JP2021027484A - Imaging apparatus and control method thereof - Google Patents

Abstract

To reduce image blurring without being greatly affected by noise included in an output signal of vibration detection means.SOLUTION: An imaging apparatus 11a includes: motion vector detection means 15 that detects a motion vector using an output from an imaging device 12a capturing a subject image; camera vibration proof means 114 that performs a vibration proof operation, namely moving the imaging device in such a manner that image blurring is reduced in accordance with a first signal indicating the motion vector; communication means 115a that receives, from an interchangeable lens 11b, information indicating whether the interchangeable lens is in the vibration isolation operation; and gain control means 115a, 16, 17, 112, and 113 that acquire gain in accordance with the first signal detected when the interchangeable lens is in vibration isolation operation. The communication means transmits the gain acquired by the gain control means to the interchangeable lens.SELECTED DRAWING: Figure 1

Description

The present invention relates to an image pickup device such as a digital camera or a digital video, and more particularly to an image pickup device capable of vibration isolation operation for reducing (correcting) image shake.

In Patent Document 1, a motion vector is detected from a video signal generated by using an image sensor, and the motion vector is used to obtain an output signal from a vibration detection unit (angular velocity meter) that detects vibration of the image pickup device. A method of correction is disclosed.

Japanese Unexamined Patent Publication No. 2012-256077

However, the method disclosed in Patent Document 1 increases the noise contained in the output from the vibration detection unit, and there is a possibility that the image shake is not satisfactorily reduced due to the influence thereof.

The present invention provides an image pickup apparatus capable of reducing image shake without being significantly affected by noise contained in an output signal of a vibration detection unit.

In the imaging device as one aspect of the present invention, an interchangeable lens can be attached and detached. The image sensor has an image sensor that captures a subject image, a motion vector detection means that detects a motion vector using an output from the image sensor, and a first signal indicating the motion vector to reduce image shake. A camera anti-vibration operation that moves the image sensor to the camera, a communication means that receives information from the interchangeable lens indicating whether or not the interchangeable lens is in the anti-vibration operation, and an interchangeable lens that performs the anti-vibration operation. It has a gain control means for acquiring a gain according to a first signal detected while the lens is inside. The communication means is characterized in that the gain acquired by the gain control means is transmitted to the interchangeable lens.

Further, the interchangeable lens as another aspect of the present invention can be attached to and detached from the image pickup apparatus, and is generated by using the vibration detecting means for detecting the vibration of the interchangeable lens and the output from the vibration detecting means. The lens anti-vibration means for moving the anti-vibration lens in response to the second signal and the gain for the second signal in the anti-vibration operation of the lens anti-vibration means are changed according to the output from the image pickup apparatus. It is characterized by having a gain changing means.

Further, the control method as another aspect of the present invention is an image pickup device in which an interchangeable lens can be attached and detached, and an image pickup element for capturing a subject image and a motion vector for detecting a motion vector using an output from the image pickup element. It is applied to an image pickup apparatus having a detection means and a camera anti-vibration means for performing an anti-vibration operation of moving an image sensor so as to reduce image shake in response to a first signal indicating a motion vector. The control method includes a step of receiving information from the interchangeable lens indicating whether or not the interchangeable lens is in the vibration isolation operation, and a first signal detected when the interchangeable lens is in the vibration isolation operation. Correspondingly, it is characterized by having a step of acquiring a gain and a step of transmitting the acquired gain to an interchangeable lens.

A computer program that causes the computer of the image pickup apparatus to execute a process according to the control method also constitutes another aspect of the present invention.

According to the present invention, image shake can be satisfactorily reduced without being significantly affected by noise contained in the output of the vibration detecting means.

The block diagram which shows the structure of the camera system in Example 1 of this invention. The figure which shows the signal processing waveform in Example 1. FIG. The figure explaining the polarity determination in Example 1. FIG. The figure which shows the polarity determination waveform in Example 1. FIG. The flowchart which shows the process in Example 1. The figure which shows the relationship between the exposure time and the extraction frequency in Example 1. FIG. The block diagram which shows the structure of the camera system in Example 2 of this invention. The figure which shows the relationship between the focal length and the extraction frequency in Example 2. FIG. The figure explaining the runout correction position and runout correction width in Example 3. The flowchart which shows the process in Example 2.

Hereinafter, examples of the present invention will be described with reference to the drawings.

FIG. 1 shows the configuration of a camera system 11 composed of a camera body 11a according to a first embodiment of the present invention and an interchangeable lens 11b detachably attached to the camera body 11a. The camera CPU 115a provided on the camera body 11a controls the operation of each part of the camera body 11a shown in the broken line frame 116a in the drawing. Further, as a communication means, the camera CPU 115a communicates with the lens CPU 115b provided on the interchangeable lens 11b to exchange various commands and data, and also has the inside of the interchangeable lens 11b shown in the broken line frame 116b via the lens CPU 115b. Control the operation of each part of.

The interchangeable lens 11b has an imaging optical system 19. The image pickup optical system 19 forms an image of light from a subject (not shown) and forms a subject image on an image pickup element 12a included in an image pickup unit 12 provided in the camera body 11a. The imaging optical system 19 includes an anti-vibration lens 19a.

The vibration detecting unit (vibration detecting means) 18 provided in the interchangeable lens 11b includes an angular velocity sensor such as a vibration gyro that outputs a signal (hereinafter referred to as an angular velocity signal) indicating the angular velocity of the camera shake due to camera shake or the like, and the angular velocity sensor. It is composed of a processing circuit that integrates the angular velocity signals from the above and generates an angular velocity signal (second signal) indicating the displacement of the angular velocity. The angular vibration signal from the vibration detection unit 18 is input to the signal processing unit 111, and is also input to the polarity determination unit 113 in the camera body 11a via the lens CPU 115b and the camera CPU 115a.

The signal processing unit 111 calculates the target shift position, which is the target movement position of the anti-vibration lens 19a for reducing (correcting) the image shake due to the camera shake, by the angular shake signal from the vibration detection unit 18, and the target shift Outputs a runout correction signal indicating the position. Further, the signal processing unit (gain changing means) 111 performs gain processing on the vibration correction signal (which will be described later), and outputs the vibration correction signal that has undergone the gain processing to the lens vibration isolation unit 110.

The lens anti-vibration unit (lens anti-vibration means) 110 movably holds the anti-vibration lens 19a in a direction orthogonal to the optical axis 10 of the imaging optical system 19, for example, in the direction of the arrow 19b (shift direction), and also anti-vibration. It has a shift actuator such as a VCM that drives the lens 19a in the shift direction. The lens vibration isolation unit 110 drives the shift actuator so as to move (shift) the vibration isolation lens 19a to the target shift position indicated by the vibration correction signal received from the signal processing unit 111. As a result, the interchangeable lens 11b is subjected to a vibration isolation operation for correcting the image shake caused on the image sensor 12a due to the camera shake. The lens CPU 115b transmits the operating state of the lens anti-vibration unit 110 (whether or not the anti-vibration operation is in progress, etc.) to the camera CPU 115a by communication.

The image sensor 12a performs photoelectric conversion (imaging) of the subject image and outputs an image pickup signal as an electric signal. The image processing unit 13 provided in the camera body 11a performs various image processing on the image pickup signal to generate a video signal. The video signal is displayed on a display unit (not shown) or recorded on a recording medium by the storage unit 14. Further, the video signal is output to the vector processing unit (motion vector detecting means) 15.

The vector processing unit 15 detects a motion vector indicating the motion of a feature point between two temporally continuous frame images included in the video signal, and a signal indicating the motion vector (first signal: hereinafter, motion vector). A signal) is output to a camera vibration isolator (camera vibration isolator) 114, a narrow band filter (first filter) 16, and a polarity determination unit (determination means) 113 provided in the camera body 11a.

The camera vibration isolator 114 movably holds the image sensor 12a in a direction orthogonal to the optical axis 10, for example, in the direction of the arrow 12b (shift direction), and drives the image sensor 12a in the shift direction, such as a shift actuator such as a VCM. Has. The camera vibration isolation unit 114 moves (shifts) the image sensor 12a to a target shift position, which is a target movement position corresponding to a signal indicating a motion vector (hereinafter, referred to as a motion vector signal) detected by the vector processing unit 15. Drive the shift actuator. In this way, in the camera body 11a, the camera anti-vibration unit 114 uses the detected motion vector to perform the anti-vibration operation.

As described above, the lens vibration isolation unit 110 performs the vibration isolation operation in the interchangeable lens 11b, but the image vibration may not be completely corrected and the image vibration may remain. In this case, the remaining image runout (hereinafter referred to as runout correction remaining) is detected as a motion vector by the vector processing unit 15. The vector processing unit 15 inputs a motion vector signal corresponding to the remaining vibration correction to the camera vibration isolation unit 114. The camera vibration isolation unit 114 performs a vibration isolation operation that shifts the image sensor 14 to a target shift position corresponding to the motion vector signal. As a result, the residual vibration correction is reduced, and good vibration isolation performance can be obtained.

FIG. 2A exemplifies the waveform 20 of the motion vector detected by the vector processing unit 15. Here, a motion vector whose size changes in the direction and magnitude shown on the vertical axis with time shown on the horizontal axis is shown.

The motion vector signal from the vector processing unit 15 is also output to the narrow band filter 16. FIG. 2B shows the motion vector waveform extracted by the narrow band filter (first filter) 16 in the waveform 21 of FIG. 2B. The narrow band filter 16 extracts a motion vector component of a predetermined frequency band from the motion vector signal from the vector processing unit 15, and outputs the motion vector component to the rectifying processing unit (rectifying means) 17. The predetermined frequency band referred to here is variably set from, for example, 1 Hz to 7 Hz according to the imaging conditions of the camera system 11.

The frequency band of parallel runout, which is one of the camera shakes, which will be described later, is generally distributed from 1 Hz to 7 Hz, and the narrow band filter 16 attenuates the motion vector component of a frequency band other than that frequency band (predetermined frequency band). The noise superimposed on the vibration detection unit 18 is cut. Comparing the waveform 20 of FIG. 2A and the waveform 21 of FIG. 2B, in the waveform 21 extracted by the narrow band filter 16, the motion vector component having a frequency lower than the predetermined frequency band is removed. The waveform is close to a sine wave within a predetermined frequency band.

There are two types of camera shake: angular runout and parallel runout. The angular runout can be detected by calculation using the angular velocity signal from the vibration detection unit 18 provided on the interchangeable lens 11b, but the parallel runout cannot be detected. .. As a result, under imaging conditions in which parallel runout becomes large (for example, close-up imaging such as macro imaging), the lens vibration isolator 110 in which the runout correction signal generated by the signal processing unit 111 is input from the angular velocity signal is a camera including parallel runout. Image shake due to runout cannot be sufficiently corrected.

Therefore, in this embodiment, the image shake due to the parallel runout is detected as a motion vector by the vector processing unit 15, and the motion vector signal is input to the rectification processing unit 17 via the narrow band filter 16. The rectification processing unit 17 rectifies (smooths) the motion vector signal of the predetermined frequency band extracted by the narrow band filter 16 as shown in the waveform 22 in FIG. 2C, and biases the smoothed signal value. It is input to the comparison processing unit (processing means) 112 as (third signal). The bias value is a value corresponding to the amplitude of the motion vector signal in a predetermined frequency band.

The comparison processing unit 112 outputs to the signal processing unit 111 a multiplication coefficient as an output according to the bias value from the rectification processing unit 17 and the determination result of the polarity determination unit 113 described later. The multiplication coefficient is used to change the gain for the angular vibration signal from the vibration detection unit 18 in the vibration isolation operation of the lens vibration isolation unit 110. The comparison processing unit 112 outputs the multiplication coefficient to the signal processing unit 111 every predetermined time (for example, 0.2 seconds) to update the gain.

The signal processing unit 111 receives the gain for the angular vibration signal from the vibration detection unit 18 in the vibration isolation operation of the lens vibration isolation unit 110 (that is, from the angular vibration isolation signal to the vibration isolation lens 19a) according to the multiplication coefficient from the comparison processing unit 112. Change the gain when calculating the target shift position). As described above, the vibration detection unit 18 cannot detect the parallel vibration, but the lens vibration isolation unit 110 also corrects the parallel vibration by changing the gain for the angular vibration signal from the vibration detection unit 18 based on the motion vector signal. be able to. This is because the parallel runout is synchronized with the angular runout, so changing the gain for the angular runout is equivalent to detecting the parallel runout.

In this way, by smoothing the output of the narrow band filter 16 by the rectification processing unit 17, the amplitude of the motion vector signal that changes in the cycle of camera shake is taken as the average value. Then, the signal processing unit 111 changes the gain of the runout correction signal based on the average value. A specific example of the multiplication coefficient used for changing the gain will be described later. The polarity determination unit 113 determines whether the motion vector signal from the vector processing unit 15 is insufficient or excessive with respect to the correction of the remaining runout correction. The polarity determination unit 113 is provided in order to prevent the image shake correction from being overcorrected by excessively increasing the gain of the vibration detection unit 18.

FIG. 3 shows the configuration of the polarity determination unit 113. As described above, the motion vector signal from the vector processing unit 15 and the angular deflection signal from the vibration detection unit 18 are input to the polarity determination unit 113, and these are narrowband filters 113a, which have the same characteristics as the narrowband filter 16, respectively. Processed by 113b. As a result, a motion vector signal and an angular deflection signal in a predetermined frequency band can be obtained.

FIGS. 4A and 4B show the waveform 41 of the angular deflection signal output from the vibration detection unit 18 and the waveforms 42 and 43 of the motion vector signal output from the vector processing unit 15. FIG. 4A shows a case where the waveform 41 of the angular deflection signal and the waveform 42 of the motion vector signal are in phase, and FIG. 4B shows that the phases are inverted (opposite phases). Show the case. The fact that the waveform 41 of the angular runout signal and the waveform 42 of the motion vector signal match in phase indicates that the motion vector signal is insufficient for the rest of the runout correction. On the contrary, the fact that the phases of the waveform 41 of the angular runout signal and the waveform 43 of the motion vector signal are opposite to each other indicates that the motion vector signal is excessive (overcorrected) with respect to the remaining runout correction. ..

The multiplication unit 113c multiplies the waveform 41 of the angular deflection signal and the waveforms 42 and 43 of the motion vector signal. FIG. 4C shows the multiplication result of the waveform 41 of the angular deflection signal and the waveforms 42 and 43 of the motion vector signal. The waveform 44 is the result of multiplying the waveform 41 of FIG. 4A and the waveform 42, and the waveform 45 is the result of multiplying the waveform 41 of FIG. 4B by the waveform 43.

The integration unit 113d obtains the time integration values of the waveforms 44 and 45. From this, it can be determined whether the image shake correction is insufficient or excessive. That is, if the time integration values of the waveforms 44 and 45 are positive, the image shake correction within the integration time is insufficient, and if it is negative, it is excessive. If the time integration value is 0, there is no excess or deficiency in image shake correction. In this way, the polarity determination unit 113 outputs to the comparison processing unit 112 a determination result indicating whether the image shake correction (that is, the vibration isolation operation of the lens vibration isolation unit 110) is insufficient, excessive, or excessive or insufficient.

The comparison processing unit 112 outputs the following to the signal processing unit 111 provided on the interchangeable lens 11b according to the determination result from the polarity determination unit 113. If the image shake correction is insufficient, a multiplication coefficient larger than 1 is output, if there is no excess or deficiency, a multiplication coefficient 1 is output, and in the case of overcorrection, a multiplication coefficient smaller than 1 is output. At this time, the comparison processing unit 112 outputs a larger multiplication coefficient as the insufficient amount of image shake correction indicated by the bias value from the rectification processing unit 17 is larger, and outputs a smaller multiplication coefficient as the overcorrection amount is larger. The camera CPU 115a may transmit the multiplication coefficient from the comparison processing unit 112 to the signal processing unit 111 via the lens CPU 115b.

As described above, the signal processing unit 111 changes the gain for the angular deflection signal from the vibration detection unit 18 according to the multiplication coefficient from the comparison processing unit 112. Specifically, the changed gain is calculated by multiplying the gain reference value by the multiplication coefficient.

In this embodiment, the gain control means is configured by the camera CPU 115a, the narrow band filter 16, the rectification processing unit 17, the comparison processing unit 112, and the polarity determination unit 113.

The vibration isolation control process including the gain process will be described with reference to the flowchart shown in FIG. The camera CPU 115a as a computer executes this process according to a computer program.

When the user half-presses the release button (not shown) provided on the camera body 11a, in step S501, the camera CPU 115a performs an imaging preparation operation such as AF or AE, and activates the vibration detection unit 18 via the lens CPU 115b. The lens anti-vibration unit 110 is started to perform anti-vibration operation. Further, the camera CPU 115a causes the vector processing unit 15 to detect a motion vector from the video signal generated by the image processing unit 13 using the image pickup signal output from the image sensor 12a, and generate the motion vector signal. Further, the camera CPU 115a causes the narrow band filter 16 to extract a motion vector signal in a predetermined frequency band, smoothes the motion vector signal extracted by the rectification processing unit 17, and transfers the bias value obtained thereby to the comparison processing unit 112. Let me enter.

Next, in step S502, the camera CPU 115a supplies the polarity determination unit 113 with the angular deflection signal and the motion vector signal in a predetermined frequency band extracted from the angular deflection signal from the vibration detection unit 18 and the motion vector signal from the vector processing unit 15. It is made to judge whether the motion vector signal is insufficient (insufficient vibration correction), excessive (excessive vibration correction), or no excess or deficiency (appropriate vibration correction) with respect to the remaining vibration correction.

Next, in step S503, the camera CPU 115a determines whether or not the determination result by the polarity determination unit 113 in step S502 is insufficient shake correction, and if the determination result is insufficient, the process proceeds to step S504. If not, step S505 is performed. Proceed to.

In step S504, the camera CPU 115a causes the comparison processing unit 112 to output the multiplication coefficient to the signal processing unit 111 of the interchangeable lens 11b. The comparison processing unit 112 outputs a multiplication coefficient larger than 1 with respect to the determination result of insufficient runout correction by the polarity determination unit 113. At this time, as described above, the larger the amount of runout correction shortage, the larger the multiplication coefficient is output. For example, if the runout correction deficiency amount is the reference amount, 1.2 is output as the multiplication coefficient, and if the runout correction deficiency amount is twice the reference amount, 1.4 is output as the multiplication coefficient, and the runout correction deficiency is insufficient. If the quantity is half of the reference quantity, 1.1 is output as the multiplication coefficient. Then, the process proceeds to step S508.

In step S505, the camera CPU 115a determines whether or not the determination result by the polarity determination unit 113 in step S502 is over-correction, and if it is over-correction, the process proceeds to step S506. If not, the process proceeds to step S507. ..

In step S506, the camera CPU 115a causes the comparison processing unit 112 to output the multiplication coefficient to the signal processing unit 111. The comparison processing unit 112 outputs a multiplication coefficient smaller than 1 with respect to the determination result of the runout correction by the polarity determination unit 113. At this time, as described above, the larger the runout overcorrection amount, the smaller the multiplication coefficient is output. For example, if the runout correction amount is the reference amount, 0.8 is output as the multiplication coefficient, and if the runout correction amount is twice the reference amount, 0.6 is output as the multiplication coefficient to correct the runout. If the quantity is half of the reference quantity, 0.9 is output as the multiplication coefficient. Then, the process proceeds to step S508.

In step S507, the camera CPU 115a causes the comparison processing unit 112 to output the multiplication coefficient to the signal processing unit 111. The comparison processing unit 112 outputs 1 as a multiplication coefficient with respect to the determination result of the proper runout correction by the polarity determination unit 113.

In step S508, the camera CPU 115a causes the signal processing unit 111 to change the gain with respect to the angular deflection signal from the vibration detection unit 18 according to the multiplication coefficient input from the comparison processing unit 112 via the lens CPU 115b.

Next, in step S509, the camera CPU 115a determines whether or not a predetermined time (for example, 0.2 seconds) has elapsed from the gain change in step S508, and when the predetermined time elapses, the process proceeds to step S510.

In step S510, the camera CPU 115a determines whether or not the release button has been fully pressed (S2-ON), and if it is S2-ON, an imaging operation for acquiring a still image is performed. The anti-vibration control process is terminated. On the other hand, if it is not S2-ON, the process returns to step S501. Waiting for a predetermined time in step S509 is a loop oscillation in which the motion vector is detected after the gain is changed and the gain is changed again by making the waiting time longer than the processing time required by the rectifying processing unit 17. This is to prevent.

As described above, the lens vibration isolation unit 110 can perform an appropriate vibration isolation operation by correcting the angular vibration signal from the vibration detection unit 18 by using the motion vector signal obtained by the vector processing unit 15. It is possible to satisfactorily correct not only the angular runout but also the parallel runout that occurs in close-up imaging such as macro imaging.

The noise component included in the angular deflection signal from the vibration detection unit 18 is increased by changing the gain by the signal processing unit 111. In particular, since the low-frequency fluctuation signal generated when the vibration detection unit 18 is activated has a large amplitude, the operation of the lens vibration isolation unit 110 driven in response to the angular vibration signal from the vibration detection unit 18 causes deterioration of the vibration isolation performance. It also becomes. Therefore, in this embodiment, the following fluctuation signal countermeasures are taken.

Due to the vibration isolation operation of the lens vibration isolation unit 110 driven by the signal processing unit 111 in response to the angular vibration signal whose gain has been changed, the image sensor 12a receives image deflection in a predetermined frequency band extracted by the narrow band filter 16. It is gradually attenuated. Then, the noise component generated by the vibration detection unit 18 accounts for a large proportion of the motion vector signal (remaining vibration correction) obtained by the vector processing unit 15.

Since the output of the vector processing unit 15 is input to the camera vibration isolation unit 114 that shifts the image sensor 12a in the shift direction, the image sensor 12a is shift-driven according to the motion vector signal in which the noise component occupies most, and the image sensor 12a Reduce the remaining runout correction above. In this way, the camera vibration isolation unit 114 corrects the remaining vibration correction that could not be completely corrected by the lens vibration isolation unit 110.

By changing the vibration isolating operation of the camera vibration isolating unit 114 and the gain for the angular vibration signal from the vibration detecting unit 18 described above, the image vibration is satisfactorily corrected without being greatly affected by the noise contained in the angular vibration signal. can do.

Even if the subject has low-frequency movement (for example, movement in a certain direction), the movement is corrected by the camera anti-vibration unit 114 by the motion vector detected by the vector processing unit 15, and is periodic. The parallel runout is corrected by the lens vibration isolator 110.

As described above, the predetermined frequency (hereinafter referred to as the extraction frequency) extracted by the narrow band filter 17 is variable, and in this embodiment, the extraction frequency is changed according to the exposure time at the time of imaging. Specifically, when the exposure time is short, the extraction frequency is set to a higher frequency than when the exposure time is long. This is because when the exposure time is short, the image shake due to the high frequency camera shake becomes dominant, so that the extraction frequency is set to a high frequency to improve the correction accuracy of the high frequency image shake.

FIG. 6 shows the relationship between the exposure time and the extraction frequency. The horizontal axis indicates the exposure time, for example, the left end indicates 1 second, the right end indicates 1/60 second, and the right end indicates a shorter exposure time. The vertical axis indicates the extraction frequency, the lower end indicates 1 Hz, the upper end indicates 7 Hz, and the upper side indicates a higher extraction frequency. The graph line 61 shows that the exposure time and the extraction frequency are inversely proportional (the extraction frequency is doubled when the exposure time is halved). In this way, by changing the extraction frequency according to the exposure time (imaging conditions), stable image shake correction can always be performed.

Further, various types of interchangeable lenses 11b having different configurations of imaging optical systems, focal lengths, and the like can be attached to the camera body 11a. If the interchangeable lens 11b has a vibration detection unit 18 and a signal processing unit 111 having a gain changing function, information for gain change based on the motion vector signal obtained by the camera body 11a (multiplication from the comparison processing unit 112). The coefficient) is transmitted to the signal processing unit 111, and the gain for the angular deflection signal from the vibration detection unit 18 is changed. As a result, even when various interchangeable lenses 11b that can be attached to the camera body 11a have a lens vibration isolator 110 having different drive characteristics and a vibration detection unit 18 having different detection characteristics, the vector processing unit 15 has them. Since the motion vector indicating the remaining vibration correction under the characteristics is detected, stable vibration isolation performance can be obtained.

In this embodiment, the case where the vibration isolation control process is performed during the image pickup preparation operation and the vibration isolation control process is not performed during the image pickup operation has been described, but from the image sensor 12a during the image pickup operation (exposure period). If non-destructive reading or divided reading of the image pickup signal can be performed, the motion vector may be detected and vibration isolation control processing may be performed.

FIG. 7 shows the configuration of the camera system 11 composed of the camera body 11a according to the second embodiment of the present invention and the interchangeable lens 11b detachably attached to the camera body 11a. The camera CPU 115a controls the operation of each part of the camera body 11a'shown in the broken line frame 116a'in the drawing. Further, the camera CPU 115a communicates with the lens CPU 115b provided on the interchangeable lens 11b to exchange various commands and data, and also operates each part in the interchangeable lens 11b shown in the broken line frame 116b via the lens CPU 115b. To control. In this embodiment, the components common to those in the first embodiment are designated by the same reference numerals as those in the first embodiment. This example differs from Example 1 in the following points.

(1) The rectifying processing unit 17, the polarity determination unit 113, and the comparison processing unit 112 shown in the first embodiment are not provided, and the rectifying processing unit (first rectifying means) 73 and the rectifying processing unit (second rectifying means) are not provided. ) 74 and a comparison processing unit (processing means) 75.
(2) A narrow band filter (second filter) 71 having the same characteristics as the narrow band filter (first filter) 16 is added to the camera body 11a.
(3) A point having an adding unit (adding means) 72 that adds the signal from the narrow band filter 16 and the signal from the narrow band filter 71.
(4) The signals from the rectifying processing units 73 and 74 and the rectifying processing units 73 and 74 that rectify the signal from the adding unit 72 and the signal from the narrow band filter 71 are compared, and the multiplication coefficient is output to the signal processing unit 111. A point that newly has a comparison processing unit 75 to be used.

The angular deflection signal from the vibration detection unit 18 provided on the interchangeable lens 11b is input to the narrow band filter 71 provided on the camera body 11a'. The narrow band filter 71 has the same characteristics as the narrow band filter 16. When the angular deflection signal from the vibration detection unit 18 is input to the narrow band filter 71 having the same characteristics as the narrow band filter 16, the angular deflection signal in the same frequency band as the motion vector signal is extracted.

The addition unit 72 adds the motion vector signal and the angular deflection signal extracted by the narrow band filters 16 and 71, respectively. The amplitude of this addition result is larger than the amplitude of the angular runout signal extracted by the narrowband filter 71 when the runout correction is insufficient for the remaining runout correction, and is extracted by the narrowband filter 71 when the runout correction is performed. It becomes smaller than the amplitude of the angle deflection signal. This can be seen from the fact that in FIG. 4, since the waveform 41 and the waveform 42 are in phase, the amplitude increases when they are added, and because the waveform 41 and the waveform 43 are in opposite phase, the amplitude decreases when they are added.

The output from the addition unit 72 is a motion vector signal indicating the remaining vibration correction and a vibration correction indicating the target shift position of the vibration isolation lens 19a by the lens vibration isolation unit 110 regardless of whether the vibration correction is insufficient or the vibration correction is excessive. Since it is the result of addition with the signal, it is equal to the angular runout signal output from the vibration detection unit 18 when the proper runout correction is performed.

The rectification processing unit 73 converts the amplitude of the addition result into a bias value. On the other hand, the rectification processing unit 74 converts the amplitude of the angular deflection signal from the vibration detection unit 18 extracted by the narrow band filter 71 into a bias value.

The comparison processing unit 75 compares the bias value from the rectification processing unit 73 with the bias value from the rectification processing unit 74 to obtain the multiplication coefficient to the signal processing unit 111. Specifically, the bias value from the rectification processing unit 73 is divided by the bias value from the rectification processing unit 74 to obtain the multiplication coefficient. When the runout correction is insufficient, the bias value from the rectification processing unit 73 is larger than the bias value from the rectification processing unit 74, so a multiplication coefficient larger than 1 is obtained. The larger the bias value from the rectification processing unit 73, the larger the multiplication. The coefficient is calculated. Further, in the case of runout correction, the bias value from the rectification processing unit 73 is smaller than the bias value from the rectification processing unit 74, so that the multiplication coefficient is smaller than 1, and the smaller the bias value from the rectification processing unit 73, the smaller the multiplication coefficient. Is required. The comparison processing unit 75 transmits the obtained multiplication coefficient to the signal processing unit 111 of the interchangeable lens 11b. The camera CPU 115a may transmit the multiplication coefficient from the comparison processing unit 112 to the signal processing unit 111 via the lens CPU 115b.

The signal processing unit 111 changes the gain for the angular deflection signal from the vibration detection unit 18 at predetermined time (for example, 0.2 seconds) according to the multiplication coefficient from the comparison processing unit 75.

As described above, in this embodiment, the excess or deficiency of the image shake correction is obtained by dividing the bias value from the rectification processing unit 73 by the bias value from the rectification processing unit 74 in the comparison processing unit 75, and multiplication according to the result. The difference from the first embodiment is that the coefficient is input to the signal processing unit 111 to change the gain with respect to the angular deflection signal. In this embodiment, the gain control means is configured by the camera CPU 115a, the narrow band filters 16, 71, the addition unit 72, the rectification processing units 73, 74, and the comparison processing unit 75.

FIG. 8 shows the vibration isolation control process mainly executed by the camera CPU 115a in this embodiment. The same steps as those shown in FIG. 5 in the first embodiment are designated by the same reference numerals as those in FIG.

When the user half-presses the release button (not shown) provided on the camera body 11a, in step S1001, the camera CPU 115a performs an imaging preparation operation such as AF or AE, and activates the vibration detection unit 18 via the lens CPU 115b. The lens anti-vibration unit 110 is started to perform anti-vibration operation. Further, the camera CPU 115a causes the vector processing unit 15 to generate a motion vector signal from the video signal generated by the image processing unit 13 using the image pickup signal output from the image sensor 12a, and causes the narrow band filter 16 to generate a motion vector signal in a predetermined frequency band. The motion vector signal is extracted and output to the addition unit 72. Further, the camera CPU 115a extracts the angular deflection signal of the predetermined frequency band from the angular deflection signal of the vibration detection unit 18 input to the narrow band filter 71 through the lens CPU 115b, and outputs the angular deflection signal to the addition unit 72. In addition, the camera CPU 115a causes the addition unit 72 to add the motion vector signal and the angular deflection signal from the narrow band filters 16 and 71, and the rectification processing unit 73 compares the bias value obtained from the addition signal by the addition unit 72 with the comparison processing unit 75. To output. Further, the camera CPU 115a causes the rectification processing unit 73 to output the bias value obtained from the angular deflection signal from the narrow band filter 71 to the comparison processing unit 75. Then, the camera CPU 115a causes the comparison processing unit 75 to divide the bias value input from the rectification processing unit 73 by the bias value input from the rectification processing unit 74, and transmits the multiplication coefficient according to the result to the signal processing unit 111. Let me.

Next, in step S1002, the camera CPU 115a determines whether or not the image sensor 12a is being exposed, and repeats this determination during the exposure. This is to avoid sudden fluctuations in the vibration correction due to the change in the gain for the vibration detection unit 18 during the exposure. That is, when the exposure is started, the gain for the vibration detection unit 18 is fixed to the gain changed immediately before that. When the exposure is completed, the process proceeds to step S508. The processing of steps S508 to S510 is the same as the processing of steps S508 to S510 described in the first embodiment, and the camera CPU 115a returns to step S1001 if it is not determined to be S2-ON in step S510.

As described above, by changing the gain for the angular deflection signal from the vibration detection unit 18 by using the motion vector signal obtained by the vector processing unit 15, the influence of the noise contained in the angular deflection signal is greatly affected. Therefore, it is possible to cause the lens vibration isolation unit 110 to perform an appropriate vibration isolation operation, and it is possible to satisfactorily correct not only the angular vibration but also the parallel vibration generated in close-up imaging such as macro imaging.

Further, in the first embodiment, the extraction frequency of the narrow band filters 16 is changed according to the exposure time, but the extraction frequencies of the narrow band filters 16 and 71 are changed according to the focal length of the imaging optical system 19 in addition to the exposure time. May be good. Specifically, when the focal length is long, the extraction frequency may be changed to the low frequency side as compared with when the focal length is short. Since the amount of image shake becomes large when the focal length is long, the lens vibration isolator 110 having a large amount of image shake that can correct the image shake of a low frequency having a large amplitude is in charge of the correction, and is corrected by the lens vibration isolator 110. This is because the camera vibration isolator 114, which has a small amount of possible image shake, does not have to correct low-frequency image shake.

FIG. 9 shows the relationship between the focal length of the imaging optical system 19 and the extraction frequency. The horizontal axis indicates the focal length, for example, the left end indicates a focal length of 24 mm, the right end indicates a focal length of 168 mm, and the right end indicates a longer focal length. The vertical axis indicates the extraction frequency. For example, the lower end indicates 1 Hz, the upper end indicates 7 Hz, and the upper side indicates a higher extraction frequency. Graph line 81 shows that the focal length and the extraction frequency are inversely proportional (the extraction frequency is halved when the focal length is doubled). By changing the extraction frequency according to the focal length (imaging condition) of the imaging optical system 19 in this way, stable image shake correction can always be performed.

When the target shift position of one of the anti-vibration lens 19a driven by the lens anti-vibration unit 110 or the image sensor 12a driven by the camera anti-vibration unit 114 approaches the end of the movable range, the other target. The extraction frequencies of the narrow band filters 16 and 71 may be changed so as to change the shift position to a position farther from the optical axis 10 (that is, to increase the shift amount). For example, as shown by the broken line waveform 91 in FIG. 10, when the target shift position of the anti-vibration lens 19a reaches the position 92 near the end of the movable range, the extraction frequencies of the narrow band filters 16 and 71 are set to higher frequencies. As shown by the solid line waveform 91', the shift amount of the anti-vibration lens 19a is suppressed so as not to increase further, and the shift amount of the image sensor 12a is increased so as to compensate for the image shake correction for the remaining shift amount. .. Similarly, when the target shift position of the image sensor 12a approaches the end of the movable range, the extraction frequency of the narrow band filters 16 and 71 is changed to a lower frequency, and the shift amount of the image sensor 12a does not increase any more. The shift amount of the anti-vibration lens 19a is increased.
(Other Examples)
The present invention supplies a program that realizes one or more functions of the above-described embodiment to a system or device via a network or storage medium, and one or more processors in the computer of the system or device reads and executes the program. It can also be realized by the processing to be performed. It can also be realized by a circuit (for example, ASIC) that realizes one or more functions.

Each of the above-described examples is only a representative example, and various modifications and changes can be made to each of the examples in carrying out the present invention.

11a Camera body 11b Interchangeable lens 12a Image sensor 15 Vector processing unit 16 Narrow band filter 17 Rectification processing unit 18 Vibration detection unit 19a Anti-vibration lens 110 Lens anti-vibration unit 111 Signal processing unit 112 Comparison processing unit 113 Polarity determination unit 114 Camera anti-vibration Department

Claims (12)

An imaging device with removable interchangeable lenses
An image sensor that captures the subject image and
A motion vector detecting means for detecting a motion vector using the output from the image sensor, and
A camera anti-vibration means that performs an anti-vibration operation by moving the image sensor so as to reduce image vibration in response to a first signal indicating the motion vector.
A communication means for receiving information indicating whether or not the interchangeable lens is in the vibration isolation operation from the interchangeable lens.
It has a gain control means for acquiring a gain in response to the first signal detected when the interchangeable lens is in a vibration isolation operation.
The communication means is an image pickup apparatus characterized in that the gain acquired by the gain control means is transmitted to the interchangeable lens. The communication means receives a second signal based on the result of vibration detection from the interchangeable lens, and the gain control means
A first filter that extracts a signal in a predetermined frequency band from the first signal, and
A rectifying means for generating a third signal obtained by smoothing the signal extracted by the first filter, and
A determination means for determining excess or deficiency of the vibration isolation operation of the lens vibration isolation means using the first signal and the second signal, and
The imaging device according to claim 1, further comprising a processing means for performing an output for changing the gain in response to the excess / deficiency determination result and the third signal. The communication means receives a second signal based on the result of vibration detection from the interchangeable lens, and the gain control means
A first filter that extracts a signal in a predetermined frequency band from the first signal, and
A second filter that extracts a signal in the predetermined frequency band from the second signal, and
An addition means for adding the signals extracted by each of the first and second filters, and
A first rectifying means that smoothes the added signal from the adding means to generate a third signal, and
A second rectifying means that smoothes the signal extracted by the second filter to generate a fourth signal, and
The imaging apparatus according to claim 1, further comprising a processing means for performing an output for changing the gain in response to the third signal and the fourth signal. The imaging device according to claim 2 or 3, wherein the camera anti-vibration means performs an anti-vibration operation in response to the first signal that is not input to the first filter. The image pickup apparatus according to any one of claims 2 to 4, wherein the gain control means changes the predetermined frequency band according to the exposure time of the image pickup device. The imaging device according to any one of claims 2 to 5, wherein the gain control means changes the predetermined frequency band according to the focal length of the interchangeable lens. The method according to any one of claims 2 to 6, wherein the gain control means changes the predetermined frequency band according to the position of one of the vibration isolation lens and the image pickup element in the vibration isolation operation. The imaging device described. The imaging device according to any one of claims 1 to 7, wherein the gain control means outputs a coefficient multiplied by a reference value of the gain in the interchangeable lens to the interchangeable lens. The imaging device according to any one of claims 1 to 8, wherein the gain control means causes the interchangeable lens to update the gain at predetermined time intervals. An interchangeable lens that can be attached to and detached from the imaging device according to any one of claims 1 to 9.
Vibration detection means to detect runout and
A lens anti-vibration means that performs an anti-vibration operation to move the anti-vibration lens in response to a fourth signal generated by using the output from the vibration detection means.
An interchangeable lens comprising a gain changing means for changing a gain with respect to the fourth signal in the vibration isolating operation of the lens vibration isolating means according to an output from the imaging device. An image pickup device in which an interchangeable lens can be attached and detached, an image pickup element for capturing a subject image, a motion vector detection means for detecting a motion vector using an output from the image pickup element, and a first signal indicating the motion vector. A control method for an image pickup device having a camera vibration isolation means for performing a vibration isolation operation for moving the image pickup element so as to reduce image shake according to the above.
A step of receiving information from the interchangeable lens indicating whether or not the interchangeable lens is in the vibration isolation operation, and
A step of acquiring a gain in response to the first signal detected when the interchangeable lens is in the vibration isolation operation.
A method for controlling an imaging device, which comprises a step of transmitting the acquired gain to the interchangeable lens. An image pickup device in which an interchangeable lens can be attached and detached, an image pickup element for capturing a subject image, a motion vector detection means for detecting a motion vector using an output from the image pickup element, and a first signal indicating the motion vector. A computer of an image pickup apparatus having a camera anti-vibration means for performing an anti-vibration operation for moving the image sensor so as to reduce image shake according to the above is caused to execute a process according to the control method according to claim 11. Computer program.