JPH0734581B2 - Imaging device - Google Patents

Description

TECHNICAL FIELD The present invention relates to a photographing device such as a video camera, and in particular,
The present invention relates to an image pickup apparatus having a vibration isolation function that can obtain a stable image even when the image pickup apparatus main body is subjected to disturbance vibration or swing.

2. Description of the Related Art In recent years, the performance of video equipment has been remarkably improved, and high-quality images have become extremely easy to obtain. Along with this, sophisticated photography techniques are required. Against such a background, there has been proposed an image pickup apparatus having an image stabilization function capable of obtaining a stable image with little screen shake regardless of the shake of the photographer and the image pickup apparatus.

Hereinafter, a conventional imaging apparatus having a vibration isolation function will be described with reference to the drawings. FIG. 13 is a block diagram showing a conventional image pickup apparatus having a vibration isolation function. Reference numeral 1301 denotes a lens barrel portion on which a plurality of lenses and an image sensor are mounted. 1302 is the lens barrel
The counterweight for 1301 is mechanically coupled to the lens barrel 1301 by a connecting rod 1303. Reference numeral 1305 is a joint that rotatably connects the connecting rod 1303 and the support rod 1304. The photographer operates the photographing apparatus by supporting the support rod 1304. In the above configuration, the lens barrel 1301, the connecting rod 1304, the counterweight 1302
The counterweight 1302 is adjusted so that the center of gravity of the movable portion 1306 configured by is located near the joint 1305.

Then, the movable portion 1306 has a large moment of inertia around the joint 1305. Therefore, even if the support bar 1304 supported by the photographer is tilted by some disturbance, the movable portion 1306 is acted by the action of the moment of inertia of the movable portion 1306.
That is, the posture of the lens barrel portion 1301 leans against a constant angle. Therefore, it is possible to obtain a stable image with little screen shake even when the photographer swings. (For example, John Jurgens, "Design of Steadicam" SMP
The E-Journal, Vol. 87, September 1978, page 587 (John Jurgens "Steadicam as a Design Problem",
SMPTE jounal Vol.87, Sep, 1978, P587)) Problems to be Solved by the Invention However, the conventional configuration as described above has a problem in that its vibration isolation characteristics are fixed. For example, when a photographer holds an image capturing device in his / her hand and takes an image while walking, the image capturing device shakes in a cycle in which the photographer visits, but it is desirable that the screen does not shake. On the other hand, when the photographer performs pan photographing, it is desirable that the photographing screen respond quickly in the direction intended by the photographer. If the image stabilization characteristic is selected so as to suppress the screen shake under the former photographing condition in the above conventional example, the photographing screen does not respond well as the photographer intends under the latter photographing condition. On the contrary, if the anti-vibration characteristic is selected according to the latter shooting condition, it is possible to quickly respond in the direction intended by the photographer, but under the former shooting condition, sufficient anti-vibration effect cannot be obtained. .

Further, according to the above-mentioned conventional example, it is necessary to provide the counterweight 1302, which makes it difficult to reduce the size and weight.

Means for Solving the Problems In order to solve the above-mentioned problems, a photographing apparatus according to the present invention is a lens barrel portion in which a plurality of lenses and an image pickup element are mounted, and an image signal from an electric signal obtained by the image pickup element. Image signal processing means for producing the above, a support member for rotatably supporting the lens barrel portion around a rotation axis orthogonal to or substantially orthogonal to the incident light ray axis to the lens barrel portion, the lens barrel portion and the support member. An actuator means mounted between the two to rotate and drive the lens barrel portion, a relative angle detecting means for detecting a relative angle between the lens barrel portion and the support, and an angular velocity with respect to the inertial coordinate of the lens barrel portion. Lens barrel angular velocity detection means, support angular velocity detection means for detecting the angular velocity of the support with respect to inertial coordinates, and when the absolute value of the output signal of the support angular velocity detection means exceeds a predetermined value, it is determined to be moving photography. And the relative Shooting mode determination means for determining still shooting when the output signal of the angle detection means is within a predetermined range and the absolute value of the output signal of the support angular velocity detection means is less than or equal to a predetermined value; The photographing mode includes a combining means for generating and outputting a combined signal from the output signal of the angular velocity detecting means and the output signal of the relative angle detecting means, and a driving means for amplifying the combined signal and supplying electric power to the actuator means. When the discriminating means discriminates the still photographing, the control gain of the synthesizing means is adjusted to control the lens barrel portion to stand still in the inertial coordinate, and when the photographing mode discriminating means discriminates the moving photographing, the synthesizing means. The control gain is adjusted so as to match the angles of the lens barrel and the support with respect to the inertial coordinates.

Operation According to the present invention, when the photographing mode discrimination means discriminates the still photographing by the above-mentioned constitution, the lens barrel part is controlled to be stationary or substantially stationary at the inertial coordinate, and the photographing mode discrimination means is operated. When it is determined that the shooting is a moving image, the respective angles of the lens barrel portion and the support body are smoothly matched in the inertial coordinates, and the lens barrel portion is controlled so as to follow the support body well. To do. Therefore, it is possible to obtain a stable image with little screen shake regardless of the shake of the photographer and the photographing device during still photographing, and operability is not impaired during moving photographing such as pan photographing and tilt photographing. It is possible to provide an imaging device that can be made compact and lightweight.

Embodiment An image pickup apparatus according to an embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of an image pickup apparatus in one embodiment of the present invention. In FIG. 1, a large number of lens groups (not shown) and an image sensor 2 (for example, CCD
A plate or a shooting tube) is attached to collect reflected light from a subject to form an image on the image pickup device 2 and a charge signal (electrical signal)
Convert to. The image signal processing circuit 10 sequentially reads out the charge signals obtained by the image pickup device 2 to obtain an image signal (video signal).
Is producing.

An actuator 5 is arranged between the lens barrel portion 1 and the support body 3 to rotate and drive the lens barrel portion 1 in the yaw direction around a rotation shaft 6 (in the use state, the lens barrel portion 1 is substantially horizontal). It can be freely rotated on. The rotation shaft 6 of the actuator 5 passes through the center of gravity G of the lens barrel portion 1 and is rotatably supported by the support body 3. Further, the support body 3 is provided with a grip portion 4 that is manually supported by the operator of the photographing apparatus.

2 (a), (b), and (c) show a specific configuration of the actuator 5. In FIG. 2, the back yoke 201 made of a ferromagnetic material of the magnet 202 is attached to the lens barrel portion 1 and rotates together with the rotating shaft 6. The magnet 202 is magnetized to have four poles and generates a magnetic field flux. Bearing 20 of rotating shaft 6
In the coil yoke 203 to which 7 is attached, the coils 204a, 2
04b and Hall element (magnetism sensitive element) 9 are fixed. In this example, the magnet 202 is attached to the lens barrel portion 1, and the coil yoke 203 is attached to the support body 3. The coils 204a and 204b are connected in series, and a rotational torque is generated by the current flowing from the terminals 205 to 206 and the magnetic flux of the magnet 202. The Hall element 9 is arranged so as to substantially face the magnetic pole switching portion of the magnet 202.
(Angle θm of barrel 1) and coil yoke 203 (support 3
The output signal corresponding to the relative angle difference θmo (= θo−θm) of the angle θo of is generated. It should be noted that θm is the angle of the lens barrel portion 1 around the rotation axis 6 viewed from the inertial coordinate, and θo is the angle of the support 3 around the rotation axis 6 viewed from the same inertial coordinate.

The output signal a of the Hall element 9 that detects the magnetic flux of the magnet 202 of the actuator 5 is input to the relative angle detection circuit 11. FIG. 3 shows a specific configuration of the relative angle detection circuit 11. The DC signal obtained at the two output terminals of the hall element 9 is differentially amplified by a predetermined multiple by the differential amplifier circuit including the operational amplifier 301 and the resistors 302, 303, 304, 305 to obtain the output signal c. + VH, -VH are suitable power supplies, resistors 306, 30
An appropriate bias voltage is applied to the Hall element 9 via 7.

Further, the angular velocity sensor 7 composed of a vibration type gyro is attached to the lens barrel portion 1 by a fixing member 8. The detection axis of the angular velocity sensor 7 coincides with the rotation axis 6 of the actuator 5, and outputs the output signal b in response to the rotation angular velocity around the rotation axis 6 of the lens barrel portion 1 in the inertial coordinates. The output signal b of the angular velocity sensor 7 is input to the angular velocity detection circuit 12,
Angular velocity ω around the rotation axis 6 of the lens barrel 1 viewed from the inertial coordinate
A signal d proportional to m is obtained. FIG. 4 shows a specific configuration of the angular velocity detection circuit 12. The forced vibration circuit 401 has a sine wave oscillating circuit having a predetermined frequency (for example, 1 kHz), and forcibly vibrates the drive element 402 made of the piezoelectric element of the angular velocity sensor 7 by the oscillating frequency signal. Since the sense element 403 made of a piezoelectric element is arranged in mechanical contact with the drive element 402, it vibrates with the drive element 402 at the same frequency. At this time, when the lens barrel unit 1 rotates about the rotation axis 6 in the inertial coordinate, a mechanical Coriolis force is generated. Since the Coriolis force is proportional to the product of the angular velocities of the two orthogonal axes of the sense element 403, it is proportional to the product of the angular velocity ωm around the rotation axis 6 of the lens barrel portion 1 in inertial coordinates and the angular velocity due to forced vibration. The sense element 403 causes a mechanical strain due to the Coriolis force and generates an electric signal due to a piezoelectric action. The output of the sense element 403 is synchronously detected by the synchronous detection circuit 404 at the same frequency as the forced vibration, and the low-pass filter 405 extracts the low-frequency component (for example, DC to 100 Hz) of the detected output. A signal d proportional to the angular velocity ωm about the rotation axis 6 of 1 is obtained.

The output signal c of the relative angle detection circuit 11 and the output signal d of the lens barrel angular velocity detection circuit 12 are input to the support angular velocity detection means 17. The support angular velocity detection means 17 outputs a signal u proportional to the angular velocity in the inertial coordinates of the support 3 operated by the photographer from these input signals. The output signal u of the support angular velocity detection circuit 17 is input to the photographing mode determination means 13. The photographing mode discrimination means 13 discriminates whether the photographer is performing still photography or pan photography from these input signals. Here, the still shooting means that the photographer operates the photographing device without changing the photographing direction,
The pan photographing means that the photographer operates the photographing apparatus so as to rotate the photographing direction in the yaw direction.

The discrimination result of the photographing mode discrimination means 13 is input to the synthesizing means 14. Further, the output signal c of the relative angle detection circuit 11 and the output signal d of the lens barrel angular velocity detection circuit 12 are input to the synthesizing means 14. When the shooting mode determination means 13 determines that the shooting mode is determined to be still shooting, the control calculation means 14 determines the relative angle detection circuit 11
Output signal c and the output signal d of the lens barrel angular velocity detection circuit 12
And the result of integration calculation of the output signal c of the relative angle detection circuit 11 are respectively added with a first predetermined gain, and when the photographing mode discrimination means 13 discriminates the pan photographing, the relative angle detection circuit 11 Output signal c, the output signal d of the angular velocity detection circuit, the result of the integral calculation of the output signal c of the relative angle detection circuit 11, and the angular velocity command obtained from the output signal c of the relative angle detection circuit 11, respectively. Addition is performed with a predetermined gain, and the calculation result is output as an output signal e.

FIG. 5 shows the configuration of the photographing mode discrimination means 13 and the composition means 14. In this embodiment, the photographing mode discrimination means 13 and the control calculation means 14 are composed of A / D converters 502 and 503, a calculation device 501, a memory 504 and a D / D converter.
It is composed of an A converter 505. The A / D converter 502 produces a digital signal p corresponding to the value of the output signal c of the position detection circuit 11. Further, the A / D converter 503 produces a digital signal q corresponding to the value of the output signal d of the angular velocity detection circuit 12. The arithmetic unit 501 operates according to a predetermined built-in program, which will be described later, stored in the ROM area (read-only memory area) of the memory 504, and outputs the digital signal p of the A / D converter 502 and the digital signal of the A / D converter 503. RAM for signal q
The combined digital signal w is fetched into the area (random access memory area), subjected to a predetermined calculation, and then combined.
It outputs to the D / A converter 505, and the synthetic signal e is obtained.

FIG. 6 shows a specific configuration of the A / D converter 502 (the same applies to the A / D converter 503). The input signal c and the output signal m of the D / A conversion circuit 607 are compared by the comparator 601, and the comparator signal n corresponding to the magnitude relation is obtained.
The oscillator circuit 605 generates a clock pulse 1 having a predetermined frequency. The signal h from the calculator 501 is normally "H" (high potential state), and becomes "L" (low potential state) when the digital signal p is read. Therefore, the inverter circuit 602 and the AND circuits 603 and 604 input the clock pulse 1 to the down pulse input terminal D or the up pulse input terminal U of the counter circuit 606 according to the comparator signal n. (When the signal h is "H"). Counter circuit 606
The internal state is decremented by 1 by the input pulse to the down pulse input terminal D, and the internal state is incremented by 1 by the input pulse to the up pulse input terminal U. The internal state of the counter circuit 606 is output as a digital signal p, and is converted into an analog signal m according to the digital signal p in the D / A converter 607. As a result, the counter circuit
The digital signal p of 606 becomes a value corresponding to the input signal c. The arithmetic unit 501 sets the signal h to "L" for a predetermined short time to stop the operation of the counter circuit 606 and read a stable digital signal p. Similarly, the calculator 501
Makes the signal k "L" for a predetermined short time so that a stable digital signal q is read.

The output signal e of the synthesizing means 14 is input to the drive circuit 16, and a voltage signal (or current signal) f proportional to the signal e is supplied to the coils 204a and 204b of the actuator 5. FIG. 7 shows a specific configuration of the drive circuit 16. The operational amplifier 701, the transistors 704 and 705, and the resistors 702 and 703 form a power amplification circuit, which outputs a voltage signal f obtained by amplifying the signal e by a predetermined number.

Now, the built-in program of the arithmetic unit 501 will be described.
First, an outline will be described based on the basic flowchart shown in FIG.

In process 801, an interrupt from the timer is awaited.
The timer generates an interrupt signal every predetermined time ΔT, and shifts to 2 when an interrupt occurs. That is, the following processing is performed at the sampling time ΔT.

In process 802, the digital signal p corresponding to the relative angle θmo between the lens barrel 1 and the support 3 is fetched from the A / D converter 502,
Store in memory 504. Further, the digital signal q corresponding to the angular velocity ωm viewed from the inertial coordinate of the lens barrel portion 1 is fetched from the A / D converter 503 and stored in the memory 504.

In process 803, the digital signal p corresponding to the relative angle θmo between the lens barrel portion 1 and the support body 3 and the digital signal q corresponding to the angular velocity ωm of the lens barrel portion 1 from the inertial coordinate are used to determine the support seen from the inertial coordinate. The angular velocity ωo of the body 3 is calculated, and it is determined from the value of the angular velocity ωo of the support body 3 whether the photographing is still photography or pan photography. Then, if it is determined to be still shooting, the process proceeds to 4, and if it is determined to be pan shooting, the process proceeds to 5.

In process 804, a control calculation is performed so that the lens barrel unit 1 is stationary or substantially stationary in the inertial coordinates, and a signal e to be a command input of the drive circuit 16 is generated. After this processing, shifts to.

In process 805, control calculation is performed so that the lens barrel portion 1 smoothly follows the support body 3, and a signal e which is a command input of the drive circuit 16 is generated. After this process, the process shifts to 1.

Incidentally, the processing 803 is the photographing mode discrimination means 13 in FIG.
The process 804 and the process 805 correspond to the control calculation means 14 in FIG.

Next, the processes 802 to 805 will be described in detail with reference to FIG.

FIG. 9A is a detailed flowchart of the process 802. First, the content of the variable Qn that holds the value of the digital signal p corresponding to the relative angle θmo between the lens barrel 1 and the support 3 at the previous sampling is stored in the variable Qn ₋₁ . Then signal h
Is set to "L", a digital signal p corresponding to the relative angle θmo between the lens barrel 1 and the support 3 is newly fetched and stored in the variable Qn, and then the signal h is set to "H" again.

Further, the content of the variable Wn holding the value at the time of the previous sampling of the digital signal q corresponding to the angular velocity ωm of the lens barrel portion 1 is stored in the variable Wn ₋₁ . Next, the signal k is set to “L”, a digital signal q corresponding to the angular velocity ωm of the lens barrel unit 1 is newly taken in and stored in the variable Wn, and then the signal k is set to “H” again. That is, at this point in time, the lens barrel portion 1 and the support body 3 are present.
That is, the information of the relative angle θmo of the above is stored in the variable Qn, and the information of the angular velocity ωm of the lens barrel portion 1 is stored in the variable Wn. Further, each information before one sampling is stored in the variable Qn _-1 and the variable Wn _-1 .

FIG. 9B is a detailed flowchart of the process 803. First, in process 901, relative to the lens barrel 1 and the support 3 one sampling before, from the variable Qn holding a value corresponding to the current relative angle θmo (= θo−θm) between the lens barrel 1 and the support 3. A variable Qn- ₁ holding a value corresponding to the angle θmo is subtracted, and the result is stored in the variable ΔQ. Next, in process 902, the variable Δ is set to the variable Wn representing the angular velocity ωm of the lens barrel 1 with respect to the inertial coordinate.
A value ΔQ / ΔT obtained by dividing Q by the sampling period ΔT is added. That is, the following calculation (1) is performed, and the calculation result is stored in the variable Wo.

Wo = Wn + ΔQ / ΔT (1) In the formula (1), the term [ΔQ / ΔT] is the lens barrel 1 and the support 3
Therefore, if the sampling period ΔT is sufficiently small, the second item of the equation (1) is the relative angular velocity (ωo−ω) of the lens barrel portion 1 of the support body 3.
m), and [Wn + ΔQ / ΔT] corresponding to the sum of this term and the angular velocity ωm of the lens barrel portion 1 indicates a value corresponding to the angular velocity ωo of the support 3. Therefore, the variables in equation (1)
Wo represents the angular velocity ωo with respect to the inertial coordinate of the support 3. In process 903, the content of the variable PFL is "1".
When it is, it indicates that it is determined to be pan shooting, and when it is "O", it indicates that it is determined to be still shooting.

Now, if PFL = 1 in the result of the processing 803 at the previous sampling, the result of the processing 903 is
Move to 904. In process 904, it is determined whether or not the absolute value of the variable Qn exceeds the predetermined value QAIS. If the absolute value of the variable Qn does not exceed the predetermined value QAIS, the process proceeds to processing 905, and if it exceeds, the process proceeds to processing 909. Variable in process 905
It is determined whether or not the absolute value of Wo exceeds a predetermined value WAIS. If the absolute value of the variable Wo does not exceed the predetermined value WAIS, the process proceeds to step 906, and if it exceeds, the process proceeds to step 909.
In process 906, the content of the variable NAIS is incremented by "1". Process 907
Then, it is determined whether the variable NAIS exceeds a predetermined value TAIS. If the variable NAIS exceeds the predetermined value TAIS, process 908
To 910, and if not exceeded, to 910. Processing 9
In 08, the variable PFL is set to "0". In process 909, the variable NAIS is set to "0". The variable NAIS acts as a kind of counter, and the series of processing shown in FIG.
Processing is performed since timer interrupts occur every T.
The process 907 corresponds to measuring time ΔT × TAIS.

That is, when it is determined that the pan shooting is performed (PFL =
1), the absolute value of the relative angle θmo between the lens barrel 1 and the support 3 is less than or equal to a predetermined value (corresponding to QAIS), and the support 3
When the absolute value of the angular velocity ωo of is less than or equal to the predetermined value (corresponding to WAIS) continues for a predetermined time (corresponding to TAIS) or more, it is determined that the still image capturing is performed (PFL = 0), and the panning is performed. While shooting, the above conditions are not satisfied and PFL = 1.

On the contrary, if PFL = 0 in the result of the previous processing 803, the processing proceeds to processing 911 as a result of processing 903. Processing 911
Then, it is determined whether or not the absolute value of the variable Wo exceeds a predetermined value WPAN. If the absolute value of the variable Wo exceeds the predetermined value WPAN, the process proceeds to step 912, and if not, the process proceeds to step 915. In process 912, the content of the variable NPAN is incremented by "1". In process 913, it is determined whether the variable NPAN exceeds a predetermined value TPAN. If the variable NPAN exceeds the predetermined value TPAN, the process proceeds to step 914, and if not, the process proceeds to step 915.
In process 914, the variable PFL is set to "1". Variable NP in process 915
Set AN to "0". The variable NPAN functions as a kind of counter, and the series of processing shown in FIG. 8 is performed by the timer interrupt that occurs every time ΔT, so the processing 91
Processes 2 to 913 correspond to measuring time ΔT × TPAN. In process 914, the variable PFL is set to "1". In process 915, the variable NPAN is set to "0".

That is, when it is determined that the robot is stationary (PFL =
0), the absolute value of the angular velocity Wo of the support 3 is a predetermined value (W
It is determined that the pan shooting has been performed (PFL = 1) for the first time after a predetermined time (corresponding to TPAN) for more than a predetermined time (corresponding to PAN).

Then, in process 910, the process proceeds to the next process based on the determination result of still image capturing or pan image capturing. That is, if it is determined that the still shooting is performed in the processing up to the preceding stage, 4
Then, the processing shifts to step 804 and the processing 804 is performed.

FIG. 9C is a detailed flowchart of the process 804. First, the variable En that holds the value of the digital signal w corresponding to the signal e that is the command input of the drive circuit 16 one sampling before is stored in the variable En ₋₁ . Next, the lens barrel 1 and the support 3
A variable Qn that holds the value corresponding to the current relative angle θmo of
Is subtracted from the variable Qn _-1 which holds the value corresponding to the relative angle θmo of the barrel 1 and the support 3 before one sampling, and the result is stored in the variable ΔQ. Next, from the variable Wn that holds the value corresponding to the current angular velocity ωm of the lens barrel unit 1 to the lens barrel unit 1
The variable Wn ₋₁ holding the value corresponding to the angular velocity ωm of one sampling before is subtracted, and the result is stored in the variable ΔW.
Then, using K1, K2, Ta, ΔT as constants, the following calculation (2)
And the calculation result is stored in the variable ΔE.

ΔE = -K1 × ΔW + K2 × (ΔQ + ΔT × Qn / Ta)
(2) Next, the variable ΔE and the variable En ₋₁ are added, and the result is newly stored in the variable En. Finally, the content of the variable En is output as a signal w to the D / A converter 505 (see FIG. 5). After that, the process shifts to 1 and waits for the next timer interrupt (see FIG. 8).

FIG. 9D is a detailed flowchart of the process 805. First, the variable En that holds the value of the digital signal w corresponding to the signal e that is the command input of the drive circuit 16 one sampling before is stored in the variable En ₋₁ . Next, the lens barrel 1 and the support 3
A variable Qn that holds the value corresponding to the current relative angle θmo of
Is subtracted from the variable Qn _-1 which holds the value corresponding to the relative angle θmo of the barrel 1 and the support 3 before one sampling, and the result is stored in the variable ΔQ. Next, a variable Wn _-1 holding a value corresponding to the angular velocity ωm of the lens barrel 1 one sampling before is subtracted from a variable Wn holding a value corresponding to the current angular velocity ωm of the lens barrel 1 The result is stored in the variable ΔW. Then, using K3, K4, Tw, Tp, ΔT as constants, the following calculation (3)
And the calculation result is stored in the variable ΔE.

ΔE = −K3 × (ΔW−ΔQ / Tw) + K4 × (ΔQ + ΔT × Qn / Tp) (3) Next, the variable ΔE and the variable En _-1 are added, and the result is newly stored in the variable En. . Finally, the content of the variable En is output as a signal w to the D / A converter 505 (see FIG. 5). After that, the process shifts to 1 and waits for the next timer interrupt (see FIG. 8).

Hereinafter, the control operation and the image stabilization effect at the time of still image capturing will be described in detail. FIG. 10 is a control block diagram of the image pickup apparatus in this embodiment during still image pickup. In the figure, S represents a Laplace operator. further,
In this figure, the delay element due to sampling and the frequency dependence of the low-pass filter 405 of the lens barrel angular velocity detection circuit 12 are omitted because they can be sufficiently ignored in the frequency region described below. Now, the relative angle θmo between the angle θm of the lens barrel portion 1 and the angle θo of the support body 3 when viewed from the inertial coordinates is easily detected by the Hall element 9 that detects the magnetic field of the magnet 202 of the actuator 5. The Hall element 9 and the relative angle detection circuit 11 are represented by a block 1001 and obtain a signal c (the output signal of the position detection circuit 11) that is K _Q · A _Q times θmo. K _Q is the magnetic flux density / voltage conversion gain of the Hall element 5, and A _Q is the voltage gain of the relative angle detection circuit 11. On the other hand, the angular velocity ωm of the lens barrel 1 viewed from the inertial coordinate is determined by the angular velocity sensor 7 and the angular velocity detection circuit.
Detected by 12 and represented by block 1008,
A signal d that is K _W · A _W times ωm is obtained. K _W is the angular velocity / voltage conversion gain of the angular velocity sensor 7, and A _W is the voltage gain of the angular velocity detection circuit 12. Further, the signal c is multiplied by K2 (block 1002) by the processing 804 (corresponding to a portion surrounded by a broken line) of the arithmetic unit 501, and the result and the result are integrated with a gain of 1 / Ta (block 1003). And the signal d multiplied by −K1 (block 1009) are combined at the addition point 1010,
Obtain the combined signal e. Block 1004 corresponding to drive circuit 16
At, the signal e is amplified Ae times to obtain the voltage signal f. In the block 1005 corresponding to the actuator 5, the voltage signal f is converted into the torque Tm. Here, R is a combined resistance value of the coils 204a and 204b, and Kt is a torque constant. A block 1006 represents the transmission from the torque Tm to the angular velocity ωm due to the mechanical moment of inertia Jm of the lens barrel portion 1, and a block 1007 represents the relationship between ωm and θm. Therefore, the transfer characteristic from θo to θm from this block diagram is as shown in FIG. 11th
In the figure, r is the frequency of the fluctuation of θo. here,
The frequencies ₁ and ₂ at the break points can be expressed as follows using the gain constants shown in FIG.

Actually, ₁ = 0.18Hz and ₂ = 10Hz. Then, the transfer characteristic of the rotation angle θm of the lens barrel portion 1 with respect to the rotation angle θo of the support 3 in the inertial coordinate is the first break point frequency.
₁ (0 dB) in the frequency range of ₁ or less, ₁
As a result, in the frequency range below the _second corner frequency ₂ , −6d
Attenuates at B / oct, -12 dB / oct in the frequency range of _{2 and} above
Is decaying at. From FIG. 11, the amount of transmission from the vibration of θo to the vibration of θm becomes smaller in the frequency range of ₁ or more. The extent is represented by the difference ZdB between OdB and the characteristic line (in this case Z = 15dB at r = 1Hz).

The fluctuation of the shooting device during still shooting is mainly 0.5Hz to 5Hz
It is known to be distributed in the range of. Therefore, if the vibration-proof characteristics of the present photographing device are set as shown in FIG. 11, the rotation angle θm of the lens barrel portion 1 hardly changes regardless of the fluctuation of the rotation angle θo of the support body 3, and the fluctuation of the photographing screen. It can be seen that is significantly reduced. That is, stable and easy-to-view video shooting becomes possible.

In addition, each term in the mixing operation (2) has the following meaning depending on the processing (Em = En ₋₁ + ΔE) in the next stage.

That is, the first term [−K1 × ΔW] in the equation (2) is a component provided for setting the angular velocity ωm of the lens barrel portion 1 to “0” and keeping the lens barrel portion 1 stationary in the inertial coordinates. The term [K2 × ΔQ] in the second term of the equation (2) is the angle θm in the inertial coordinate of the lens barrel 1 and the inertial coordinate of the support 3 when the relative angle θmo between the lens barrel 1 and the support 3 is “0”. The angle θo at is a component provided to make the states substantially coincide. Where gain K2 is gain K1
It is set to be much smaller than. This is because the main purpose is to set the angular velocity ωm of the lens barrel portion 1 to “0”. That is, when K2 is increased, the followability of the angle θm of the lens barrel portion 1 with respect to the angle θo of the support 3 is improved, and the vibration isolation effect is reduced. Therefore, the value of K2 is selected to be sufficiently small, and when the torque Tm generated by the actuator 5 due to the term [K2 × ΔQ] is small and the loss around the bearing of the actuator 5 is large, the generated torque Tm must exceed the loss. Without this, the relative angle θmo between the lens barrel 1 and the support 3 cannot be set to “0”. That is, the angle θm of the lens barrel portion 1
And the angle θo of the support 3 has a steady deviation ωm = 0.
Will be controlled. Therefore, the term [K2 × ΔT × Qn / Ta] in the second term of the equation (2) is provided to compensate for impairing the effect of the term [K2 × ΔQ] due to the bearing loss of the actuator 5. That is, the term [K2 × ΔT × Qn
/ Ta] is θm by the next process (En = En _-1 + ΔE)
It has the meaning of an integral calculation of o and compensates for a steady deviation between the angle θm of the lens barrel 1 and the angle θo of the support 3. That is, θ
When there is a steady deviation between o and θm, the term [K2 × ΔT × Qn / Ta] is added to the variable En for each sampling. As a result, the actuator 5 according to the term [K2 × ΔQ]
Even if the torque generated by is less than the bearing loss,
Since the term [K2 × ΔT × Qn / Ta] is added to En at each sampling, the torque generated by the actuator 5 increases with time, overcoming the bearing loss, and eliminating the steady deviation between θo and θm. it can.

Now, the control operation at the time of pan photographing will be described. First, since the lens barrel unit 1 is controlled to stand still in the inertial coordinate as described above during still image capturing,
If the photographer operates the support body 3 from this state (still photography) to start pan photography, the lens barrel portion 1 is controlled so as not to follow the support body 3 and stand still (actually, the mixing calculation ( 2)
(It slightly moves due to the action of the term [K2 × ΔQ] and the term [K2 × ΔT × Qn / Ta]). The relative angle θmo between the lens barrel portion 1 and the support 3 gradually increases. However, the angular velocity ω of the support 3 is determined by the photographing mode determination means 13 (process 803).
The absolute value of o is a predetermined value (corresponding to QPAN in Fig. 9 (a))
When it exceeds, it is determined to be pan photographing, and the control calculation is performed by the process 805. In process 805, the control calculation is performed so that the angle θm of the lens barrel portion 1 smoothly matches the angle θo of the support body 3 and follows well. This will be described in detail below.

Each term in the mixed operation (3) is processed in the next process (En
= En _-1 + ΔE) has the following meaning.

That is, the first term of the equation (3) [−K3 × (ΔW−ΔQ / Tw)]
Serves to match the angular velocity ωm of the lens barrel portion 1 with the angular velocity ωo of the support 3 with a time constant Tw. This is because in the steady state (when the pan photographing is determined and the processing 805 is always performed), this item is “0”. Therefore, the following equation (4) is obtained.

ΔW = ΔQ / Tw (4) From the above description, the equation (4) can be rewritten as the following equation (5).

Wn−Wn ₋₁ = (Qn−Qn ₋₁ ) / Tw (5) Further, the equation (4) can be rewritten as the following equation (5).

(Wn−Wn ₋₁ ) / ΔT = [Wn + (Qn−Qn ₋₁ ) / ΔT−Wn] / Tw (6) Equation (6) is obviously a difference equation, and the limit ΔT of the sampling period ΔT → When 0 is taken, the equation (6) becomes a differential equation of the following equation (7) where t is time.

dWn / dt = (Wn + dQn / dt−Wn) / Tw (7) where Wn corresponds to the angular velocity ωn of the lens barrel 1, and Qn corresponds to the relative angle θmo of the support 3 to the lens barrel 1. Therefore, the equation (7) can be rewritten as the following equation (8).

dωm / dt = (ωm + dθmo / dt−ωm) / Tw (8) Now, in the equation (8), the term [dθmo / dt] represents the relative angular velocity between the lens barrel portion 1 and the support 3. [Ωm + d, which is the sum of this term and the angular velocity ωm of the lens barrel portion 1,
θmo / dt] represents the angular velocity ωo of the support 3.
Therefore, the equation (8) becomes the following equation (9).

dωm / dt = (ωo-ωm) / Tw (9) When the differential equation (9) is taken with respect to ωm, the following equation (10) is obtained.
To get

ωm = ωo × (1−e ^{−t / Tn} ) (10) That is, in equation (10), the angular velocity ωm of the lens barrel 1 is the time constant Tw.
Indicates that it matches the angular velocity ωo of the support 3.
That is, ΔQ / Tw in the equation (3) gives an angular velocity command to the lens barrel portion 1. However, the term [−K3 × (ΔW−ΔQ /
Tw)] alone, the angle θm of the lens barrel portion 1 and the angle θ of the support body 3
O does not match, and is in the form of being shifted by a difference at the time when the shooting mode determination means 13 determines the pan shooting. Therefore, the term [K4 × ΔQ] is the relative angle θ between the lens barrel 1 and the support 3.
It is provided so that mo is “0” and the angle θm in the inertial coordinate of the lens barrel portion 1 and the angle θo in the inertial coordinate of the support body 3 are substantially matched. However, if the gain K4 is increased, the angle difference between the angle θm of the lens barrel portion 1 and the angle θo of the support 3 at the time when the pan photographing is detected is steeply controlled to “0”. Has a feeling of strangeness because the shooting screen suddenly flows. Conversely, when the gain K4 is reduced, the torque generated by the actuator 5 due to the term [K4 × ΔQ]
When Tm is small and the bearing loss of the actuator 5 exists, the angle θm of the lens barrel portion 1 and the angle θo of the support body 3 do not match. Therefore, the term [K4 × ΔT × Qn / Tp] is provided. This term [K4 × ΔT × Qn / Tp] means the integral calculation of the relative angle θmo between the lens barrel portion 1 and the support body 3 by the next process (En = En ₋₁ + ΔE). Angle θm
And a steady deviation between the angle θo of the support 3 is compensated. That is, if there is a steady deviation between θo and θm, the term [K4 × ΔT × Qn / Tp] is added to the variable En for each sampling. As a result, even when the torque generated by the actuator 5 due to the term [K4 × ΔQ] is smaller than the bearing loss, the term [K4 × ΔT × Qn / Tp] is added to En for each sampling, The torque generated by the actuator 5 increases with time and exceeds the bearing loss.
It leads to the elimination of the steady deviation of m.

By the way, when the photographer operates the photographing apparatus of the present embodiment in a state of still photography, pan photography, and still photography again, the angle θo in the inertial coordinate of the support body 3 and the angle θm in the inertial coordinate of the lens barrel portion 1 The movement will be described with reference to FIG. FIG. 12 shows the angle θo in the inertial coordinates of the support body 3 and the angle in the inertial coordinates of the lens barrel portion 1 when the photographer operates the photographing apparatus of the present embodiment such as still photographing, pan photographing, and still photographing. It shows the movement of θm. The curve 1201 is the angle θo in the inertial coordinate of the support 3.
The photographer performs still photography until time t ₁ , and pan photography starts from this time (t ₁ ) at time t ₃
Indicates that the camera is in still photography again. A curve 1202 shows the movement of the lens barrel 1 at the angle θm in the inertial coordinates. When the pan photographing is started at time t ₁ , the photographing mode discriminating means 13 (process 803) still discriminates the still photographing at this point, so that the lens barrel 1 moves slightly as described above. The ability to follow 3 is poor and the lens barrel 1
And the relative angle θmo of the support 3 increases the spread. Eventually, after a time corresponding to TPAN determined by the process 803, at time t ₂ , the relative angle θmo between the lens barrel portion 1 and the support 3 increases to a value corresponding to θPAN determined by the process 803. When it is determined that the pan photographing is performed, the control calculation is entrusted to the process 805. Then, as described above, the angle θm of the lens barrel portion 1
Is smoothly controlled with respect to the angle θo of the support 3 and is controlled so as to follow well. Then, at the time t ₃ , when the photographer enters the still shooting, TAIS determined by the process 803 is set.
After a time corresponding to, at the time t ₄ , the process 803 determines again that the still image shooting is performed. In the above operation, the angle θm of the lens barrel portion 1 slightly exceeds the angle θo of the support body 3 (overshoot).
This is the integral term [K of the mixed operation (3) in process 805.
4 × ΔT × Qn / Tp], but the degree of the overshoot is slight and the photographer does not feel uncomfortable.

In the above description, the embodiment of the present invention is applied to the shake prevention in the yaw direction and the pan photographing, but it goes without saying that the invention can also be applied to the shake prevention in the pitch direction and the tilt photographing. Further, in the above description, the angular velocity of the support is calculated by the support angular velocity detecting means from the relative angular difference between the support and the lens barrel and the angular velocity of the lens barrel, but an angular velocity sensor is newly attached to the support. Needless to say, it may be directly detected. Further, the application range of the present photographing device is not limited to the video camera, and various modifications can be made without changing the gist of the present invention.

EFFECTS OF THE INVENTION As described above, the anti-vibration mechanism of the image pickup apparatus of the present invention does not require the counterweight of the lens barrel portion required in the conventional example, and can be reduced in size and weight. Moreover, the number of sensors is small and the cost is low. Further, since the relative position is detected by the Hall element (magnetism sensitive element) that detects the magnetic field of the magnet of the actuator, the configuration is simple and the number of parts is small. Further, the photographing apparatus of the present invention has a photographing mode discrimination means, and achieves both the vibration-proof effect required for still photographing and the good followability required for pan photographing. Therefore, if, for example, a video camera is constructed according to the present invention, it can be easily reduced in size and weight.
You can get a high-performance video camera with anti-vibration mechanism.

[Brief description of drawings]

FIG. 1 is a block diagram of a photographing device according to an embodiment of the present invention,
2 is a configuration diagram showing a specific configuration of the actuator of FIG. 1, FIG. 3 is a circuit diagram showing a specific configuration of the relative angle detection circuit of FIG. 1, and FIG. 4 is a mirror of FIG. FIG. 5 is a configuration diagram showing a specific configuration of the tubular portion angular velocity detection circuit, FIG. 5 is a configuration diagram showing a specific configuration of the photographing mode discrimination means and the composition means in FIG. 1, and FIG. 6 is A / in FIG. FIG. 7 is a configuration diagram showing a specific configuration of the D converter, FIG. 7 is a circuit diagram showing a specific configuration of the drive circuit of FIG. 1, and FIG. 8 is stored in the ROM area of the memory of FIG. A basic flowchart of the built-in program, FIGS. 9 (a), (b), (c), and (d) are detailed flowcharts of the processes of FIG. 8, and FIG. Block diagram, FIG. 11 is a Bode diagram for explaining the operation of FIG. 1, FIG. 12 is an operation explanatory diagram for explaining the operation of FIG. 1, and FIG. It is a block diagram of the image capturing device in come. 1 ... Lens barrel, 3 ... Support, 5 ... Actuator,
6 ... Rotation axis, 7 ... Angular velocity sensor, 9 ... Hall element, 10 ... Image signal processing circuit, 11 ... Relative angle detection circuit, 12 ... Lens barrel angular velocity detection circuit, 13 ... Shooting mode discrimination Means, 14 ... Control computing means, G ... Center of gravity of lens barrel 1,
16 ... Drive circuit, 501 ... Calculator, 502,503 ... A / D converter, 504 ... Memory, 505 ... D / A converter.

Claims (5)

[Claims] 1. A lens barrel portion on which a plurality of lenses and an image pickup device are mounted, image signal processing means for generating an image signal from an electric signal obtained by the image pickup device, and an axis of a light ray incident on the lens barrel portion. Alternatively, a support body that rotatably supports the lens barrel portion around substantially orthogonal rotation axes, and an actuator means that is mounted between the lens barrel portion and the support body and that drives the lens barrel portion to rotate, Relative angle detecting means for detecting a relative angle between the lens barrel and the support, lens barrel angular velocity detecting means for detecting an angular velocity of the lens barrel with respect to the inertial coordinates, and angular velocity with respect to the inertial coordinates of the support. When the absolute value of the output signal of the support angular velocity detection means and the support angular velocity detection means exceeds a predetermined value, it is determined to be moving photography, and the output signal of the relative angle detection means is within a predetermined range and Body angular velocity detection hand When the absolute value of the output signal becomes less than or equal to a predetermined value, a photographing mode discrimination means for discriminating still photography and an output signal of the lens barrel angular velocity detection means and a combined signal from the output signal of the relative angle detection means are combined. It is composed of a synthesizing means for generating and outputting, and a driving means for amplifying the synthesized signal and supplying electric power to the actuator means. When the photographing mode discrimination means discriminates the still photographing, the control gain of the synthesizing means is adjusted. The lens barrel is controlled so as to be stationary in the inertial coordinate, and when the photographing mode discrimination means discriminates the moving photographing, the control gain of the synthesizing means is adjusted to adjust the angle between the lens barrel and the support with respect to the inertial coordinate. An image pickup apparatus, which is controlled so as to match with each other. 2. The support angular velocity detecting means is configured to add the differential value of the output signal of the lens barrel angular velocity detecting means and the differential value of the output signal of the relative angle detecting means. The image capturing apparatus according to item (1). 3. The photographing apparatus according to claim 1, wherein the rotation axis of the actuator section passes through the center of gravity of the lens barrel section or in the vicinity of the center of gravity. 4. The photographing apparatus according to claim 1, wherein an angular velocity sensor using a vibration type gyro is used as the angular velocity detecting means. 5. The transfer characteristic of the lens barrel rotation angle with respect to the support rotation angle in the inertial coordinate is set to 1 in the frequency range of the first break point frequency f1 or lower, and at the f1 or higher, the second break point frequency f2 ( The image pickup apparatus according to claim (1), characterized in that it is attenuated at −6 dB / oct in a frequency range of f1 <f2) or less and is attenuated at −12 dB / oct in a frequency range of f2 or more.