 12.
The end-of-movement detecting section (46) determines the end of the movement.
Value (ω ₀fin) and the vibration detection signal or the reference value
The absolute value of the difference is calculated, and the absolute value of the difference ends moving.
The state larger than the threshold (ω "fin) is the movement determination time
(T_PAN) When the movement is continued, the movement end determination value (ω
₀fin) is updated.
It is.

According to a fourteenth aspect of the present invention, in the shake detecting device according to the twelfth or thirteenth aspect, the movement end detecting section (46) detects that the movement start detecting section (44) has started the movement. And determining whether or not to update the movement end determination value (ω ₀ fin) based on the calculation duration of the reference value (ω ₀ ) calculated before the movement detection.

According to a fifteenth aspect of the present invention, in the shake detecting device according to the fourteenth aspect, the reference value calculation control section (52).
Is the update reference time (T_CAL2)
If it is longer, the movement end determination value (ω ₀ fin)
Is not updated, and when the calculation continuation time is shorter than the update reference time (T_CAL2), the movement end determination value (ω ₀ fin) is updated.

According to a sixteenth aspect of the present invention, in the shake detecting device according to any one of the twelfth to fifteenth aspects, the movement end detecting unit (46) is configured to update the vibration detection signal (ω ) Or the reference value (ω ₀ ) is updated as a new movement end determination value (ω ₀ fin).

According to a seventeenth aspect of the present invention, in the shake detecting device according to any one of the third to sixteenth aspects, after the movement end detecting section (46) detects the end of the movement, A shake detection device comprising a reference value calculation control unit (48) for controlling a reference value calculation method of the reference value calculation unit (52).

According to an eighteenth aspect of the present invention, in the shake detecting device according to the seventeenth aspect, the reference value calculation control section (48).
Is a control device for controlling whether or not the provisional reference value is used for a reference value calculation.

According to a nineteenth aspect of the present invention, in the shake detecting device according to the seventeenth or eighteenth aspect, the reference value calculation control unit (48) controls the movement start detection unit (44) to start the movement. The reference value (ω
₀ ), the movement end detection unit (4)
6) controls the method of calculating the reference value (ω ₀ ) after detecting the end of the movement.

According to a twentieth aspect, in the shake detecting device according to the nineteenth aspect, the reference value calculation control section (48).
Indicates that the operation continuation time is the operation control reference time (T_CAL)
If it is longer than 1), the provisional reference value is used for reference value calculation, and the calculation continuation time is calculated by the calculation control reference time (T
_CAL1), the provisional reference value is controlled so as not to be used for the reference value calculation.

According to a twenty-first aspect of the present invention, in the shake detecting device according to the twentieth aspect, the reference value calculation control section (48)
Is the case where the provisional reference value is used for the reference value calculation, and immediately after the movement end detection unit (46) detects the end of the movement, the reference value calculation unit calculates the provisional reference value. Calculating a reference value (ω ₀ ) that reflects the value of the vibration detection signal (ω) at that time as a reference value (ω ₀ ) and gradually elapses thereafter;
Is a shake detection device.

According to a twenty-second aspect of the present invention, in the shake detecting device according to the twentieth aspect, the reference value calculation control section (48)
Is the case where the provisional reference value is not used for the reference value calculation, and immediately after the movement end detection unit (46) detects the end of the movement, the reference value calculation unit (52) uses it for the calculation. The number of samples of the vibration detection signal (ω) is reduced, and the number of samples of the vibration detection signal used for calculation by the reference value calculation unit is gradually increased as time elapses thereafter. This is a shake detection device.

According to a twenty-third aspect of the present invention, in the shake detecting device according to any one of the first to twenty-second aspects, the movement is a movement intentionally performed by a user. This is a shake detection device.

According to a twenty-fourth aspect of the present invention, there is provided a shake detecting apparatus according to any one of the first to twenty-third aspects, a blur correction optical system (80) for correcting a blur of an image, and the blur correction optical system. A driving unit (70a, 70b) for driving a system, a driving signal is calculated from the vibration detection signal (ω) and the reference value (ω ₀ ), and the driving result is output as a driving signal to drive the driving unit And a control unit (50) for controlling.

According to a twenty-fifth aspect of the present invention, in the blur correcting optical apparatus according to the twenty-fourth aspect, the blur correcting optical system (8
0) and a photographing unit (170) for photographing an image formed by the photographing optical system.

[0041]

Embodiments of the present invention will be described below in more detail with reference to the drawings. (First Embodiment) FIG. 1 is a block diagram for explaining an outline of a shake detecting apparatus and a blur correction optical apparatus according to a first embodiment of the present invention. In the present embodiment, a description will be given by taking a shake correction camera using a silver halide film as an example.

(Outline of Camera for Shake Correction) The half-press switch SW1 is a switch that is turned on in conjunction with a half-press operation of a release button (not shown). When the half-press switch SW1 is turned ON, a series of photographing preparation operations such as photometric calculation by a photometric unit (not shown) and autofocus driving by an autofocus driving unit (not shown) are started. When the half-press timer 100 is OFF, the half-press timer 100 turns ON in synchronization with the ON of the half-press switch SW1.

The full-press switch SW2 is turned on in conjunction with a full-press operation of further pressing the release button. When this switch is turned on,
A series of photographing operations such as an operation of raising the mirror 130, opening and closing of a shutter by a shutter mechanism (not shown), operation of lowering the mirror 130, and winding of the film 140 by the feeding motor 150 are performed.

The half-pressing timer 100 is turned on at the same time as the half-pressing switch SW1 is turned on, and remains on while the half-pressing switch SW1 is on. Is a timer that remains ON for a certain period of time. The half-pressing timer 100 starts counting at the same time as
During N, the counting is continued.

The power supply unit 110 supplies power to each unit of the camera, here, the angular velocity sensor 10.
While the half-press timer 100 of the camera is ON, the angular velocity sensor 10 is started, and power is continuously supplied to the power required in the camera system. Also, half-press timer 10
When 0 is OFF, the power supply unit 110 stops supplying power to the angular velocity sensor 10 and the like. Therefore, as long as the half-press timer 100 of the camera is ON, the angular velocity sensor 1
0 enables camera vibration detection.

The angular velocity sensor 10 is a sensor that detects the vibration applied to the camera as an angular velocity value. The angular velocity sensor 10 detects the angular velocity using the Coriolis force applied to the angular velocity sensor 10, and outputs the detection result as a voltage signal. Department. The angular velocity sensor 10a is a sensor that detects angular blur in the X-axis direction in the figure, and the angular velocity sensor 10b is a sensor that detects angular blur in the Y-axis direction in the figure. In this way, by arranging the angular velocity sensors 10a and 10b in different axial directions, it becomes possible to detect the camera vibration two-dimensionally. The voltage signal output by the angular velocity sensor 10 is transmitted to the amplifier 20. Note that the angular velocity sensor 1
0 indicates that the angular velocity can be detected only while the power is supplied from the power supply unit 110.

The amplification section 20 is an amplification section for amplifying the output of the angular velocity sensor 10. Generally, since the output from the angular velocity sensor 10 is small, even if it is digitized by the A / D converter 30 and processed in the microcomputer 90, the resolution of the angular velocity value is too low (the angular velocity value per bit is large). And the vibration cannot be accurately detected, and the accuracy of blur correction cannot be improved. Therefore,
Before inputting to the A / D converter 30, the angular velocity signal is amplified. Then, the resolution of the angular velocity value in the microcomputer 90 is increased (the angular velocity value per bit is reduced).
This can increase the accuracy of the shake correction. The amplification unit 20 is provided with two amplification units 20a and 20b corresponding to the angular velocity sensors 10a and 10b, respectively. Also, here, a low-pass filter may be added for the purpose of reducing high-frequency noise included in the sensor output as well as amplifying the signal. Amplifier 2
Angular velocity signal amplified by 0 (hereinafter, shake detection signal)
Is transmitted to the A / D converter 30.

The A / D converter 30 is a converter for converting an analog signal into a digital signal. In the present embodiment, A
/ D converters 30a, 30b and A / D converters 30c, 3
0d is provided. A / D converters 30a, 30b
Is a converter for converting the analog shake detection signal sent from the amplifier 20 into a digital signal. By converting the shake detection signal into a digital signal, arithmetic processing in the microcomputer 90 becomes possible. The converted shake detection signal is input to the drive signal calculation units 50a and 50b and the composition change detection units 40a and 40b. A / D converter 30c,
Reference numeral 30d denotes a converter for converting the blur correction lens position information (analog signal) sent from the drive unit into a digital signal. The converted blur correction lens position information is transmitted to the drive signal calculation units 50a and 50b.

In this embodiment, the A / D converter 30
Is premised on using the one built in the microcomputer 90. However, the present invention is not limited to this example, and an A / D converter separate from the microcomputer 90 may be used. Further, in the present embodiment, A / A is set so as to correspond to the amplification units 20a and 20b.
Although two A / D converters of the D converters 30a and 30b are provided, the conversion operation may be temporally distributed by using one A / D converter. For example, after the signal of the amplifier 20a is converted, the signal of the amplifier 20b is converted, and thereafter, the amplifier 20a, the amplifier 20b, the amplifier 20a. . . May be repeated. This is the same for the A / D converters 30c and 30d.

The composition change detecting section 40 includes an A / D converter 30
And a drive signal calculation unit 50
This is a part for detecting the state of camera shake from the reference value or the like calculated in step (1). The shake state is a state in which the composition is not changed at all and the photographer vibrates due to a shake not intended by the photographer (stationary state), and a state in which the photographer is shaken by the photographer's intention such as panning (composition change state) ), The composition change detection unit 40 detects which state the camera is in. This detection result is transmitted to the drive signal calculation unit 50,
The drive signal calculation unit 50 changes the calculation method according to the result. Details of the internal configuration of the composition change detection unit 50 will be described later with reference to FIG. 2, and details of the operation will be described with reference to FIG.
This will be described with reference to FIGS.

The drive signal calculation section 50 is provided with the A / D converter 30
Is a calculation unit that calculates a drive signal for driving the shake correction lens 80 from the shake detection signal and the shake correction lens position information transmitted from the camera, and outputs the drive signal. First,
A reference value is calculated from the shake detection signal, and the reference value is subtracted from the shake detection signal value. By integrating this, the angular velocity signal is converted into an angular displacement signal, and the target drive position of the blur correction lens 80 is calculated by taking into account various conditions such as the focal length of the entire imaging optical system (not shown). Further, the drive signal calculation unit 50 calculates the calculated target drive position information and the drive unit 7.
It is also a control unit that calculates a drive signal from the position information of the shake correction lens 80 sent from 0 and performs control. The drive signal calculation unit 50 changes the drive signal calculation method according to the detection result (stationary state or composition change state) of the composition change detection unit.

In the present embodiment, the drive signal calculation units 50a,
Two drive signal calculation units 50b are provided. However, it is also possible to combine them into one to temporally distribute the drive signal calculation operation. For example, after calculating a drive signal for a signal in the X-axis direction, a drive signal is calculated for a signal in the Y-axis direction, and then X, Y, X, Y. . . Alternatively, the driving signal may be calculated alternately. Details of the internal configuration of the drive signal calculation unit 50 will be described separately with reference to FIG.

The D / A converter 60 is a D / A converter for converting the drive signal (digital signal) calculated by the drive signal calculator 50 into an analog signal. The converted analog signal is transmitted to the driving unit 70. In the present embodiment, it is assumed that the D / A converter 60 is built in the microcomputer 90. However, the present invention is not limited to this, and the D / A converter 60 is separate from the microcomputer 90. A converter may be used. Further, in the present embodiment, the D / A converter 60a,
60b, two D / A converters are provided.
A single converter may be used to temporally distribute the conversion operation. For example, after converting the signal in the X-axis direction, the signal in the Y-axis direction is converted, and then X, Y, X,
Y,. . . May be converted.

The driving section 70 is a driving section for driving the blur correction lens 80 based on the driving signal (analog signal) transmitted from the D / A converter 60. The drive unit 70 includes an actuator for driving the shake correction lens 80,
It has a position detection sensor for detecting the position of the shake correction lens 80 and the like. The output of the position detection sensor is transmitted to the drive signal calculation unit 50 via the A / D converter 30. Since it is necessary to drive the blur correction lens 80 in the two-dimensional direction, it is necessary to provide two driving units 70a and 70b.

The blur correction lens 80 is a part of an imaging optical system (not shown) built in the lens barrel 160 of the photographing apparatus, and is a single lens or a plurality of lenses that can move in a plane substantially perpendicular to the optical axis I. This is a blur correction optical system composed of two lenses. The blur correction lens 80 is driven by the drive unit 70 in a direction substantially orthogonal to the optical axis I, and deflects the optical axis I of the imaging optical system.

The blur of an image such as a photograph is caused by vibration applied to the camera such as camera shake or the like.
This is caused by the movement of the image of the (0 plane). However, in a shake correction camera as shown in FIG. 1, a vibration detection sensor such as an angular velocity sensor 10 is built in.
The vibration applied to the camera can be detected by the vibration detection sensor. When the vibration applied to the camera is detected, the motion of the image on the image plane due to the vibration can be known. Therefore, the motion compensation lens 80 is moved so that the motion of the image on the image plane stops. Thereby, the movement of the image on the image forming plane, that is, the blur can be corrected.

The microcomputer 90 includes an A / D converter 30, a composition change detection unit 40, a drive signal calculation unit 50, and a D / A converter 6
This is a microcomputer in which 0 and the like are incorporated. In addition to the operations described here, the microcomputer 90 may also perform control such as auto-focus driving (not shown).

The mirror drive motor 120 is connected to the power supply 1
Power is supplied from the mirror 10 and the mirror 130 is
This motor performs up and down operations. In the present embodiment, an example in which the mirror driving motor 120 uses a motor that is an electromagnetic actuator has been described. However, the present invention is not limited to this, and mechanical means such as a spring may be used.

The mirror 130 is a mirror that deflects light from an imaging optical system (not shown) and sends it to a pentaprism and a finder (not shown). During the exposure operation, the mirror is raised, and the light from the imaging optical system reaches the surface of the film 140.

The film 140 is a film for recording an image formed by the image forming optical system. In the present embodiment, it is assumed that the camera is a silver halide camera. However, the present invention is not limited to this, and an area sensor such as a CCD or a C-MOS sensor may be used.

The feed motor 150 is a motor for feeding the frame of the film 140 after the exposure is completed. When an area sensor such as a CCD is used instead of the film 140 for the image pickup medium, the motor itself is unnecessary.

The camera body 170 is a camera body of a single-lens reflex camera having a photographing section and having a replaceable lens barrel 160. In the present embodiment, an example is shown in which the camera is a single-lens reflex camera. However, the present invention is not limited to this. For example, a non-interchangeable lens such as a compact camera may be used.

FIG. 2 is a diagram showing the internal configuration of the drive signal calculation section 50 and the composition change detection section 40. Since the contents described hereinafter are the same in both the X direction and the Y direction, the directions will not be specifically described, and will be described collectively. The drive signal calculation section 50 includes a reference value calculation section 52, a reference value storage section 53, an integration calculation section 54, and a drive signal calculation section 55.

The reference value calculation section 52 is provided with the A / D converter 30
A calculation unit that calculates a reference value for drive signal calculation from the shake detection signal transmitted from (30a, 30b). Normally, the reference value in the stationary state may be an output value (hereinafter, zero output) in a state in which the angular velocity sensor 10 is completely stationary. However, since this zero output value fluctuates due to environmental conditions such as drift and temperature, if the reference value is set to a fixed value, the accuracy of blur correction may decrease or unnatural behavior may occur. . Therefore, it is desirable to calculate the reference value from the actual use state, that is, the signal of the camera shake of the photographer, to obtain the zero output. In the present embodiment as well, the reference value calculation unit calculates the reference value from the shake detection signal. 52 are provided. The reference value calculation unit 52 changes the calculation formula of the reference value between the stationary state and the composition change state. An example is shown below.

[0065]

(Equation 1)

Here, ω is a shake detection signal, and ω ₀
Is a reference value of the shake. Further, the suffix t attached to these variables is a variable representing the elapsed time, and in the present embodiment, is represented by the number of times of sampling, and is an integer value. Each of these equations represents a moving average of the shake detection signal, but the number of data used for averaging differs between the stationary state and the composition change state.

It is desirable that the reference value in the stationary state is a value close to the zero output value of the angular velocity sensor 10. The frequency of the zero output signal of the angular velocity sensor 10 is much lower than the frequency of the hand shake of a human. Therefore, the reference value may be obtained by extracting the low frequency component of the shake detection signal. Therefore, the reference value of the shake detection signal is calculated by calculating the shake detection signal, that is, the moving average of the shake. In order to extract only low frequency components as much as possible, the number K0 of data used for the moving average is increased.

As for the reference value in the composition change state, the shake detection signal fluctuates more greatly than in the stationary state. Since the composition change state has a shake intended by the photographer, it is not preferable to correct even the intended shake. Therefore, in the composition change state, the response of the reference value to the shake detection signal is made faster in order to suppress blur correction. Here, the number K1 of data used for the moving average is set to K0 in the case of the stationary state.
Less than that. By doing so, the response of the reference value can be made faster, and the blur correction operation can be suppressed as compared with the stationary state, and as a result, the composition can be determined as intended for shooting. .

The reference value calculated by the reference value calculator 52 is
It is transmitted to the drive signal calculation unit 50. The calculation of the reference value is not limited to the moving average as described above, and a low-pass filter such as an FIR filter or an IIR filter may be used.
In this case, the cutoff frequency in the stationary state may be set lower than the cutoff frequency in the composition change state.

The reference value storage section 53 is a memory for storing a predetermined number of reference values calculated by the reference value calculation section 52. Until the memory becomes full, the calculated reference values are sequentially stored. After the memory is full,
Of the stored reference values, the oldest one (hereinafter, the oldest reference value) is deleted from the memory and replaced with the latest reference value. The operation of the reference value storage unit 53 will be described with reference to FIG.

The integration operation unit 54 is an operation unit that integrates the shake detection signal (angular velocity), converts it into shake angle information, and further calculates the target drive position of the shake correction lens. An example of the calculation performed by the integration calculation unit 54 is described below.

[0072]

(Equation 2)

Each symbol in equation (3) is θ (t): target drive position, ω (t): shake detection signal, ω ₀ (t): reference value, t: time (integer value), C is a constant determined by conditions such as the focal length of the lens.

The drive signal calculator 55 calculates the target drive position calculated by the integration calculator 54 and the drive unit 70 from the A / D converter 3.
This is a part for calculating a signal for driving the blur correction lens 80 from the position of the blur correction lens 80 transmitted via 0 (30c, 30d). Generally, PID control is performed in which a deviation between a target drive position and a position of a blur correction lens is calculated, and a driving signal is calculated by adding a term proportional to the deviation, a term proportional to the integral of the deviation, and a term proportional to the derivative of the deviation. It is. The method of calculating the drive signal is not limited to the PID control, but may be another method.

The composition change detection section 40 has a composition change start detection section 44 and a composition change end detection section 46. The composition change start detection unit 44 is a movement start detection unit that detects the start of a composition change state due to the photographer's intention. The start of the composition change state is detected using the shake detection signal sent from the A / D converter 30 (30a, 30b) and the reference value sent from the reference value calculation unit 52. The composition change start detection unit 44 operates only when the shake state is the stationary state, and does not operate when the composition change state is set. When the start of the composition change state is detected, the oldest data (oldest reference value) among the reference values stored in the reference value storage unit 53 is stored as the provisional reference value, and this is changed to the end of the composition change. Transmit to the detection unit. When detecting the start of the composition change state, the composition change start detection unit 44 transmits the information to the reference value calculation unit 52, and the reference value calculation unit 52 changes the calculation method of the reference value. In the present embodiment, the reference value calculated by Expression (1) is switched to be calculated by Expression (2).

The composition change end detecting section 46 is a movement end detecting section for detecting the end of the composition change by the photographer's intention. The composition change end detection section 46 is a composition change start detection section 4
4 based on the oldest reference value sent from No. 4 and the reference values sequentially calculated by the reference value calculation unit 52. The composition change end detection unit 46 operates in the composition change state, and does not operate in the stationary state. Upon detecting the end of the composition change state, the composition change end detection unit 46 transmits the information to the reference value calculation unit 52, and
2 changes the method of calculating the reference value. In this embodiment,
The reference value calculated by equation (2) is switched to be calculated by equation (1).

(Operation of Image Stabilizing Camera) Next, the operation of the image stabilizing camera in this embodiment will be described. FIG.
5 is a flowchart illustrating a flow of the entire camera system including the shake detection device according to the embodiment.

In step (hereinafter referred to as S) 10, it is determined whether or not the half-press switch SW1 is ON. If it is ON, the process proceeds to S20, and if it is OFF, the process proceeds to S160.

In S20, the counter Tsw1 is reset, and the count value is set to 0 (Tsw1 = 0). The counter Tsw1 is a counter for measuring an elapsed time from when the half-press switch SW1 is turned off, and the count value is an integer. The counter Tsw1 remains at 0 while the half-press switch is ON, and operates only while the half-press switch is OFF and the half-press timer 100 is ON.

At S30, the half-press timer 100 is turned off.
It is determined whether it is F or not. If it is OFF, the process proceeds to S40, and if it is ON, the process proceeds to S190.

In S40, the counter t is reset, and the count value is set to 0 (t = 0). The counter t is
This is a counter that measures the time during which the half-press timer 100 is ON. This counter t is an integer value counter, and starts a counting operation at the same time when the half-pressing timer 100 is turned on, and continues the counting operation while the half-pressing timer 100 is on.

In S60, the detection result of the composition change detection unit 40 is set to a stationary state.

In S70, the half-press timer 100 is turned on.
To In S80, the angular velocity sensor 10 is turned on, and detection of vibration is started. In addition, the conversion operation by the A / D converter 30 is also started here.

In S90, since the stationary state is set in S60, the calculation of the reference value for the stationary state is started. In the present embodiment, the reference value is calculated by Expression (1). S1
In 00, since the stationary state is set in S60, the composition change detection unit 40 starts an operation for detecting the start of the composition change state.

In S110, the drive signal calculation section 50 starts calculation of the drive signal. In S120, the drive unit 70 starts driving the blur correction lens 80 based on the drive signal obtained from the drive signal calculation unit 50.

In S130, the full-press switch SW2 is set to O
It is determined whether or not N. Full-press switch SW2 is O
In the case of FF, the process proceeds to S150. Full-press switch SW2
Is ON, the process proceeds to S140.

In S140, the mirror 130 is raised, the shutter is opened and closed, the mirror 130 is lowered, and the feeding motor 15
A photographing operation such as driving of 0 is performed.

In S150, the counter t of the half-press timer 100 is incremented by one (calculation of t = t + 1 is performed).

In S160, the half-pressing timer 100 is
It is determined whether or not N. Half-press timer 100 is O
If N, proceed to S170, half-press timer 100 is turned off
If F, the process returns to S10, and the detection of the half-press switch is continued.

At step S170, at the point when the process proceeds to this step, the camera is in a state where the half-press switch SW1 is OFF and the half-press timer 100 is ON. In order to measure the time during which this state continues, a counter Tsw1 is used.
Is advanced by one (the operation of Tsw1 = Tsw1 + 1 is performed).

In S180, it is determined whether the value of the counter Tsw1 is smaller than the threshold value T_SW1. Here, the threshold value T_SW1 is a constant for determining the upper limit of the counter Tsw1, and determines the time from when the half-press switch is turned off until the half-press timer turns off. If the counter Tsw1 is less than the threshold, that is, if the determination is affirmative, the half-press timer 100
Does not turn OFF, and proceeds to S190. On the other hand, the counter T
When sw1 becomes equal to the threshold value T_SW1,
That is, if a negative determination is made in this step, S2
Proceeding to 80, a process for turning off the half-press timer 100 (S310) and a process associated with turning off the half-press timer 100 (S290, S300) are performed.

In S190, the angular velocity sensor 10 is turned on.
And the detection of shake is continuously performed. Also, A
The conversion operation by the / D converter 30 also continues.

In S200, the detection result of the composition change detection unit 40 is monitored. If the detection result is stationary, S21
0, and if the composition is changed, the process proceeds to S230.

In S210, the reference value calculation section 52 calculates a reference value in a stationary state. In the present embodiment, the reference value is calculated by Expression (1).

In step S220, when proceeding to this step, the camera is in a stationary state, so the composition change start detecting section 44
Performs an operation for detecting the start of the composition change state. Here, when the start of the composition change state is detected, the composition change detection unit 40
From the stationary state to the composition change state. Details of this calculation will be described with reference to FIG.

In S230, the reference value calculation section 52 calculates a reference value in the composition change state. In this embodiment,
The reference value is calculated by equation (2).

In S240, when the process proceeds to this step, the composition is in the composition change state, so that the composition change end detection unit 46 performs an operation for detecting the end of the composition change state. Here, when the end of the composition change state is detected, the detection result of the composition change detection unit 40 is changed from the composition change state to the stationary state. Details of this calculation will be described with reference to FIG.

In S250, the reference value is stored in the reference value storage unit 53. Details of this operation will be described with reference to FIG.

In S260, the reference value calculated in S210 or S230, the shake detection signal, and the blur correction lens 80
The drive signal calculation unit 50 calculates a drive signal based on the position information.

In S270, the drive section 70 drives the blur correction lens 80 based on the drive signal calculated in S260.

In S280, the composition change detection unit 40 stops the calculation for detecting the start or end of the composition change state.

In S290, the reference value calculation unit stops calculating the reference value.

In S300, the supply of power to the angular velocity sensor 10 is stopped, and the angular velocity sensor 10 is turned off.

In S310, the half-pressing timer 100 is turned off.
The process returns to S10 with FF, and the state of the half-press switch SW1 is detected.

(Operation for Saving Reference Value) FIG. 4 is a flowchart showing the flow of the operation of the reference value storage unit 53. The reference value storage unit 53 stores the reference value calculation unit 5 as described above.
This is a memory for storing a predetermined number of reference values calculated by 2, and in this embodiment, M reference values can be stored. The data stored in the reference value storage unit 53 is all cleared when the half-press timer 100 is turned on, that is, at the same time as the start of shake detection. However, this may be synchronized with the half-pressing timer 100 being turned off, that is, the end of the shake detection. The storage of the reference value in the reference value storage unit 53 is controlled using a variable k indicating the address in the reference value storage unit 53. The variable k is returned to the beginning (k = 0) when the half-press timer 100 is turned on (or may be turned off). Also, the variable k advances over time,
When it reaches the end of the memory, it returns to the beginning again. Based on the above contents, the flow of operation of the reference value storage unit 53 will be described.

In S400, it is determined whether or not the counter t of the half-press timer 100 is larger than the number M of reference values that can be stored in the reference value storage unit 53. When the determination is affirmative, the process proceeds to S410, and when the determination is negative, the process proceeds to S420. Each time the counter t advances by one, the reference value storage unit 53
Save the two reference values. When the counter t is less than M, there is room for storing the reference value in the reference value storage unit 53, so that the reference value may be stored in the reference value storage unit 53 as it is. On the other hand, if the counter t is larger than M, the reference value storage unit 53 has already been filled with the reference value. Therefore, in that case, the reference value storage unit 53
Of the reference values stored in, the oldest reference value is discarded and the latest reference value is stored.

In S410, the time when the process proceeds to this step
Then, at the k-th address in the reference value storage 53,
Old reference value ω₀(t−M) is stored. Latest standards
Value ω ₀To save (t) this oldest reference value reference value
ω₀Discard (t−M).

In S420, the latest reference value ω ₀ (t) is stored at the k-th address.

In S430, the address k in the reference value storage unit 53 is advanced by one for the storage processing in the next sampling.

In S440, it is determined whether or not the address k is equal to or larger than M. If it is not less than M, the process proceeds to S450, and the address is returned to the beginning (k = 0). If k is smaller than M, the process proceeds to S260.

In S450, k = 0 and the address is returned to the head. As described above, while the shake is being detected, the address in the reference value storage unit 53 is k =
.., M-2, M-1, 0, 1, 2,.
.., M-2, M-1, 0, 1, 2,... Here, it is assumed that one reference value is stored for each sampling. However, the present invention is not limited to this. For example, an average value of a plurality of reference values may be calculated and stored. .

(Detection of Start of Composition Change State) FIG. 5 is a flowchart showing the flow of operation of the composition change start detection unit 44, and is a diagram showing the operation in S220 of FIG. 3 in detail. In S500, the reference value ω ₀ (t) is subtracted from the shake data ω (t) to calculate the detected angular velocity ω ′ (t). ω ′ (t) = ω (t) −ω ₀ (t) Equation (4)

In S510, it is determined whether or not the sign of the angular velocity detection value ω '(t) calculated in S500 has changed from the angular velocity detection value ω' (t-1) at the time of the previous sampling. As an example of the determination method, the previous angular velocity detection value ω ′
There is a method in which (t-1) is multiplied by the current angular velocity detection value ω '(t), and if the sign is positive, the sign is unchanged, and if the sign is negative, the sign is changed. If the determination is affirmative (the sign is unchanged) in S510, the process proceeds to S520, and if the determination is negative (the sign is changed), the process proceeds to S570.

At S520, the counter Tstrt is advanced by one. This counter Tstrt is used for detecting the angular velocity detection value ω ′.
Is a counter for measuring the time during which the state in which the sign of? Does not change continues, and is reset to 0 when the half-press timer 100 is ON or when the sign of? 'Changes. The fact that the sign of the angular velocity detection value ω ′ does not change is equivalent to the fact that the camera is swung in one direction.

In S530, it is determined whether or not the counter Tstrt is equal to or greater than a threshold value T_START. Here, the threshold value T_START is a movement start reference time set to determine the start of movement. Affirmative decision (T
If (strt ≧ T_START), the process proceeds to S540,
If negative determination (Tstrt <T_START), S5
Go to 80.

In S540, it is determined whether or not the absolute value of the detected angular velocity ω ′ is equal to or larger than the threshold ω′START. Here, the threshold value ω'START is a movement start reference value set to determine the start of the composition change state. If the determination is affirmative (| ω '|>ω'START), S
Proceeding to 550, it is detected that the composition change state has been started, and the accompanying processing is performed. Negative judgment (| ω '| ≦ ω'
If it is (START), the process proceeds to S570 and the counter Tst
Reset rt to 0.

In general, the detected angular velocity ω ′ in the composition change state has a larger amplitude and a lower frequency than the detected angular velocity ω ′ at rest. Therefore, the composition change start detection unit 44
First, it is determined whether or not the data has a low frequency by measuring the time during which the state in which the detected angular velocity ω ′ is unchanged continues (S510, 520). Further, it is determined whether or not a signal having a large amplitude is input by determining whether or not the detected angular velocity value ω ′ is larger than a predetermined threshold value ω ′ START (S540). As described above, the start of the composition change state is detected by monitoring the frequency and the magnitude.

In S550, among the reference values stored in the reference value storage unit 53, the oldest one (ω ₀ (t−M):
(The oldest reference value) is stored as the composition change end determination value ω ₀ fin. The composition change end determination value ω ₀ fin is a movement end determination value that is used later by the composition change end detection unit 46 to detect the completion of the composition change.

In S560, the judgment is changed to the composition change state. In S570, the counter Tstrt is reset to 0. In S580, the stationary state is continued.

(Detection of End of Composition Change State) FIG. 6 is a flowchart showing the flow of the operation of the composition change end detection unit 46, and is a diagram showing the operation in S240 of FIG. 3 in detail.

In S600, a composition change angular velocity value ω ″ for detecting the end of the composition change state is calculated. The composition change angular velocity value ω ″ is a reference value ω ₀ and a composition change end determination value ω _0.
This is the difference from fin, and is expressed as follows in the equation. ω ″ (t) = ω ₀ (t) −ω ₀ fin Equation (5) Here, since the composition is being changed, the reference value ω ₀ in Equation (5) is calculated by Equation (5) in the present embodiment. This is a value calculated by 2).

In S610, it is determined whether or not the absolute value of the composition change angular velocity value ω ″ calculated in S600 is smaller than the threshold value ω ″ fin. Here, the threshold value ω ″ fin
Is a movement end threshold set to determine whether or not the shake has converged. Affirmative determination (| ω "| <ω" fi
If n), the reference value ω ₀ is the composition change end determination value ω ₀ f
Since it can be determined that the value is close to “in”, S620
If the determination is negative (| ω ″ | ≧ ω ″ fin),
The reference value ω ₀ is changed from the composition change end determination value ω ₀ fin to ω ″ f
Since the value is more than “in” apart, the process proceeds to S650.

In S620, the counter Tfin is incremented by one (Tfin = Tfin + 1). This counter Tfi
n is the absolute value of the composition change angular velocity value ω ″ is the threshold value ω ″ f
This is a counter for measuring the time during which the state smaller than “in” continues.
0 when a negative determination is made at 0 (| ω ″ | ≧ ω ″ fin)
Is reset to

In S630, it is determined whether or not the counter Tfin is equal to or larger than a threshold value T_FINISH. Here, the threshold value T_FINISH is a movement end reference time set to determine the end of the composition change. If the determination is affirmative (Tfin ≧ T_FINISH), the process proceeds to S640, and the detection result of the composition change detection unit 40 is set to the stationary state. If the determination is negative (Tfin <T_FINISH), S660 is performed.
Then, the detection result of the composition change detection unit 40 is continued in the composition change state.

The stored composition change end determination value ω₀fi
n is a value substantially equal to the zero output value of the angular velocity sensor 10
It is believed that there is. In the composition change state, the camera
This composition change end determination value ω₀fin
Near reference value ω₀Is not very common,
When the change of the drawing is completed and the vehicle is stopped again, the reference value ω ₀
Is the composition change end determination value ω₀close to fin
And that state continues.

Thus, it is determined whether or not the reference value ω ₀ is close to the composition change end determination value ω ₀ fin (S6).
10), the continuation time of the state where the values are close to each other is measured (S620, 630), and the result is calculated as a predetermined time (T_F
(INISH) By detecting the continuation (S630), the end of the composition change state is detected.

In S640, the detection result of the composition change detection unit 40 is set to the stationary state. In S650, the counter Tfi
Reset n to zero. In S660, the detection result of the composition change detection unit 40 is continued in the composition change state.

Here, the composition change end determination value ω ₀ fin
And by determining the composition change velocity value omega "which is the difference between the reference value omega _0, although detecting the end of the composition change state,
Instead, the difference between the composition change end determination value ω ₀ fin and the shake detection signal ω may be determined. However, since the shake detection signal ω is the hand shake of the user, a signal having a relatively high frequency is also included. Then, the composition change end determination value ω ₀ fin and the shake detection signal ω
Is not easily settled within the threshold ω ″ fin,
There is a possibility that the detection of the end of the composition change state is delayed. However, if the reference value ω ₀ is used, the high frequency component of the shake detection signal ω is attenuated, so that such a phenomenon hardly occurs. Therefore, it is desirable to use the reference value ω ₀ for detecting the completion of the composition change.

(Shake Correction in the Still State Immediately After the End of the Composition Change State) The reference value calculation unit 52 calculates the reference value ω ₀ using the slightly earlier shake detection signal ω. ], Immediately after the composition change state ends, the shake detection signal ω due to a large shake in the immediately preceding composition change state is used. Therefore, immediately after the completion of the composition change state, the reference value ω ₀ obtained by the calculation according to the equation (1) may not be close to the zero output value of the angular velocity sensor 10, and in this case, the reference value ω _It takes a long time before the precision of ₀ becomes high. Therefore,
In the present embodiment, the composition change end determination value ω ₀ fin stored when the start of the composition change state is detected is used for the reference value calculation immediately after the completion of the composition change state, thereby allowing the composition change state to be obtained immediately after the completion of the composition change state. Even if there is, high-precision blur correction was made possible. Reference value ω ₀ immediately after the end of the composition change state
An example of the calculation method is shown below.

[0130]

(Equation 3)

Here, t ′ represents the time when the end of the composition change is detected. ω ₀ ′ is a reference value immediately after the completion of composition change detection. ω ₀ is a reference value calculated by the equations (1) and (2). W is a function of tt ′ and has the following features. (1) W (tt ′ = 0), that is, the value at the time when the composition change end is detected is 1. (2) It is a monotonically decreasing function of tt ′, where 0 ≦ W ≦ 1. (3) When tt ′ is equal to or longer than the predetermined time TMAX, it becomes 0. Equation (6) indicates that a weighted average of the composition change end determination value ω ₀ fin and the calculated reference value ω ₀ is calculated.

In this manner, immediately after the completion of the composition change state, a value close to the composition change end determination value ω ₀ fin is output as the reference value ω ₀ ′, and as represented by the equation (1) over time. Weight is given to the reference value ω ₀ calculated by the following equation, and finally, a value equal to the reference value ω ₀ calculated by the equation (1) is output as the reference value ω ₀ ′. become. As described above, the composition change end determination value ω ₀
fin is a value substantially close to the zero output value of the angular velocity sensor 10. Therefore, by using the composition change end determination value ω ₀ fin for the reference value calculation immediately after the completion of the composition change state, highly accurate blur correction can be performed even immediately after the completion of the composition change state. Become.

(Example of First Embodiment) FIG. 7 is a diagram showing an output example of the shake detection signal ω and the reference value ω ₀ of the shake detection device according to the present embodiment. Hereinafter, a specific example of detection of the start of the composition change state to detection of the end of the composition change state will be described with reference to FIG. The shake data shown in FIG.
When the half-press switch SW1 is turned on (time t1
1) is an example of data when the camera is in the stationary state, pans from the stationary state to the composition change state, and ends panning again to the stationary state. Actual panning is performed from around t0 to around t5.

(Detection of Start of Composition Change State) Time t11
, The half-press switch SW1 is turned ON, the reference value calculation in the stationary state is performed by the equation (1), and the reference value ω ₀ is output. Until the time t1, the camera is almost stationary, so that the output of the reference value ω ₀ is a value substantially equal to the zero output value of the angular velocity sensor 10. From the time t11 to the vicinity of the time t0, the camera is in the stationary state. Therefore, the start of the composition change state is monitored by the flow of FIG. 5 while performing the motion compensation operation at the stationary state.

In [0135] the time t7 to t8, not changed the sign of the formula (4) obtained the angular velocity detection value omega '(i.e., in FIG. 7, the shake detection signal omega does not intersect the reference value omega ₀₎ However, before the state continues for the time of the threshold value T_START, the sign of the time t8 has changed, so that the start of the composition change state is not detected.

In the period from time t9 to t10, the detected angular velocity ω ′
Has continued for the time of the threshold value T_START, but at time t10, the angular velocity detection value ω ′
Absolute value does not exceed the threshold Omega'START (i.e., in FIG. 7, the shake detection signal omega is the reference value omega ₀ in the range of the threshold Omega'START) for, the composition change state Start is not detected.

At times t1 to t2, the sign of the detected angular velocity ω ′ has not changed. Further, in that state, the threshold value T_
At the time t2 after the elapse of START, the detected angular velocity ω ′ exceeds ω ′ START. Therefore, the start of the composition change is detected at this time t2, and the arithmetic expression of the reference value ω ₀ is changed from Expression (1) to Expression (2).

Further, at this time t2, the composition change
End judgment value ω₀Work to save fin. But when
In the interval t2, the time has elapsed since the panning was actually started.
Effect of the shake detection value ω due to panning
The reference value ω₀Is the zero output value of the angular velocity sensor 10
It is out of alignment. Therefore, the calculation was performed at this time t2.
Reference value ω₀(T2) is the composition change end determination value ω₀fin and
Storage is not preferred. Therefore, the reference value
Reference value ω stored in storage unit 53₀The oldest of
Is the composition change end judgment value ω₀Save as fin. FIG.
Then, the oldest reference value at time t2 is performed at time t0.
Calculated reference value ω₀(T0) is the composition change end determination value ω ₀
Save as fin.

(Detection of End of Composition Change State) Next, a case where the composition change state returns to the stationary state will be described.
At time t3, since the absolute value of the composition change angular velocity value ω ″ is smaller than the threshold value ω ″ fin, measurement of the time during which this state continues is started. However, since the measurement time has increased again at time t4 before reaching the threshold value T_FINISH, the end of the composition change state is not detected here.

At time t5, since the absolute value of the composition change angular velocity value ω ″ is smaller than the threshold value ω ″ fin, the time during which this state continues from this time t5 is measured. This measurement time starts from the time t5, and the threshold value T_FIN
The end of the composition change state is detected when the time t6 has elapsed by ISH, and the detection result of the composition change detection unit 40 is set to the stationary state.

In this embodiment, the detection of the completion of the composition change state is intentionally delayed. This is because the end of the composition change state is not erroneously detected due to panning or the like that changes direction in the middle. In addition, the reference value used for detecting the end of the composition change state is higher than that at the time when the start of the composition change state is detected. it can.

According to the present embodiment, the oldest reference value at the start of the composition change state is determined as the composition change end determination value ω ₀ fin.
As the composition change end determination value ω ₀ f
In becomes closer to the output value of the angular velocity sensor 10 when the angular velocity sensor 10 is completely stationary, and the detection accuracy of the end of the composition change state can be improved. Further, by monitoring both the sign and the magnitude of the angular velocity detection value ω ′ and detecting the start of the composition change, the probability of erroneously determining that the composition is changed when the vehicle is at rest can be reduced, and operability and user-friendliness can be reduced. It is possible to provide a good device. Furthermore, by intentionally delaying the completion detection of the composition change state, the image can be determined according to the photographer's intention without the image fluctuating immediately after the composition change. A comfortable camera can be provided.

(Second Embodiment) In the first embodiment, it has been assumed that the composition change end determination value ω ₀ fin is substantially equal to the zero output value of the angular velocity sensor 10. However, when the calculation time (the number of data) of the reference value stored as the composition change end determination value ω ₀ fin is not sufficient,
The composition change end determination value ω ₀ fin may have a deviation from the zero output value of the angular velocity sensor 10.
In the second embodiment, even if the reference value is slightly different from the zero output value, highly accurate blur correction can be performed immediately after the completion of the composition change state.

FIG. 8 is a diagram showing the internal configuration of the drive signal calculation section 50 and the composition change detection section 40 in this embodiment. A new reference value calculation control unit 48 is added to the composition change detection unit 40.
Are the same as in the first embodiment, except for the provision of the, and the description of the overlapping parts will be omitted.

The reference value calculation control unit 48 is a reference value calculation control unit that controls the method of calculating the reference value ω ₀ after detecting the end of the composition change state. The reference value calculation control unit 48 receives the signal when the composition change start detection unit 44 detects the start of the composition change state, and at the same time, how long the stationary state has continued, that is, the composition change state. before the start of the is detected, and stores the operation time of the reference value omega _0.
When the composition change end detection unit 46 detects the end of the composition change, the reference value calculation control unit 48 stores the stored reference value ω _0.
A signal for controlling the method of calculating the reference value is transmitted to the reference value calculating unit 52 in accordance with the calculation time. The reference value calculation control unit 48 selects which of the following two calculations is to be performed based on this signal. (1) When the duration of the stationary state is long: A weighted average is calculated as in the first embodiment. (2) When the duration of the stationary state is short: The weighted average is not calculated.

For the calculation of the reference value ω ₀ , starting with the moving average,
A general method is to extract a low frequency component from the shake detection signal using a low-pass filter or the like. However, these low-pass filters require a certain amount of calculation time until a desired output (zero output of the angular velocity sensor) is obtained. The reference value ω ₀ is sufficiently close to a desired value (zero output of the angular velocity sensor) if the calculation time is sufficient, but is not limited if the calculation time is short. Therefore, in the second embodiment, it is monitored how long the stationary state has continued before the start of the composition change state is detected, that is, how long the reference value calculation in the stationary state has continued (FIG. 7). Then, the time corresponding to t11 to t2) is determined, and whether or not to execute the weighted averaging as in the first embodiment is controlled in accordance with the length of the time, so that the end of the composition change state is detected. The method of calculating the reference value ω ₀ is changed.

When the duration of the stationary state is long, weighted averaging [Equation (6)] as in the first embodiment is executed.
This is because the calculation of the reference value ω ₀ was performed for a sufficiently long time before the detection of the start of the composition change state, so that the composition change end determination value ω ₀ f
This is because in is considered to be sufficiently close to the zero output value of the angular velocity sensor 10. As described above, by calculating the weighted average as in the first embodiment, highly accurate blur correction can be performed even immediately after the completion of the composition change.

On the other hand, if the duration of the stationary state is short, the weighted averaging as in the first embodiment is not performed. This is because if the calculation of the reference value is not performed for a sufficient period of time, the composition change end determination value ω ₀ fin may greatly deviate from the zero output value of the angular velocity sensor 10. In this case, the weighted average as in the first embodiment is not calculated, but the number of data of the moving average is changed stepwise according to the elapsed time from the detection of the completion of the composition change, for example, as follows.

[0149]

(Equation 4)

Here, t 'is the time when the end of the composition change state is detected. This is because immediately after the completion of the composition change, the response of the reference value to the shake detection signal is increased by reducing the number of data of the moving average, and the reference value ω ₀ becomes a value close to the zero output value of the angular velocity sensor 10. It is to speed up. Then, as the reference value approaches the zero output value with the passage of time, the number of moving average data is increased to increase the accuracy.

When the composition change end determination value ω ₀ fin is largely deviated from the zero output value of the angular velocity sensor 10, when the weighted average of the equation (6) is executed, the reference value ω ₀ becomes a value close to the zero output value. It may take some time to complete. That is, it may take some time from the completion of the composition change to the improvement of the blur correction accuracy. However, if the number of data of the moving average is increased stepwise as in the present embodiment, the time until the accuracy is improved can be shortened as compared with the weighted average as in Expression (6). . Note that there is a method in which the cutoff frequency of the low-pass filter is gradually reduced as time elapses after the completion of the composition change.

FIG. 9 is a flowchart of the operation performed when the reference value is calculated in the stationary state after the end of the composition change state is detected. The operation corresponding to the operation at the position of S210 in FIG. FIG. Hereinafter, FIG.
The control operation of the reference value calculation in the stationary state after detecting the end of the composition change state will be described below.

In S700, it is determined whether or not the calculation time of the reference value ω ₀ before the detection of the start of the composition change state is larger than the threshold value T_CAL1. If the calculation time of the reference value ω ₀ is longer than the threshold value T_CAL1, the composition change end determination value ω ₀ fin is a value substantially equal to the zero output value of the angular velocity sensor 10, so the process proceeds to S710 and the equation When the weighted average according to (6) is executed and the calculation time of the reference value ω ₀ is not larger than the threshold value T_CAL1,
Since the composition change end determination value ω ₀ fin may deviate from the zero output value of the angular velocity sensor 10, the process proceeds to S720, and the weighted averaging based on Expression (6) is not performed.

In S710, weight averaging according to equation (6) is performed.

In S720, it is determined whether or not the elapsed time (tt ') from the detection of the end of the composition change state is smaller than a predetermined time 1. If the elapsed time (tt ′) is shorter than the predetermined time 1, it is immediately after the completion of the composition change state, and the process proceeds to S730, and the (K5) reference value ω ₀ having a small number of data according to the equation (7). Is calculated. If the elapsed time (tt ′) is not smaller than the predetermined time 1, the process proceeds to S740, and the elapsed time is further determined.

In S730, the reference value ω is calculated by the equation (7).
Performs an operation of ₀ . This makes it possible to quickly respond reference value omega ₀ for shake detection signal omega. Therefore, even in the shooting immediately after the end of the panning, the blur correction can be performed without significantly lowering the accuracy.

In S740, it is determined whether or not the elapsed time (t-t ') from the detection of the end of the composition change state is smaller than the predetermined time 2. Here, there is a relationship of predetermined time 1 <predetermined time 2. The elapsed time (tt ′) is equal to the predetermined time 2
If the time is smaller than the predetermined time 1, the predetermined time 1 or more has elapsed from the end of the composition change state, and the predetermined time 2 has not been reached. Therefore, the predetermined time 1 ≦ the elapsed time (tt ′) <the predetermined time 2 is satisfied. Therefore, the process proceeds to S750, and the calculation of the (K4) reference value ω ₀ with a slightly larger number of data is performed by equation (8). If the elapsed time (tt ′) is not shorter than the predetermined time 2, the process proceeds to S760.

In step S750, a reference value ω ₀ having a slightly larger number of data (K4) is calculated according to equation (8). In this case, it is possible to perform the shake correction with higher accuracy than the equation (7).

In S760, the (K0) reference value ω ₀ having a large number of data is calculated by the equation (9). Expression (9) is the same expression as the expression for calculating the reference value ω ₀ in the stationary state shown in Expression (1), and can perform the most accurate blur correction as in the normal stationary state.

According to the present embodiment, when the continuation time of the stationary state is long, the weighted average is calculated as in the first embodiment, whereby highly accurate blur correction can be performed even immediately after the completion of the composition change. . On the other hand, when the duration of the stationary state is short, the weighted averaging as in the first embodiment is not performed. Immediately after the composition change is completed, the number of data of the moving average is reduced and the reference value for the shake detection signal is reduced. Since the response is made faster, the reference value ω ₀ can be more quickly brought to a value close to the zero output value of the angular velocity sensor 10. Therefore, even if the reference value reference value ω ₀ is slightly deviated from the zero output value, high-precision blur correction can be performed immediately after the completion of the composition change state, and the accuracy of the blur correction can be further increased in a short time thereafter. Can be higher.

(Third Embodiment) In the first and second embodiments, the stationary state exists before the composition change state starts, and the oldest value of the reference value ω ₀ calculated for the stationary state is , The end of the composition change state is detected by the oldest reference value ω.
₀ (t−M) was stored as a composition change end determination value ω ₀ fin and used as a reference for detection.

However, when the stationary state before the composition change state is started is very short, the calculated oldest reference value ω ₀ may not be close to the zero output value of the angular velocity sensor 10. is there. Also, when the half-press switch SW1 is pressed while the camera is being shaken, the camera operation is immediately performed from the composition change state without the stationary state, and the calculated oldest reference value ω ₀ is , The value when a large swing in the composition change state is received. Therefore, in such a case, even if the oldest reference value ω ₀ (t−M) is stored as the composition change end determination value ω ₀ fin and used as a detection reference, the shake detection signal ω or Even when the panning is completed without approaching the reference value ω _0, there is a case where the composition change state is still determined. Further, in some cases, it is determined that the composition change state has ended even though panning is being performed, and the end of the composition change state has not been correctly detected. Therefore, in the present embodiment, an embodiment is shown in which even in such a case, the end of the composition change can be correctly detected.

The present embodiment is a modification of the operation of FIG. 6 corresponding to the portion of S210 in FIG. 3 of the first or second embodiment, and the duplicate description will be omitted as appropriate, and the operation of FIG. The following description focuses on the operation shown in FIG. FIG. 10 is a flowchart illustrating the flow of the operation of the composition change end detection unit.

In S800, a composition change angular velocity value ω ″ for detecting the end of the composition change state is calculated. The composition change angular velocity value ω ″ is a reference value ω ₀ and a composition change end determination value ω _0.
This is the difference from fin, and is expressed as follows in the equation. ω ″ (t) = ω ₀ (t) −ω ₀ fin Equation (5) Here, since the composition is being changed, the reference value ω ₀ in Equation (5) is calculated by Equation (5) in the present embodiment. This is a value calculated by 2).

In S810, it is determined whether or not the absolute value of the composition change angular velocity value ω ″ calculated in S800 is smaller than the threshold value ω ″ fin. Affirmative determination (| ω "| <ω" f
in), the reference value ω ₀ is the composition change end determination value ω ₀
Since it can be determined that the value is close to fin, S82
The process goes to 0, a negative decision if the (| | ω "≧ ω" fin), the reference value ω ₀ is a distant value from the composition change end determination value ω ₀ fin ω "fin or more, the process proceeds to S860.

In S820, the counter Tfin is incremented by one (Tfin = Tfin + 1). This counter Tfi
n is the absolute value of the composition change angular velocity value ω ″ is the threshold value ω ″ f
This is a counter for measuring the time during which the state smaller than “in” continues.
0 when a negative determination is made at 0 (| ω ″ | ≧ ω ″ fin)
Is reset to

In S830, it is determined whether or not the counter Tfin is equal to or larger than a threshold value T_FINISH. If the determination is affirmative (Tfin ≧ T_FINISH), the process proceeds to S840, and the detection result of the composition change detection unit 40 is set to the stationary state,
If negative determination (Tfin <T_FINISH), S92
The process proceeds to 0, and the detection result of the composition change detection unit 40 is continued in the composition change state.

Composition change end judgment value ω stored₀fi
n is a value substantially equal to the zero output value of the angular velocity sensor 10
It is believed that there is. In the composition change state, the camera
This composition change end determination value ω₀fin
Near reference value ω₀Is not very common,
When the change of the drawing is completed and the vehicle is stopped again, the reference value ω ₀
Is the composition change end determination value ω₀close to fin
And that state continues. Therefore, the reference value ω₀But
Figure change end judgment value ω₀Whether the value is close to fin
Is determined (S810), and the state of the state where the values are close to each other is determined.
The duration is measured (S820, 830), and the result is
Time (T_FINISH) has been detected
(S830), thereby detecting the end of the composition change state.
You.

In S840, the counters Tfin and Tp
Reset an to zero. In S850, the detection result of the composition change detection unit 40 is set to the stationary state. In S860, the counter Tfin is reset to 0.

In S870, it is determined whether or not the calculation time of the reference value ω ₀ before the detection of the start of the composition change state is larger than the threshold value T_CAL2. When the calculation time of the reference value ω ₀ is larger than the threshold value T_CAL2, the composition change end determination value ω ₀ fin is almost equal to the zero output value of the angular velocity sensor 10, so the composition change end determination value ω
There is no need to correct ₀ fin, and the process proceeds to S920. on the other hand,
When the calculation time of the reference value ω ₀ is not longer than the threshold value T_CAL1, the composition change end determination value ω ₀ fin
Since there is a possibility that the zero output value of the angular velocity sensor 10 has shifted, the process proceeds to S880, and S880 to S880.
The flow which may update the composition change end determination value ω ₀ fin shown in S910 is executed.

When this embodiment is implemented in combination with the second embodiment, it is necessary to satisfy T_CAL1 ≧ T_CAL2. If this condition is not satisfied, a case may occur in which, even when the composition change end determination value ω ₀ fin is updated, the reference value ω ₀ is calculated using Expression (6) after returning to the stationary state. This is because there are times. T_
If CAL1 ≧ T_CAL2, when the composition change end determination value ω ₀ fin is updated, after returning to the stationary state,
The calculation of the reference value ω ₀ is always performed using the equations (7) to (9). When the composition change end determination value ω ₀ fin is updated, there is a high possibility that the oldest reference value ω ₀ is not an accurate value. Therefore, in this case, the equations (7) to (9) are always used.
It is correct to calculate the reference value ω ₀ after returning to the stationary state by using.

In S880, the counter Tpan is advanced by one. The counter Tpan has a composition change angular velocity value ω ″
Is a counter for measuring the time during which the state in which the absolute value of is larger than the threshold value ω ″ fin continues, and is reset to 0 when the half-press timer 100 is ON or when the process proceeds to S840 or S910.

In S890, it is determined whether or not the counter Tpan is equal to or larger than a threshold value T_PAN. If the determination is affirmative (Tpan ≧ T_PAN), the process proceeds to S900, and if the determination is negative (Tpan <T_PAN), the process proceeds to S920.

Here, the affirmative determination means that the reference value ω ₀ does not remain near the composition change end determination value ω ₀ fin. In other words, this affirmative determination indicates that even if the composition change end determination value ω ₀ fin is used as it is, it is unlikely that the condition for ending the composition change state described in S830 will occur. .

When the composition change state ends and the apparatus enters the stationary state, the reference value ω ₀ becomes a value close to the zero output value of the angular velocity sensor 10. Therefore, the affirmative determination here also indicates that there is a high possibility that the composition change end determination value ω ₀ fin is deviated from the zero output value of the angular velocity sensor 10. Therefore, if an affirmative determination is made here, the process proceeds to S900 to perform an operation of updating the composition change end determination value ω ₀ fin.

In S900, the composition change end determination value ω ₀ f
Update in. Specifically, the reference value ω ₀ (t) at the time of proceeding to this step is changed to a new composition change end determination value ω
₀ fin. As mentioned in the description of S830, composition change end detection unit 46, by detecting that the stay long in the vicinity of a certain value the reference value omega _0, detects the end of the composition change state. However, as I mentioned in S890, if the process proceeds to the S890, likely to remain the reference value omega ₀ a long time in the vicinity of the composition change end determination value omega ₀ fin at that time, little. Therefore, here, the composition change end determination value ω ₀ fin is updated, and an operation of determining whether or not the composition change state end condition is satisfied again is performed using the updated composition change end determination value ω ₀ fin.

In S910, the counter Tpan is reset to 0. In S920, the detection result of the composition change detection unit 40 is continued in the composition change state.

In this embodiment, the end of the composition change state is detected by obtaining the composition change angular velocity value ω ″, which is the difference between the composition change end determination value ω ₀ fin and the reference value ω ₀ . Can be obtained as the difference between the composition change end determination value ω ₀ fin and the shake detection signal ω, similarly to the first and second embodiments, and the reference value ω ₀ is changed to the composition change end The fact that it is desirable to use it for detection is the same as in the first and second embodiments.

(Example of Third Embodiment) FIG. 11 is a diagram showing an output example of the shake detection signal ω and the reference value ω ₀ of the shake detection device according to the present embodiment. Hereinafter, a specific example of detecting the end of the composition change state according to the present embodiment will be described with reference to FIG. In the example shown in FIG. 11, a certain point in time during which panning is actually performed (t = t0)
In this case, the detection of the shake is started, and the panning is completed in the vicinity of t4 to t5, and thereafter, the still state is continued. Note that, as described in the first embodiment, immediately after the half-press switch SW1 is turned on, in the original algorithm, the stationary state is set (see S60 in FIG. 3).
However, since subsequent operations are not affected, here, for the sake of simplicity, description will be made assuming that the state has already been determined as the composition change state.

In the range of t0 ≦ t <t1, the absolute value of the reference value ω ₀ is smaller than the composition change end determination value ω ₀ fin. However, when this state is equal to the threshold
Since the absolute value of the reference value ω ₀ is larger than the composition change end determination value ω ₀ fin at time t1 before continuing for H, the end of the composition change state is not detected. In the flow of FIG. 10, S800, S810, S820, S83
0 and S920.

In the range of t1 ≦ t <t2, the absolute value of the reference value ω ₀ is larger than the composition change end determination value ω ₀ fin. In addition, since the detection of the shake is started during the panning, the calculation time of the reference value before the start of the composition change is T_
In this range, S800, S810, S860, S870, S870
The processing is performed in the order of 880, S890, and S920.

At the time when t = t2, the counter Tpa
Since n becomes equal to the threshold value T_PAN, the composition change end determination value ω ₀ fin is updated here. As shown in FIG. 11, the composition change end determination value ω ₀ fin after the update is t =
This is the reference value ω ₀ at t2. In the flow of FIG.
00, S810, S860, S870, S880, S8
90, S900, S910, and S920 are executed in this order.

In the range of t2 ≦ t <t3, t0 ≦ t <t
The same operation as in the range of 1 is performed. In the range of t3 ≦ t <t4, the same operation as in the range of t1 ≦ t <t2 is performed. t = t
At time point 4, the same operation as at time point t = t2 is performed. t
In the range of 4 ≦ t <t5, the same operation as in the range of t0 ≦ t <t1 is performed.

At time t = t5, the state in which the absolute value of the reference value ω ₀ is smaller than the composition change end determination value ω ₀ fin has continued for T_FINISH, so that the end of the composition change state is detected at this time t5. The detection result of the composition change detection unit 40 is set to the stationary state. In the flow of FIG.
0, S810, S820, S830, S840, S85
It is executed in the order of 0.

According to the present embodiment, while the composition is being changed, the composition
Value for detecting the end of change (composition change end judgment value)
Is detected, shake detection starts during composition change.
Even then, even in a situation where it becomes stationary,
Detection and shake correction suitable for the state of
It works. What is the zero output value of the angular velocity sensor?
The correct reference value ω ₀Can be calculated by
The value of the shake detection signal in the stationary state is
Can also deal with. Therefore, the direct detection of the shake detection signal
Since there is no need to perform operations such as cutting the flow components,
A circuit for processing the detection signal can be simplified.

(Modifications) Various modifications and changes are possible without being limited to the embodiments described above, and they are also within the equivalent scope of the present invention. For example, in each of the embodiments, the arithmetic expressions for calculating various values such as the reference value ω ₀ have been described. However, the present invention is not limited thereto, and the contents of the arithmetic expressions may be changed as appropriate.

In each of the embodiments, the example in which the present invention is applied to a shake correction camera using a silver halide film has been described. However, the present invention is not limited to this. For example, an image is recorded electrically using a CCD or the like. A so-called digital camera can be used,
A video camera or other optical devices may be used.

[0188]

As described in detail above, according to the first aspect of the present invention, the movement start detecting unit detects the vibration detection signal and / or
Or, when the reference value outputs a low-frequency signal for a predetermined time or more, it is detected that the movement has started, so that blurring due to camera shake in a state where the camera is held normally without changing the composition is corrected with high accuracy. In addition, the start of the movement can be reliably detected.

According to the second aspect of the present invention, the movement start detecting section calculates a difference value between the vibration detection signal and the reference value, and indicates that the difference value is moving in the same direction as the movement start reference time. In addition, since the start of movement is detected when the difference value at the elapse of the movement start reference time is equal to or greater than the movement start reference value, both the direction and magnitude of the shake can be monitored, and the start of the composition change can be improved. The detection can be performed reliably. Therefore, the composition change can be started without giving the user a feeling of strangeness.

According to the third aspect of the present invention, the movement end detecting section detects the movement end from the movement end determination value serving as a reference for detecting the movement end and the vibration detection signal and / or the reference value. Can be reliably detected.

According to the fourth aspect of the present invention, a movement end determination value serving as a reference for detecting the end of movement, a vibration detection signal and / or
Alternatively, since the movement end detecting unit that detects the end of the movement from the reference value is provided, the end of the movement can be reliably detected.

According to the fifth aspect of the present invention, the movement end detecting section detects the movement when the vibration detection signal and / or the reference value is within a predetermined range from the movement end determination value for a movement end reference time. Since the end is detected, both the direction and magnitude of the shake can be monitored, and the end of the composition change can be detected more reliably. Therefore, the composition change can be started without giving the user a feeling of strangeness.

According to the invention of claim 6, there is provided a reference value storage unit for sequentially storing a predetermined number of reference values, and among the reference values stored in the reference value storage unit, the movement start detection unit Saves a provisional reference value based on a reference value in the past from when the movement was detected, the provisional reference value can be used for subsequent calculations, and the accuracy of the blur correction operation can be improved.

According to the seventh aspect of the present invention, there is provided a reference value storage section for sequentially storing a predetermined number of reference values, and the movement start detecting section is selected from the reference values stored in the reference value storage section. Since the oldest reference value is stored as the provisional reference value when the movement is detected, the reliability of the provisional reference value to be stored can be increased.

According to the eighth aspect of the present invention, since the movement end detecting section sets the provisional reference value as the movement end determination value, it is possible to more reliably detect the end of the movement, and to give the user a sense of discomfort. The composition change operation can be performed without any change.

According to the ninth aspect of the present invention, the movement end detecting section calculates the absolute value of the difference between the provisional reference value and the vibration detection signal or the reference value, and determines whether the absolute value of the difference is equal to or less than the movement end threshold value. Detects the end of movement when the movement has continued for the reference time, so that both the direction and magnitude of the shake can be monitored,
The end of the composition change can be more reliably detected, and the accuracy of the blur correction operation can be increased. Therefore, the composition change can be started without giving the user a feeling of strangeness.

According to the tenth aspect, the reference value calculating section calculates the moving average of the vibration detection signal as a reference value, and when the movement is detected, the reference value calculating section uses the moving average rather than the calculation in a substantially stationary state. Since the number of data of the detected vibration signals is reduced, it is possible to calculate a reference value suitable for each of the substantially stationary state and the moving state, and to determine the composition as intended by the photographer. Further, during movement, by suppressing the blur correction operation more than in the stationary state, the shooting result can be improved even in panning.

According to the eleventh aspect of the present invention, the reference value calculating section has a low-pass filter, and when a movement is detected, the cut-off frequency used for the low-pass filter is higher than that in the calculation in a substantially stationary state. Since the setting is made, a reference value suitable for each of the substantially stationary state and the moving state can be calculated, and the composition can be determined as intended by the photographer.

According to the twelfth aspect of the present invention, the movement end detecting section updates the movement end determination value when the movement end is not detected even after the lapse of a predetermined time or more. Even if it has not been obtained, the movement end determination value is more quickly converged to the zero output value of the angular velocity sensor, so that a blur-corrected photograph can be taken even after completion of the composition change. Further, even in a case where the detection of the shake is started during the movement and the state becomes substantially stationary thereafter, the stationary state can be quickly detected.

[0200] According to the thirteenth aspect, the movement end detecting section calculates the absolute value of the difference between the movement end determination value and the vibration detection signal or the reference value, and the absolute value of the difference is larger than the movement end threshold value. Since the movement end judgment value is updated when the state continues for the movement judgment time, both the direction and magnitude of the shake can be monitored from the vibration detection signal or the reference value, and the update of the movement end judgment value can be reliably updated when necessary. It can be carried out.

According to the fourteenth aspect of the present invention, the movement end detecting section updates the movement end determining value based on the calculation duration of the reference value calculated before the movement start detecting section detects the start of movement. When the reliability of the reference value is high, the end of the movement can be accurately detected by using the reference value as the movement end determination value, and when the reliability of the reference value is low, the movement end determination value Can be surely updated.

According to the fifteenth aspect of the present invention, the reference value calculation control unit does not update the movement end determination value when the calculation duration is longer than the update reference time, and the calculation duration is longer than the update reference time. If the distance is also short, the movement end determination value is updated, so that the criterion of whether to perform the update is clear, and the control of the reference value calculation control unit can be simplified. Adjustment can be easily performed by changing the update reference time.

According to the sixteenth aspect of the invention, the movement end detecting section updates the vibration detection signal or reference value at the time of updating as a new movement end judgment value, so that the movement end judgment value can be easily set to an appropriate value. I can guide you.

According to the seventeenth aspect, since the reference value calculation control unit for controlling the reference value calculation method of the reference value calculation unit after the movement end detection unit detects the end of the movement, the movement is completed. After that, the calculation of the reference value can be performed by an appropriate method.

According to the eighteenth aspect of the invention, the reference value calculation control section controls whether or not to use the provisional reference value for the reference value calculation. Can be used for

According to the nineteenth aspect of the present invention, the reference value calculation control unit determines whether the movement end detection unit has moved based on the reference value calculation duration calculated before the movement start detection unit detects the start of movement. Since the calculation method of the reference value after detecting the end of the reference value is controlled, the reliability of the provisional reference value can be easily and accurately determined.

According to the twentieth aspect, the reference value calculation control section uses the provisional reference value for the reference value calculation when the calculation continuation time is longer than the calculation control reference time. When the reference time is shorter than the reference time, the provisional reference value is controlled so as not to be used for the reference value calculation, so that the control of the reference value calculation control unit can be simplified, and the adjustment can be performed by changing the calculation control reference time. Can also be easily performed.

According to the twenty-first aspect, immediately after the movement end detecting section detects the end of the movement, the provisional reference value is set as the reference value calculated by the reference value calculating section, and gradually as the time elapses thereafter. Since the reference value reflecting the value of the vibration detection signal at that time is calculated, the reference value immediately after the end of the movement can be calculated with high accuracy, and the calculation accuracy in the subsequent calculation can be improved. .

According to the twenty-second aspect, immediately after the movement end detecting section detects the end of the movement, the reference value calculating section reduces the number of samples of the vibration detection signal used for the calculation, and sets the time for the subsequent time. As the reference value calculation unit gradually increases the number of samples of the vibration detection signal used in the calculation as time passes, the response of the reference value immediately after detecting the end of the movement becomes faster, and the accuracy of the calculation result of the reference value increases. Improvement can be accelerated, and highly accurate blur correction can be performed even when the image is shot immediately after the completion of the composition change.

According to the twenty-third aspect of the present invention, since the movement is intentionally performed by the user, it is possible for the photographer to follow the moving object without feeling uncomfortable or uncomfortable in the case of a panning shot or the like. it can.

According to the twenty-fourth aspect of the present invention, since a shake detection device, a blur correction optical system, a drive unit, and a control unit are provided, composition change and the like can be performed without discomfort, and highly accurate blur correction can be performed. A blur correction optical device that can be performed can be obtained.

According to the twenty-fifth aspect of the present invention, since a photographing optical system including a blur correction optical system and a photographing section for photographing an image formed by the photographing optical system are provided, composition change and the like can be performed without discomfort. Shooting can be performed while performing highly accurate blur correction.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an outline of a shake detection apparatus and a shake correction optical apparatus according to a first embodiment of the present invention.

FIG. 2 is a diagram showing an internal configuration of a drive signal calculation unit 50 and a composition change detection unit 40.

FIG. 3 is a flowchart illustrating a flow of the entire camera system including the shake detection device according to the embodiment.

FIG. 4 is a flowchart showing a flow of an operation of a reference value storage unit 53;

5 is a flowchart showing a flow of an operation of a composition change start detection unit 44, and is a diagram showing the operation in S220 of FIG. 3 in detail.

6 is a flowchart showing a flow of an operation of a composition change end detection unit 46, and is a diagram showing the operation in S240 of FIG. 3 in detail.

7 is a diagram showing an example of the output of the shake detection signal omega and the reference value omega ₀ of the shake detecting device in the present embodiment.

FIG. 8 is a diagram showing an internal configuration of a drive signal calculation unit 50 and a composition change detection unit 40 according to the present embodiment.

9 is a flowchart of an operation performed when a reference value calculation is performed in a stationary state after detecting the end of the composition change state, and illustrates an operation corresponding to the operation at the position of S210 in FIG. 3; It is.

FIG. 10 is a flowchart illustrating a flow of an operation of a composition change end detection unit.

11 is a diagram showing an example of the output of the shake detection signal omega and the reference value omega ₀ of the shake detecting device in the present embodiment.

FIG. 12 is a block diagram illustrating a basic configuration of a conventional shake correction optical apparatus including a shake detection device.

[Explanation of symbols]

 10a, 10b Angular velocity sensor 44 Composition change start detection unit 46 Composition change end detection unit 48 Reference value calculation control unit 50 Drive signal calculation unit 52 Reference value calculation unit 53 Reference value storage unit 70a, 70b Drive unit 80 Shake correction lens 160 Lens mirror Tube 170 Camera body

Claims (25)

[Claims] A vibration detection unit that detects vibration and outputs a vibration detection signal; a reference value calculation unit that calculates a reference value of the vibration detection signal; and a vibration detection unit that detects the vibration from the vibration detection signal and the reference value. A movement start detection unit that detects that a large movement has been started as compared to a substantially stationary state of the device including the unit. In the shake detection device, the movement start detection unit includes the vibration detection signal and / or the vibration detection signal. Detecting a start of the movement when the reference value outputs a low-frequency signal for a predetermined time or more. 2. A vibration detection unit that detects vibration and outputs a vibration detection signal, a reference value calculation unit that calculates a reference value of the vibration detection signal, and detects the vibration from the vibration detection signal and the reference value. A movement start detection unit that detects that a large movement has been started as compared to a substantially stationary state of the device including the unit.In the shake detection device, the movement start detection unit includes the vibration detection signal and the reference value. And the difference value indicates that the difference value is moving in the same direction as the movement start reference time, and the difference value at the elapse of the movement start reference time is equal to or greater than the movement start reference value. Detecting the start of the movement when the movement occurs. 3. A vibration detection unit that detects vibration and outputs a vibration detection signal, a reference value calculation unit that calculates a reference value of the vibration detection signal, and compares the vibration detection unit with a substantially stationary state including the vibration detection unit. A movement end detection unit that detects that a large movement has ended. The movement detection unit includes: a movement end determination value serving as a reference for detecting the end of the movement; Detecting the end of the movement from a detection signal and / or the reference value. 4. The shake detection apparatus according to claim 1, wherein the movement end is determined from a movement end determination value serving as a reference for detecting the end of the movement and the vibration detection signal and / or the reference value. The shake detection device, comprising: the movement end detection unit for detecting the movement end. 5. The shake detection device according to claim 4, wherein the movement end detection unit determines that the state in which the vibration detection signal and / or the reference value is within a predetermined range from the movement end determination value is the movement end. Detecting the end of the movement when the reference time has continued. 6. The shake detecting apparatus according to claim 4, further comprising a reference value storage unit for sequentially storing a predetermined number of the reference values, wherein the reference value storage unit stores the reference number in the reference value storage unit. And storing a provisional reference value based on a reference value that is earlier than a time point at which the movement start detection unit detects the movement among the reference values. 7. The shake detecting apparatus according to claim 4, further comprising a reference value storage unit for sequentially storing a predetermined number of the reference values, wherein the reference value is stored in the reference value storage unit. Wherein the oldest reference value is stored as a provisional reference value when the movement start detection unit detects the movement from among the values. 8. The shake detection device according to claim 6, wherein the movement end detection unit sets the provisional reference value as the movement end determination value. 9. Any one of claims 6 to 8
In the shake detection device according to the paragraph, the movement end detection unit calculates the absolute value of the difference between the provisional reference value and the vibration detection signal or the reference value, and the absolute value of the difference is equal to or less than the movement end threshold. Detecting the end of the movement when the state of (1) continues for the movement end reference time. 10. The shake detection device according to claim 4, wherein the reference value calculation unit calculates a moving average of the vibration detection signal as a reference value, and A vibration detection device that, when detected, reduces the number of data of the vibration detection signal used in the calculation as compared with the calculation in the substantially stationary state. 11. The shake detection device according to claim 4, wherein the reference value calculation unit has a low-pass filter, and when the movement is detected, the low-pass filter includes a low-pass filter. A cut-off frequency used for the filter is set higher than the calculation in the substantially stationary state. 12. The shake detection device according to claim 3, wherein the movement end detection unit detects that the movement end is not detected even after a predetermined time has elapsed. Updating the movement end determination value. 13. The shake detection device according to claim 3, wherein the movement end detection unit calculates a difference between the movement end determination value and the vibration detection signal or the reference value. And calculating the absolute value of the difference, and updating the movement end determination value when the state in which the absolute value of the difference is larger than the movement end threshold continues for the movement determination time. 14. The shake detection device according to claim 12, wherein the movement end detection unit calculates a reference value calculated before the movement start detection unit detects the start of the movement. Determining whether or not to update the movement end determination value based on a continuation time. 15. The shake detection device according to claim 14, wherein the reference value calculation control unit does not update the movement end determination value when the calculation continuation time is longer than an update reference time. The movement ending determination value is updated when the operation continuation time is shorter than the update reference time. 16. The shake detection device according to claim 12, wherein the movement end detecting unit sets the vibration detection signal or the reference value at the time of updating to a new end of the movement. The shake detection device is updated as a determination value. 17. The shake detection device according to claim 3, wherein the reference value of the reference value calculation unit after the movement end detection unit detects the end of the movement. A shake detection device, comprising: a reference value calculation control unit that controls a calculation method. 18. The shake detection apparatus according to claim 17, wherein the reference value calculation control unit controls whether the provisional reference value is used for reference value calculation. . 19. The shake detection device according to claim 17, wherein the reference value calculation control unit calculates a reference value calculated before the movement start detection unit detects the start of the movement. A method of calculating a reference value after the movement end detection unit detects the end of the movement, based on a calculation duration. 20. The shake detection apparatus according to claim 19, wherein the reference value calculation control unit uses the provisional reference value for reference value calculation when the calculation duration is longer than a calculation control reference time. And controlling the temporary reference value not to be used for reference value calculation when the calculation duration is shorter than the calculation control reference time. 21. The shake detection device according to claim 20, wherein the reference value calculation control unit uses the provisional reference value for reference value calculation, and the movement end detection unit detects the end of the movement. Immediately after the detection, the provisional reference value is set as a reference value calculated by the reference value calculation unit, and a reference value reflecting the value of the vibration detection signal at that time is calculated gradually as time elapses thereafter. A shake detection device. 22. The shake detection device according to claim 20, wherein the reference value calculation control unit does not use the provisional reference value for reference value calculation, and the movement end detection unit detects the end of the movement. Immediately after the detection, reduce the number of samples of the vibration detection signal used by the reference value calculation unit for calculation,
A vibration detection device, wherein the number of samples of the vibration detection signal used in the calculation by the reference value calculation unit is gradually increased as time elapses thereafter. 23. The shake detection device according to claim 1, wherein the movement is a movement intentionally performed by a user. 24. A shake detection apparatus according to claim 1, wherein: a shake correction optical system that corrects a shake of an image; and a drive unit that drives the shake correction optical system. A control unit that calculates a drive signal from the vibration detection signal and the reference value and outputs a calculation result as a drive signal to control the drive of the drive unit. 25. The blur correction optical apparatus according to claim 24, further comprising: a photographing optical system including the blur correction optical system; and a photographing unit that photographs an image formed by the photographing optical system. Optical blur correction optics.