JP3278212B2 - Image stabilizer - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a shake correcting device suitable for use in a photographing device such as a camera.

[0002]

2. Description of the Related Art Conventionally, in the field of a photographing apparatus such as a camera, automation and multi-functionality have been achieved in all aspects such as exposure setting and focus adjustment, so that good photographing can always be performed regardless of the photographing environment. I'm sorry

However, it is often the case that the quality of a photographed image is significantly degraded by camera shake. In recent years, various shake correction devices for correcting the camera shake have been proposed and are attracting attention.

[0004] Shake correction devices are broadly divided into those that correct optically in a correction system, those that correct electrically by image processing, and those that physically detect vibration in a detection system.
The method is roughly classified into a method in which a motion vector of an image is detected by image processing, and many forms are conceivable.

FIG. 19 shows an example of the configuration of a conventional shake correction apparatus.

In FIG. 1, reference numeral 1 denotes an angular velocity detector including an angular velocity sensor such as a vibrating gyroscope, which is attached to a photographing device such as a camera. 2 is an angular velocity detector 1
This is a DC cut filter that cuts off the DC component of the speed signal output from the controller and passes only the AC component, that is, the vibration component. As the DC cut filter, a high-pass filter (hereinafter, referred to as an HPF) that blocks a signal in an arbitrary band may be used. Reference numeral 3 denotes an amplifier for amplifying the angular velocity signal output from the DC cut filter to an appropriate sensitivity.

An A / D converter 4 converts an angular velocity signal output from the amplifier 3 into a digital signal, an integrator 5 integrates an output of the A / D converter 4 and outputs an angular displacement signal, and 6
Is a pan / tilt judging circuit for judging panning and tilting from an integral signal of the angular velocity signal output from the integrating circuit 5, that is, an angular displacement signal, and 7 is an output of the pan / tilt judging circuit to an analog signal or a pulse output such as PWM. This is a D / A converter for converting and outputting. The A / D converter 4, the integrator 5, the pan / tilt determination circuit 6, and the D / A converter 7 are constituted by, for example, a microcomputer COM1. Reference numeral 8 denotes a drive circuit for driving a subsequent image correcting means based on the displacement signal output from the microcomputer so as to suppress the shake, and 9 denotes an image correcting means, for example, by displacing the optical axis to cancel the shake. There is used an optical correction unit that electronically shifts the readout position of an image from a memory that stores an image and cancels a shake.

[0008]

However, the above-described shake correcting apparatus has the following problems.

FIG. 4 shows a frequency characteristic of a vibration component signal obtained by the image correction means 9 when a vibration of a constant amplitude is applied to a device such as a camera equipped with a shake correction device using an existing angular velocity sensor.

In the frequency characteristics shown in FIG. 4, focusing on the characteristics at 10 Hz, the gain characteristics are almost 0 dB, but the phases are shifted by about 7.5 deg. Due to the phase shift as a main factor, the applied vibration is suppressed to 1/100 or less at 3 Hz, but can be suppressed to only about 1/8 at 10 Hz. ing.

In consideration of the effect of the shake correction on such characteristics, even if a sufficient effect can be expected in ordinary photographing, vibration of a frequency of about 10 Hz is continuously applied to a device such as a camera. When it is, the periodic vibration becomes more noticeable.

As described above, in the above-described current shake correction device using the angular velocity sensor, when continuous vibration is continuously applied in a frequency band where a sufficient vibration suppression effect cannot be expected, the magnitude of the applied vibration is increased. Depending on the situation, there is a problem that the remaining correction is conspicuous.

Therefore, the present applicant detects a shake applied to a photographing apparatus including a shake correction mechanism, that is, a center frequency of a shake to be corrected, and changes a gain characteristic and a phase characteristic according to an output of the shake. Proposed to correct the gain shift and the phase shift of the center frequency of Japanese Patent Application No. Hei.
No. 3485), however, since the vibration frequency detecting means is independently provided outside, there is still room for improvement in terms of simplification of the configuration and adjustment.

An object of the present invention is to solve the above-mentioned problems and to further improve the prior application filed by the present applicant.

[0015]

In order to achieve the above object, the present invention provides a first detecting means for detecting vibration of a device,
Correction means for correcting the movement of the image due to the vibration; and first control for controlling the correction means based on an output of the first detection means and driving the correction means in a direction for correcting the movement of the image. Means, second detecting means for detecting the displacement of the vibration from the output of the first detecting means, and second means for controlling the characteristics of the first control means based on the output of the second detecting means. Control means, and the second detection
Means for integrating an output of the first detection means;
Means for detecting the frequency of vibration from the output signal of the integrating means
A vibration correction device having a frequency detecting means
It is. Further, the present invention provides a
First detecting means for detecting the vibration of the vessel,
Correction means for correcting the movement of an image, and the first detection means
The correction means is controlled based on the output of
Control for driving the correction means in a direction to correct
Means and the displacement of the vibration from the output of the first detecting means
Detecting means for detecting the output of the second detecting means;
Controlling a characteristic of the first control means based on a force;
Control means, and the second detection means includes the first detection means.
The frequency of the discrete value signal of the output of the
Is a shake correction device for detecting the vibration.

[0016]

As a result, appropriate shake correction can be performed for shakes of various frequencies.

[0017]

[0018]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of the present invention;

(First Embodiment) FIG. 1 is a block diagram showing a main configuration of a first embodiment of a shake correcting apparatus according to the present invention. In this figure, the same components as those of the preceding example shown in FIG. 19 are denoted by the same reference numerals, and detailed description thereof will be omitted.

In FIG. 1, an angular velocity detecting means 1 comprising an angular velocity sensor such as a vibrating gyroscope attached to a photographing device such as a camera, and a DC cut filter (or HPF) for cutting off a DC component of a velocity signal outputted from the angular velocity detecting means 1 ), An amplifier 3 for amplifying the angular velocity signal to a predetermined sensitivity,
The drive circuit 8 and the image correcting means 9 are the same as those described in the nineteenth embodiment.
The same configuration as that of the preceding example shown in the drawing can be used, and the difference in the present invention is the internal configuration of the microcomputer COM2 that controls the entire device. In this embodiment, for example, a variable apex prism (VAP) described later or a memory control method is used as the image correcting means 9.

Looking at the configuration in the microcomputer COM2, the A / D converter 4 for converting the angular velocity signal output from the amplifier 3 into a digital signal, the HPF 10 having the function of changing the characteristics in an arbitrary band, and the HPF 10 An integration circuit 5 that integrates the extracted signal of the predetermined frequency component to obtain an angular displacement signal at the frequency component, and detects the phase and gain of the integrated output signal output from the integration circuit 5, that is, the phase and gain of the angular displacement signal described later. There is provided a phase and gain correction circuit 11 for performing correction in accordance with the means, and a D / A converter 7 for converting an output signal of the phase and gain correction circuit 11 into an analog signal or a pulse output such as PWM and outputting the same.

Numeral 12 determines panning, tilting and a photographing state from the angular velocity signal output from the A / D converter 3 and the displacement signal output from the integration circuit 5,
This is a pan / tilt determination circuit that changes the characteristics of the HPF according to the determination result.

The specific operation of the pan / tilt determination circuit is as follows. The presence or absence of vibration of the angular velocity signal output from the A / D converter 4 and the angular displacement signal output from the integration circuit 5 are input so that the angular velocity is constant. If the angular displacement signal obtained by integrating the angular velocity signal indicates a monotonous increase, it is determined that the panning or the tilting is being performed.
The characteristic is changed so that the shake correction system does not respond to the low-frequency band.

When panning / tilting is detected, the VAP is centered at the center of the moving range. During this time, the detection of the angular velocity signal and the angular displacement signal is performed. When the panning / tilting is completed, the operation of lowering the low-frequency cut-off frequency again to extend the shake correction range is performed.

Reference numeral 13 denotes frequency detecting means for detecting a vibration applied to the apparatus from the angular velocity signal output from the A / D converter 3 and controlling the characteristics of the phase and gain correction circuit 11 according to the vibration frequency. .

Here, the drive circuit 8 and the image correcting means 9 for actually correcting the image blur according to the control signal output from the microcomputer COM2 will be described with reference to an example.
For example, those shown in FIGS.

FIG. 8 shows a variable apex prism (hereinafter referred to as VAP:
A control system that uses a voice coil as the drive system, uses an encoder to detect angular displacement and feeds back to the drive system to control the amount of drive. It is.

First, the VAP will be described in detail. As shown in FIG. 12, the variable apex angle prism 306 has a transparent high refractive index (refractive index n) between two opposing transparent parallel plates 340a and 340b. An elastic body or an inert liquid 342 is filled in a sandwiched manner, and the outer periphery thereof is elastically sealed with a sealing material 341 such as a resin film.
In this structure, the optical axes 0a and 340b are swingable, and the optical axis is displaced by swinging the transparent parallel plates 340a and 340b to correct the shake.

FIG. 13 shows the variable apex angle prism 30 of FIG.
6 is connected to the swing shaft 301 (31).
Incident light beam 34 when rotated by angle σ around 1)
FIG. 4 is a diagram showing a passing state of the light beam 4, and as shown in FIG. 4, a light beam 344 incident along an optical axis 343 is deflected by an angle φ = (n−1) σ according to the same principle as a wedge prism. Emit. That is, the optical axis 343 is decentered (deflected) by the angle φ.

Returning to the description of FIG. 8, the VAP described above
Reference numeral 306 is fixed to the lens barrel 302 via a holding frame 307 so as to be rotatable around 301 and 311 as axes.

313 is a yoke, 315 is a magnet, 3
Reference numeral 12 denotes a coil, which constitutes a voice coil type actuator capable of changing the apex angle of the VAP about the point 311 by flowing a current through the coil 312. 31
Reference numeral 0 denotes a slit for detecting the displacement of the VAP.
Rotating together with the holding frame 307, that is, the VAP 306, coaxially with the position 1, the position is displaced. 308 is a slit 310
309 is a light emitting diode for detecting the position of
sition Sensing Detector)
By detecting the displacement of the slit 310 together with 08, an encoder for detecting the angular displacement of the VAP apex angle is constituted.

The light beam whose incident angle has been changed by the VAP 306 is subjected to CC by the taking lens unit 303.
An image is formed on the imaging surface of the imaging device 304 such as D.

Reference numeral 305 denotes the center of the other rotation axis orthogonal to the rotation axis composed of the axes 301 and 311 of the holding frame 307.

Next, the basic configuration and operation of the control circuit for driving and controlling the VAP will be described with reference to the block diagram of FIG.

In FIG. 3, reference numeral 306 denotes a VAP; 322, an amplifier; 323, a driver for driving an actuator; 324, a voice coil type actuator for driving the VAP 306; 326, an encoder for detecting a vertical angle displacement of the VAP; Is a micro computer CO
The adder adds the control signal 320 for shake correction output from M2 and the output signal of the angular displacement encoder 326 in opposite polarities. The control signal 320 for shake correction and the angular displacement encoder output from the microcomputer COM2. 3
Since the control system operates so that the output signal of the encoder 26 becomes equal to the output signal of the encoder 326, the VAP 306 is driven so that the control signal 320 matches the output of the encoder 326. Is controlled.

FIG. 10 shows another example of the image correcting means, wherein the above-described VAP is driven not by a voice coil type actuator but by a stepping motor.

This is based on the rotation shaft 301 as the center of rotation.
The VAP is driven by the stepping motor 401 via the holding frame 307. That is, a stepping motor 401 having a lead screw 401a mounted on a rotation shaft thereof is attached to a support frame 403 attached to the lens barrel 302, and is movable in the optical axis direction by a guide shaft 405 of the support frame 403. The carrier 404 to be guided is always engaged with the lead screw 401a, and the carrier 404 is rotatably connected to a connecting rod 407 fixed to a support frame 307 with a rotating shaft 406, so that a step motor 401 is provided. Is rotated to move the carrier 404 in the optical axis direction, and the holding frame is turned around the turning shafts 301 and 311 via the connecting rod 407, so that the VA
P is driven. Reference numeral 402 denotes a reset sensor for detecting a reference position of the VAP. Note that a similar VAP drive mechanism is provided for the shaft 305, but a description thereof will be omitted.

The circuit configuration for driving and controlling the system of FIG. 10 is as shown in the block diagram of FIG.

In FIG. 11, the microcomputer C
The control signal 320 output from the OM2 is drive-calculated in the drive calculation circuit 410, converted into a VAP drive signal, and output to the driver IC 411. Then, the stepping motor 401 is driven by the driver IC to change the apex angle of the VAP.

FIG. 14 shows a third example of the image blur correcting means, in which image information is stored in a memory, and a cut-out range of the image from the memory is set to be smaller than that of the stored image. By shifting the image cutout position from the memory in a direction to cancel the motion of the image, the shake is corrected, and further, the cutout image signal is enlarged to correct the screen size, and then output. 2 illustrates a memory control type image blur correction unit. This method is characterized in that shake correction can be performed electronically without using an optical correction mechanism such as a VAP.

In the figure, reference numeral 100 denotes a zoom lens, 1
01 is an imaging device (CCD image sensor or the like) for converting an optical image into an electric signal, 102 is an A / D converter, and 103 is
Control signal (vibration signal) input from microcomputer COM2
In order to reduce the shake component in the image pickup signal based on 110, the cutout position of predetermined image information is shifted from the field memory 106 to correct the shake of the image, and the image is enlarged to the image read from the field memory 106. This is an image processing circuit that constitutes an angle-of-view correction processing unit that performs a process to perform a so-called electronic zoom and converts it to a normal screen size, and is realized by a micro computer.

Reference numeral 104 denotes an interpolation process for interpolating one pixel signal from image information of two or more adjacent pixels based on zoom information when performing electronic zoom for correcting an image read from the field memory to a normal angle of view. Means.
For this interpolation method, known means may be used.
For example, interpolation between pixels may be performed using an average value between adjacent pixels. Reference numeral 105 denotes a D / A converter, and reference numeral 107 denotes an encoder that detects a zoom magnification ratio of the zoom lens 100.

Next, the operation will be described.
The optical image passing through 0 is converted into an electric signal by the image sensor 101 and output as an image signal. The image signal is converted into a digital signal by the A / D converter 102, and image information for one field is written to the memory 106 via the memory control unit 103. Here, based on the shake signal input from the microcomputer COM2 and the zoom information from the encoder 107, the cutout position of the image signal from the field memory 106, that is, the readout range and the readout position are determined.

Next, in order to convert the signal read from the field memory 106 into the original size of the scanning width of the output image, that is, the angle of view, in accordance with the cut-out size, the number of pixels in a normal one pixel output period is converted. It is determined whether or not to output, and the interpolation processing is performed by the interpolation processing circuit 104 on the pixel having no pixel information. This signal is converted into an analog signal by the D / A converter 105 and output.

The specific example of the image correcting means for correcting the shake has been described above.

Next, the processing operation of the microcomputer COM2 in this embodiment shown in FIG. 1 will be described with reference to the flowchart of FIG. When the control is started, in step S201, the DC component is removed via the DC cut filter 2 and the amplifier 3 and the angular velocity signal from the angular velocity detecting means 1 amplified to a predetermined level is subjected to A / D conversion. The signal is converted into a digital signal by the device 4 and is taken into the microcomputer COM2.

Subsequently, in step S202, the panning / tilting and photographing state are performed by using the angular velocity signal and the angular displacement signal obtained by integrating the predetermined high frequency component extracted from the angular velocity signal by the HPF 10 by the integration circuit 5. Make a judgment.

In step S203, the coefficients for setting the characteristics of the HPF 10 are read from a table (not shown) prepared in the microcomputer COM2 in advance, according to the result of the determination. That is, if the HPF 10 is constituted by a digital filter, the characteristics of the HPF 10 can be freely changed by reading and setting a predetermined coefficient from a table storing the coefficient. These coefficients according to the panning / tilting and the photographing state are empirically obtained.

In step S204, the operation of the HPF 10 is performed based on the characteristic setting coefficient to set its characteristics. In step S205, the signal output from the HPF 10 is integrated by the integration circuit 5 to calculate the angular displacement signal ( (Shake signal).

In step S206, the frequency detecting means 1
3, the angular velocity signal output from the A / D converter 4 is calculated to detect the center frequency of the shake, and step S2 is performed.
In step 07, the phase and the correction coefficient of the gain correction circuit 11 corresponding to the center frequency of the shake obtained in step S206 are read from a table (not shown) prepared in the microcomputer COM2 in advance.

The phase and gain correction circuit 11 is for compensating the deterioration of the shake correction characteristic due to the phase delay of the shake correction system, has a phase lead element, and is constituted by, for example, a digital filter as described later. The digital filter reads the correction coefficient and sets the phase and gain correction characteristics corresponding to the shake frequency.

In step S208, step S207
The D / A converter 7 converts the calculation result obtained in step S209, that is, the corrected angular displacement signal into an analog signal,
Alternatively, the microcomputer COM2 may be used as a pulse output of PWM or the like.
Output more.

Since the HPF 10, the integrating circuit 5, and the phase and gain correcting circuit 11 use a digital filter or the like, the sampling time must be relatively high (for example, about 1 kHz). The judgment circuit 12 for judging the state and the frequency detection means 13 may perform processing with a relatively slow cycle (for example, 100 Hz). That is, it can be changed according to the situation.

As the frequency detecting means 13 using the angular velocity signal shown in FIG. 1, for example, a threshold value is provided near the center of the signal, and the detection can be performed based on the time at which the center intersects or the number of times the signal intersects at a certain time. However, in this method,
It depends on the stability of the DC signal. That is, in hand-held photographing and the like, since a low frequency component is considerably included, it is difficult to accurately detect a shake frequency.

Therefore, in this embodiment, the frequency is detected by paying attention to the increase / decrease of the signal for each sampling. That is, one set of signal increase / decrease is regarded as one shake, and the frequency is determined based on how many sets are detected within a predetermined time. In this scheme,
It is possible to detect every 1 Hz in 1 second and every 0.5 Hz in 2 seconds.

Here, the frequency detecting means 1 in the present embodiment
An example of the third method and the operation will be described with reference to the flowchart shown in FIG. Note that this process is repeatedly performed once every certain time.

In step S301, the frequency detection time T
In step S302, the reading of the clock function counter t, that is, the counting operation of the counter is started. In step S303, the frequency detection time T is compared with the clock counter t, and the count value of the clock counter is counted. It is determined whether or not t has reached a predetermined time T. If the clock counter t has reached T, the process proceeds to step S319.
If not reached, the process moves to step S304.

In step S304, "1" is added to the clock counter. Therefore, this "1" count up coincides with the processing time when the processing shown in the flowchart of FIG. 3 is executed once.

In step S305, an increase flag 1 for confirming whether the increase in the angular velocity signal has occurred up to the previous time.
To load. This increase flag is set to "H" when the increase has occurred up to the previous time, and is set to "L" when no increase has occurred in the past.

In step S306, it is determined by the flag 1 whether or not the increase in the angular velocity signal has occurred up to the previous time. If the flag 1 = “H”, it is determined that the increase has occurred in the past, and step S307 is determined. To flag 1 =
If “L”, it is determined that the increase has not occurred in the past, and the process proceeds to step S312.

If it is determined in step S306 that the increase in the angular velocity signal has occurred before the previous time, the process proceeds to step S30.
At 7, a decrease flag 2 is loaded to check if the decrease has occurred up to the previous time. Decrease flag 2
"H" is set if the decrease has occurred before, and "L" is set if the decrease has not occurred in the past.

In step S308, it is determined whether or not the decrease has occurred before the last time by using the flag 2, and the flag 2 =
If "H", that is, the decrease has occurred in the past, the flow proceeds to step S309.

In step S309, a shake counter N1 for counting the number of shakes (vibrations) is loaded. In step S310, "1" is added to the shake counter N1, and the process proceeds to step S311 to increase the increase flag. 1, the reduction flag 2 is reset, and the process is terminated.

On the other hand, when it is determined in step S306 that the increase flag 1 is not "H", and when it is determined in step S308 that the decrease flag 1 is not "H", that is, when neither increase nor decrease has occurred in the past. Moves to step S312, one sample before (previous processing)
Is loaded with the angular velocity data ω-1 at step S3, and then at step S3
The program proceeds to step 13 to load the current angular velocity data ω detected by the angular velocity detecting means 1.

In step S314, a threshold level a corresponding to the change that determines that the angular velocity data has increased or decreased within one sampling period is loaded. With the threshold level a and the sampling time,
The value can be set according to the frequency and the amplitude.

In step S315, the absolute value of the change in the angular velocity data within one sampling period is compared with the threshold level a.
4 and the processing is terminated. If the processing has reached (if the absolute value of the amount of change is equal to or greater than the threshold level a), step S3 is performed.
The process proceeds to step S16, and it is determined whether the change amount of the angular velocity during one sampling period is positive (increase) or negative (decrease). If it is positive, the process proceeds to step S317 to set the increase flag 1 to “H”. If it is not positive (if it is a decrease), the process proceeds to step S318, and the decrease flag 2 is set to “H”.
Then, the process proceeds to step S324 to end the process.

If the count value of the clock counter t has reached the frequency detection time T in step S303, the flow shifts to step S319 to load the shake counter N1, and in step S320 the shake is counted. The number of times N1 is divided by the detection time T to determine the number of shakes (shake frequency F) per unit time (one second).

Subsequently, in step S321, the shake counter N1 is cleared, and in step S322, the clock counter t is cleared.
Is cleared, and in step S323, the shake frequency F is stored in a predetermined storage area, and the flow shifts to step S304. Subsequent operations are as described above.

As described above, by considering one set of increase / decrease of the angular velocity signal as one vibration, it is easy to detect the swing in a low frequency range (for example, 1 Hz or less), and the threshold level is reduced. By setting a, the influence of the noise component can be removed. Further, there is an advantage that the processing can be easily realized by performing the processing using the microcomputer.

In this method, the detection accuracy can be varied by setting the frequency detection time T. For example, the resolution is 1 Hz for 1 second, and 0. 0 for 2 seconds.
The resolution is 5 Hz.

Therefore, by changing the frequency detection time T according to the detected shake frequency F, it is possible to realize detection accuracy according to the detection frequency. For example, when the detection frequency F is 〜１０10 Hz, T = 2 [SEC] and the detection is performed with an accuracy of 0.5 Hz. When the detection frequency F is 10 Hz or more, T = 1 [SEC] and the detection is performed with an accuracy of 1 Hz. Thus, the time required for detecting the shake frequency can be reduced.

According to this method, when a low frequency swing and a high frequency swing occur simultaneously, a high swing frequency can be extracted.

Normally, the correction frequency range in such a shake correction apparatus is about 1 to 15 Hz for camera shake, but depending on the shooting state, for example, an amplitude of about 3 to 5 Hz for still shooting, and 6 to 10 Hz for a vehicle. It is known that, depending on the skill of the photographer and the photographing state, the fluctuation of the relatively large amplitude is distributed in a range (band) where the frequency is narrow to some extent. In addition, when a tripod is used, high frequency vibration is a concern, and components are distributed to 20 to 30 Hz or more.

Here, the effect of the above embodiment will be described. First, since the center frequency of the shake is detected,
As described above, this is effective as one item for determining the shooting state.

Further, by combining with the phase and gain correction means first, the optimum correction can be made according to the skill of the photographer, the photographing state, and the like.

It is assumed that a sufficient effect of the shake correction can be obtained if the residual shake component can be reduced to -30 dB or less with respect to the added shake. This assumption is based on the fact that the focal length of the photographing device has a large effect on the shake correction effect. For example, if the focal length is simply twice, an image obtained without the double suppression effect can be obtained. This is because the same effect cannot be obtained.

However, at present, the total amount of residual shake after shake correction for the shake is -30 dB in the overall shake suppression effect including the shake sensor for detecting the shake and the image correction system.
Can be obtained only in a narrower range than the above-mentioned shake band (for example, 1 to 15 Hz) of the shake to be corrected.

For a specific example, it is assumed that the frequency characteristics obtained by the current angular velocity sensor and the image correcting means shown in FIG. 8 are as shown in FIG.

This shows the frequency characteristics of the image correction system with respect to the input of a sine wave-like shake applied to the shake sensor 1 which is the angular velocity detecting means in FIG. The upper figure shows the gain characteristic, and the lower figure shows the phase characteristic. The vertical axis represents gain (upper figure) and phase (lower figure), and the horizontal axis represents frequency (1 to 50 Hz).

FIG. 5 shows a characteristic 1 in which the range is expanded. Paying attention to the phase of the characteristic 1, the phase coincides at 3 Hz. At frequencies before and after that, the HPF 5 (here, the cutoff frequency is set to 0.06 Hz) or toward the lower frequency side, or The phase advances under the influence of the integrating circuit 6 (having a cutoff frequency of 0.07 Hz).
And the phase is delayed by the influence of the image correction means 9. The gain is almost constant from about 1 to 10 Hz.

FIG. 7 shows the suppression characteristics at this time.
6 shows the effect of the correction on the frequency (horizontal axis) (indicated by the vertical axis in dB). As can be seen from this figure, the best suppression can be achieved at 3 Hz where the phases match, and the suppression of -30 dB is almost achieved at 2 to 4 Hz, but at 10 Hz, the suppression effect is about -18 dB. ing.

That is, the phase shift causes the effect of suppressing the shake to deteriorate.

Here, in order to correct the phase delay of the characteristic 1, a characteristic having a phase lead element represented by a transfer function of H (S) = a (S + z) (1) If they are connected in series, the phase can be corrected at a predetermined frequency. This is to set the frequency higher than the range of shake suppression and advance the phase. This is performed by the phase and gain correction circuit 11.

For example, focusing on the phase delay at 10 Hz of the characteristic 1, the delay is about 7.5 deg. Therefore, 10
If the phase is advanced by 7.5 deg in Hz, this phase lag can be corrected, and eventually a sufficient vibration suppression effect can be obtained.

That is, the characteristic 2 is that the phase shift of the characteristic 1 is corrected at 10 Hz (the phase is matched and the gain is also matched). This is z (= 0) in equation (1).
Can be set by

Further, a is adjusted so that the gain becomes 0 dB at 10 Hz. If this is realized by the phase and gain correction circuit 11, the characteristic 1 can be changed to the characteristic 5, and the characteristic 7 showing the suppression effect can be obtained. It can be seen from the characteristic 7 that the best damping effect is obtained at around 10 Hz. The characteristic 7 indicates a residual vibration component, and is represented by 20 Log (OUT / IN) (2) OUT: residual vibration component after vibration correction IN: vibration amount

Similarly, at 20 Hz, if the characteristic 4 is realized by the phase and gain correction means 11 with respect to the characteristic 1,
Characteristic 5 can be obtained. Characteristic 8 is the suppression effect realized from this characteristic, and the best suppression effect is obtained near 20 Hz.

In order to realize the characteristic of equation (1) in the phase and gain correcting means 11, if a digital filter is used, desired characteristics can be set only by changing the coefficient (FIG. 1), which is suitable for control using a microcomputer. If a first-order IIR filter is used for the digital filter at this time, u0 = a0 · w0 + a1 · w1 w0 = e0 + a2 · w1 w1 = w0 (w1 is a state variable) e0: input u0: output a0 , A 1, a 1: Filter coefficients can be realized by the calculation of filter coefficients a 0, a 1, a 2
, The frequency characteristics can be set. Therefore, the data of the filter coefficients a0, a1, and a2 corresponding to the shake frequency may be prepared as a table, and the above-described IIR filter may be calculated using the filter coefficients obtained from the table. .

That is, according to the above-described embodiment, the center frequency of the shake is detected by the frequency detection circuit 13 from the angular velocity signal output from the angular velocity detection means 1, and the control system as a whole is detected at the center frequency of the shake. The filter characteristics of the digital filter of the phase and gain control circuit 11 are varied in order to provide the control system with phase lag and gain of 0 deg and 0 dB, that is, the best suppression characteristics are obtained. The best suppression effect can always be obtained for the shake frequency.

More specifically, the phase and gain control circuit 11
It is only necessary to read and set the filter coefficient stored in advance in the data table in the microcomputer according to the shake frequency so that the characteristics of the digital filter constituting the filter have frequency characteristics corresponding to each shake frequency. ,Constitution,
It is easy to control and is particularly suitable for processing by a microcomputer.

For example, when a video movie or the like is shot by hand, the camera shake is distributed over a relatively wide frequency range, but the shake having a relatively large amplitude is within a narrow range to some extent depending on the skill of the photographer and the shooting conditions. Distribute. Therefore,
If it can be configured to give the best correction effect to the center band of such shake, the optimum correction can be made according to the photographer's skill level and shooting conditions (handheld, mounted on a tripod, from the top of the vehicle, etc.) Becomes

(Second Embodiment) Next, a second embodiment of the present invention will be described with reference to FIG. 15. In the circuit configuration of the shake correction apparatus according to the second embodiment, the configuration of FIG. The difference lies in the processing inside the microcomputer. The output signal of the integrating circuit 5 in the microcomputer COM3 in FIG. 15 is input to the frequency detecting means 13 and used for the calculation. Both are the same as those in FIG. 1 and the description is omitted.

That is, in the present embodiment, in consideration of the angular velocity signal, since the gain and the sensitivity are low at low frequencies, the sensitivity and the accuracy are reduced. Therefore, the integrated output of the existing integrating circuit 5, that is, the angular displacement signal is used. Thus, the detection capability in the low frequency range is improved, and the frequency detection is performed using the angular displacement signal output from the integration circuit 5 in the same manner as the frequency detection based on the angular velocity signal of the first embodiment. Do.

If the angular velocity signal is used as it is, the detection capability in the low frequency range may be limited due to the dynamic range of the captured signal. Therefore, by performing detection using the angular displacement signal, the detection capability in a low frequency range is improved.

That is, if the amplitude of the angular velocity signal is the same, the signal becomes larger as the frequency becomes higher. However, the gain is limited to a certain value due to the configuration of the system. Then, a sufficient signal cannot be obtained in the low frequency range, and the detection capability is limited.

However, since the angular displacement signal is an integral signal,
If the vibration has the same amplitude, the amplitude is naturally constant regardless of the frequency. Therefore, it is suitable for frequency detection of relatively low-frequency vibration. However, as the frequency becomes higher, the vibration tends to be smaller, so that detection in a high frequency range becomes more difficult.

In consideration of such characteristics of both signals, the accuracy is improved in the entire low-frequency range and high-frequency range by detecting a shake using a portion having excellent detection capability of both signals. .

The detection from the angular displacement signal output from the integration circuit 5 is narrowed down to a low frequency range, and the sampling cycle of the signal is configured to be later than the sampling cycle of the detection of the angular velocity signal.

It is to be noted that, finally, the larger of those detected by both may be taken.

(Third Embodiment) FIG. 16 shows a third embodiment. This is a combination of the HPF 10 with the processing means at the time of panning and the phase and gain correction circuit 11. The feature is that there is no need to provide the gain and phase correction circuit 11 alone, and the frequency lower than the center frequency of the shake is reduced. Operability is improved due to cutoff (the lower the low-frequency vibration suppression performance,
Since the followability of the image blur correction device with respect to a camera or the like is improved, the operability is improved as the cutoff frequency is higher.
That is, the center frequency of the shake is sufficiently corrected, and if the center frequency of the shake is high, the cutoff frequency of the low frequency by the HPF increases, and the operability is improved.

Next, the microcomputer COM in this embodiment
The processing operation 4 will be described with reference to the flowchart of FIG.

In the figure, when the process is started, in step S401, the angular velocity signal from the angular velocity detecting means 1 via the DC cut filter 2 and the amplifier 3 is converted into a digital signal by the A / D converter 4 and taken in. In step S402, based on the angular velocity signal and the angular displacement signal, the panning / tilting and the shooting state are determined in the same manner as in the first embodiment. This is to judge hand-held shooting, shooting from a vehicle, and the like based on a change in shake, and to judge panning or tilting when the angular velocity signal is constant and the angular displacement signal monotonically increases.

In step S403, the angular velocity signal is calculated to detect the center frequency of the shake. Subsequently, in step S404, the characteristics of the HPF 10 are set from the results of the determination of the panning / tilting and photographing states and the result of the detection of the shake center frequency. The characteristic coefficient to be read out.

As a method of determining the coefficient at this time, there are various methods such as a method of searching a data table from both values, a method of comparing each characteristic and setting a coefficient of a frequency with a higher cutoff frequency, and the like. However, basically, the control is performed such that the cut-off frequency of the HPF goes to a higher frequency side during panning / tilting with reference to the cut-off frequency of the HPF obtained from the shake center frequency detection result.

The coefficients according to the panning / tilting and the photographing state have been empirically obtained.

In step S 405, the operation of the HPF 10 is performed based on the characteristic coefficient, and step S 4
At 06, the output signal of the HPF 10 is integrated by the integration circuit 5 and converted into an angular displacement signal (vibration signal).

In step S407, as a result of the integration operation,
In other words, the corrected angular displacement signal is converted into an analog value by the D / A converter 7 or output from the microcomputer COM4 as a pulse output such as a PWM signal.
The image correcting means 9 is operated in the direction of correcting the shake to perform the image stabilization operation.

(Fourth Embodiment) Further, an angular velocity detecting means,
When the frequency characteristics of the entire system are taken into consideration depending on the image correction system or a combination thereof, depending on the frequency band, for example, a region where the vibration correction is adversely affected by a shake of 30 Hz or more occurs.

This is because of the limitation on the detection characteristics of the angular velocity sensor constituting the angular velocity detecting means and the image correcting means such as VAP
When the vibration frequency increases, the shake detection system and the correction system cannot follow without delay, and the phase delay increases. At a certain frequency, the vibration phase and the drive phase of the VAP may be in phase, and conversely, the vibration may be greatly amplified.

If the frequency detecting means continuously detects a frequency in a region where the shake correction has an adverse effect, the operation of the image correction system is stopped to thereby stop the state of the adverse effect. Can be avoided.

The circuit configuration may be the circuit shown in FIG. 1. When the frequency detecting means detects the above-mentioned frequency band, the driving system 8 is controlled by, for example, a VAP correction system in the central state of correction. The hold signal may be calculated inside the microcomputer so as to be held and output to the drive circuit 8.

Next, the processing operation on the microcomputer in this embodiment will be described with reference to the flowchart of FIG. Note that the state in which the above-described shake correction has the opposite effect is referred to as 30
Hz or higher.

In FIG. 11, when the control is started, the angular velocity signal from the angular velocity detecting means 1 is converted into a digital signal by the A / D converter 4 through the DC cut film 2 and the amplifier 3 in step S501. take in.

In step S502, the shake frequency is set to 30.
If the frequency has reached Hz, the flow proceeds to step S511, and if not, the flow proceeds to step S503.

In step S503, the panning / tilting and the photographing state are determined based on the angular velocity signal and the angular displacement signal, and in step S504, a coefficient for determining the characteristics of the HPF 10 is read according to the result. The coefficients according to the panning / tilting and the photographing state are empirically obtained.

In step S505, the HPF is calculated using the characteristic coefficient. In step S506, the HPF1 is calculated.
The output signal of "0" is integrated by the integration circuit 5 and converted into an angular displacement signal (vibration signal). In step S507, the angular velocity signal is calculated to detect the vibration center frequency.

In step S508, step S507
A coefficient for phase and gain correction corresponding to the shake frequency obtained in is read out, and in step S509, a correction operation is performed based on the coefficient obtained in step S508.

In step S510, the obtained operation result, that is, the corrected angular displacement signal is converted into an analog value by the D / A converter 7, or PW
A pulse output of M or the like is output from the microcomputer and supplied to the drive circuit 8, the image correction means 9 such as the VAP is driven, and the process is terminated.

If the shake frequency F is equal to or higher than 30 Hz in step S502, that is, if it is determined that the shake correction is in a frequency range in which the shake correction has an adverse effect, step S502 is performed.
The flow shifts to step 511, where the image correction system is reset, and the correction system is held at the middle point of the shake correction range.
In step 2, the center frequency of the shake is detected from the angular velocity signal, and the process is terminated, and the process of the flowchart in FIG. 18 is terminated once.

The flowchart of FIG. 18 shows one process, and in fact, the process of this flowchart is repeatedly performed.

The resetting means of the image correction system will be described.
When a motor is used, if no signal is sent, the motor is stationary in that state, so it is returned to the center position and stopped there.

When a voice coil motor as shown in FIG. 4 is used in the image correction system, the control signal 32
It is sufficient to output the center signal to 0. However, due to the relationship between the mass of the VAP or the like and the torque of the voice coil, when a high frequency is applied, it may not be possible to sufficiently maintain the center position. Therefore, although not shown, it is desirable to use a mechanical lock mechanism that mechanically holds the VAP.

[0123]

As described above, according to the present invention,
An object of the present invention is to provide a shake correction device capable of performing appropriate shake correction for shakes of various frequencies.

[0124]

[Brief description of the drawings]

FIG. 1 is a block diagram showing a first embodiment of a shake correction apparatus according to the present invention.

FIG. 2 is a flowchart for explaining a control operation in the first embodiment.

FIG. 3 is a flowchart for explaining a shake frequency detection operation in the first embodiment.

FIG. 4 is a frequency characteristic diagram of gain and phase of the angular velocity sensor.

FIG. 5 is a frequency characteristic diagram of gain and phase for explaining the operation of the shake correction apparatus according to the present invention.

FIG. 6 is a frequency characteristic diagram of gain and phase for explaining the operation of the shake correction apparatus according to the present invention.

FIG. 7 is a frequency characteristic diagram showing a suppression characteristic of the shake correction apparatus according to the present invention.

FIG. 8 uses VAP as an image correction means,
FIG. 3 is a diagram illustrating an example in which a voice coil type actuator is used for a drive system.

FIG. 9 is a block diagram showing a driving circuit of an image correcting means using the VAP of FIG. 8;

FIG. 10 is a diagram showing an example in which a VAP is used as an image correction means and a step motor is used in a drive system.

11 is a block diagram showing a driving circuit of an image correcting means using the VAP of FIG.

FIG. 12 is a diagram illustrating the configuration and operation of a VAP.

FIG. 13 is a diagram illustrating the configuration and operation of a VAP.

FIG. 14 is a block diagram showing an example in which a memory control method for electronically performing shake correction is used as image correction means.

FIG. 15 is a block diagram showing the configuration of a second embodiment of the shake correction apparatus according to the present invention.

FIG. 16 is a block diagram showing the configuration of a third embodiment of the shake correction apparatus according to the present invention.

FIG. 17 is a flowchart for explaining the operation of the third embodiment shown in FIG. 16;

FIG. 18 is a flowchart showing a fourth embodiment of the shake correction apparatus according to the present invention.

FIG. 19 is a block diagram showing a configuration of a conventional shake correction apparatus.

Claims (3)

(57) [Claims] A first detecting unit for detecting a vibration of a device; a correcting unit for correcting a motion of an image caused by the vibration; First control means for driving the correction means in a direction for correcting the motion of the image; second detection means for detecting the displacement of the vibration from the output of the first detection means; and second detection means and a second control means for controlling a characteristic of said first control means based on the output, the second detection means, the product of the output of said first detection means
Integral means for performing minute operation, and vibration from an output signal of the integral means.
And a frequency detecting means for detecting the frequency of the vibration. 2. A first detecting means for detecting vibration of a device.
Correcting means for correcting the movement of the image due to the vibration , and controlling the correcting means based on an output of the first detecting means.
Controlling the correction means in a direction for correcting the motion of the image.
First driving means for driving, and detecting displacement of the vibration from an output of the first detecting means.
Second detection means, and the first control means based on an output of the second detection means.
Second control means for controlling the characteristics of the stage, wherein the second detection means separates the output of the first detection means.
A shake correction apparatus , wherein a frequency is detected by increasing or decreasing a signal of a scattered value . 3. The shake correction apparatus according to claim 1 , wherein a sampling time for frequency detection is changed according to an output of the first detection unit and an output of the second detection unit.