JP2006285093A - Image blur correction signal processing circuit and optical observation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain an image blur correction signal processing circuit excluding an aperiodic signal unrelated to the vibration of an optical observation device from the object of image blur correction. <P>SOLUTION: An MPU 4 decides rocking frequency being the object of image blur correction by applying autocorrelation processing to an angular velocity signal inputted from a shake sensor 8. Since the correlation value of the aperiodic signal unrelated to the vibration of a telescope such as a low-frequency drift component included in the angular velocity signal outputted from the shake sensor 8, the angular velocity signal generated by panning of the telescope and noise superposed on the angular velocity signal is calculated to be low, the aperiodic signal is easily excluded from the object of image blur correction. When a plurality of peak intervals of autocorrelation values are detected, the MPU 4 prioritizes the value having the narrow peak interval (that is, high frequency area) and sets it as the object of image blur correction. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to image blur correction technology such as a telescope.

  2. Description of the Related Art An image blur correction apparatus that detects image blur caused by a shake by detecting the shake of a telescope or the like and shifting a lens based on the detected shake information is known (see Patent Document 1). According to Patent Document 1, a frequency to be subjected to image blur correction is determined according to a value of a swing signal (detection signal from an angular velocity sensor) that has passed through a high-pass filter. By varying the cut-off frequency of the band-limited high-pass filter in accordance with the determined frequency, a frequency component signal for performing image blur correction is extracted from the fluctuation signal. The target value (correction amount) for image blur correction is determined according to the signal value of the extracted frequency component.

JP 2000-39640 A

  In general, an oscillation (vibration) signal that causes image blur is represented by a periodic signal having a resonance frequency component that is unique to the source of vibration or the medium that transmits the vibration. In the above-described method using a high-pass filter, if the frequency of non-periodic noise generated from a factor unrelated to vibration is included in the pass band of the high-pass filter, a factor that erroneously determines the target frequency for image blur correction. Become.

  The present invention relates to an image blur correction signal processing circuit that calculates an image blur correction signal in order to correct blur of an observation image by an optical observation device based on vibration of the optical observation device. Autocorrelation calculation means for performing the above and frequency determination means for determining the frequency of the image blur correction signal according to the calculation result of the autocorrelation calculation means.

  According to the present invention, when correcting blurring of an observation image by an optical observation device based on vibration of the optical observation device, the frequency of image blur correction can be appropriately determined.

  The best mode for carrying out the present invention will be described below with reference to the drawings. FIG. 1 is a perspective view of a telescope provided with an image blur correction apparatus according to an embodiment of the present invention. In FIG. 1, the telescope includes a telescope body 10 and a lens hood 30. An eyepiece lens 12 is disposed at one end of the telescope body 10, and an objective lens 20 is disposed at the other end of the telescope body 10.

  A focus ring 14 configured to be rotatable is provided at a substantially central portion of the telescope body 10. When the focus ring 14 is rotated, the telescope is focused. A pedestal 13 for attaching the telescope to a tripod (not shown) is provided at the lower part of the telescope body 10.

  The lens hood 30 is attached to the objective lens 20 side of the telescope body 10. The lens hood 30 has a hood function that protects the objective lens 20 and prevents irregular reflection during backlighting and adhesion of raindrops and the like to the objective lens 20.

  FIG. 2 is a diagram for explaining the internal structure of the telescope main body 10, and is a front view of the image blur correction device viewed from the optical axis direction of the telescope. 3 is a cross-sectional view taken along the line AA ′ in FIG. 2, and FIG. 4 is a cross-sectional view taken along the line AOB in FIG.

  In FIG. 2, the image blur correction apparatus includes a shake correction optical member (21, 22, 23) and a drive member (24, 25, 26, 27, 28, 28) that drives the shake correction optical member (21, 22, 23). 29, 40, 41). The image blur correction device corrects image blur due to shaking by shifting (vibrating) the shake correction optical member (21, 22, 23) so as to eliminate the shake caused by the swing of the telescope.

  The shake correction optical member includes a shake correction lens 21, a lens chamber 22, and a lens frame 23. The shake correction lens 21 is held by a lens chamber 22. The lens chamber 22 is held by a lens frame 23.

  The driving member includes a coil 24, a coil 25, a magnet 26, a magnet 27, a yoke 28, a yoke 29, a yoke 40, and a yoke 41. The coil 24 and the coil 25 are attached to the lens frame 23. A predetermined gap is provided between the yoke 28 and the magnet 26 and the yoke 40, and the yoke 28 and the magnet 26 are opposed to the optical axis direction (direction perpendicular to the paper surface in FIG. 2) so as to sandwich the coil 24 in this gap. Arranged. In addition, a predetermined gap is provided between the yoke 29 and the magnet 27 and the yoke 41, and the yoke 29 and the magnet 27 are arranged to face each other in the optical axis direction so that the coil 25 is sandwiched between the gaps.

  When a current is supplied to the coil 24 and the coil 25, an electromagnetic force is generated by the respective magnetic field and current. The electromagnetic force generated by the coil 24, the magnet 26, the yoke 28, and the yoke 40 is generated in the Y direction orthogonal to the optical axis in FIG. 2, and drives the lens frame 23 (that is, the shake correction lens 21) in the Y direction. Further, the electromagnetic force generated by the coil 25, the magnet 27, the yoke 29, and the yoke 41 is generated in the X direction orthogonal to the optical axis in FIG. 2, and drives the lens frame 23 (the shake correction lens 21) in the X direction.

  The lens frame 23 is attached to the casing of the telescope by a plurality of elastic bodies (four wires in the example of FIG. 2) 36-39. The elastic bodies 36 to 39 are mounted substantially parallel to the optical axis AX (FIGS. 3 and 4), and the lengths of the elastic bodies 36 to 39 are substantially equal to each other. Even when the lens 23 (the shake correction lens 21) is driven in a direction orthogonal to the optical axis AX, the lens frame 23 (the shake correction lens 21) is configured not to be inclined with respect to the optical axis AX.

  A position sensor 34b configured by a photo interrupter or the like detects the driving amount of the lens frame 23 in the Y direction. Specifically, a photo interrupter 34b is disposed so as to sandwich a slit 33 (FIG. 3) formed in a lens position detection unit 31 that is a part of the lens frame 23, and a detection signal from the photo interrupter 34b is transmitted to the lens frame. 23 is used as a signal indicating the position in the Y direction.

  Similarly, the position sensor 34a configured by a photo interrupter or the like detects the driving amount of the lens frame 23 in the X direction. Specifically, a photo interrupter 34a is disposed so as to sandwich a slit (not shown) formed in the lens position detection unit 30 that is a part of the lens frame 23, and a detection signal from the photo interrupter 34a is transmitted to the lens frame 23. Is used as a signal indicating the position in the X direction.

  The hole 42 penetrating the lens frame 23, the solenoid shaft 43 and the solenoid 44 constitute a lock device. In FIG. 4, the solenoid 44 is configured to drive the solenoid shaft 43 in the optical axis AX direction (left-right direction). When the solenoid 44 drives the solenoid shaft 43 in the left direction, the tapered portion at the tip of the solenoid shaft 43 is fitted into the hole 42. Since the solenoid 44 is fixed to the casing of the telescope, the lens frame 23 (that is, the shake correction optical member) into which the solenoid shaft 43 is fitted is set in a fixed state (locked state).

  When the solenoid 44 drives the solenoid shaft 43 in the right direction, the tapered portion at the tip of the solenoid shaft 43 is separated from the hole portion 42. As a result, the lens frame 23 (that is, the shake correction optical member) can be driven (unlocked) in directions orthogonal to the optical axis AX (X direction and Y direction in FIG. 2).

  In the unlocked state, the lens frame 23 (shake correction optical member) of the telescope is driven in the X direction and the Y direction so as to eliminate image blur due to the swing of the telescope.

  FIG. 5 is a block diagram illustrating a circuit configuration of the image blur correction apparatus. The image blur correction apparatus includes a position sensor 1, a shake correction lens 2, a lock device 3, an MPU 4, a drive circuit 5, an operation switch (SW) 6, and a shake sensor 8. When the operation switch 6 is turned on, power from a power source 9 such as a battery is supplied to each block, and the MPU 4 starts a predetermined program.

  The position sensor 1 corresponds to the position sensors 34a and 34b in FIG. The shake correction lens 2 corresponds to the shake correction optical member (21, 22, 23) of FIG. The locking device 3 corresponds to the hole 42, the solenoid shaft 43, and the solenoid 44 in FIG.

  The shake sensor 8 is configured by an angular velocity sensor such as a piezoelectric gyro, and detects the swing of the telescope (movement in the X direction and the Y direction orthogonal to the optical axis). A detection signal (angular velocity signal) from the shake sensor 8 is input to the MPU 4. The MPU 4 A / D converts the angular velocity signal input from the shake sensor 8, and both the X and Y directions of the shake correction optical member (21, 22, 23) according to the digitized angular velocity signals in the X direction and the Y direction. Are respectively calculated. The drive circuit 5 supplies a current for driving the shake correction optical member (21, 22, 23) in the Y direction to the coil 24 based on the calculation result of the MPU 4, and the shake correction optical member (21, 22, 23). Is supplied to the coil 25 for driving in the X direction.

  The present invention is characterized by a method for determining the oscillation frequency to be subjected to image blur correction.

  FIG. 6 is a block diagram illustrating a method for determining the oscillation frequency according to the present embodiment. In FIG. 6, the A / D conversion unit 4A, the low-pass filter 4B, the high-pass filter 4C, the autocorrelation processing unit 4D, the selector 4E, the determination processing unit 4F, and the target value calculation unit 4G are configured by the MPU 4.

  The A / D converter 4A converts the angular velocity signal input from the shake sensor 8 into digital data at predetermined intervals. The angular velocity signal digitized by the A / D conversion unit 4A is sent to the low-pass filter 4B, the high-pass filter 4C, and the autocorrelation processing unit 4D. For example, the low-pass filter 4B passes components Sig, B of 10 Hz or less out of the input angular velocity signal and outputs the components to the selector 4E. For example, the high-pass filter 4C passes the component Sig.C exceeding 10 Hz out of the input angular velocity signal and outputs the signal to the selector 4E.

  The autocorrelation processing unit 4D obtains an autocorrelation value for the input angular velocity signal. Details of the autocorrelation calculation will be described later. The determination processing unit 4F determines an oscillation frequency to be subjected to image blur correction according to the correlation value calculated by the autocorrelation processing unit 4D, the presence / absence of periodicity, and the frequency, and a filter corresponding to the determined frequency. A control signal is sent to the selector 4E to make a selection. The selector 4E selects an angular velocity signal (either Sig.b or Sig.C) corresponding to the control signal from the determination processing unit 4F, and sends it to the target value calculation unit 4G. The target value calculation unit 4G integrates the angular velocity signal input from the selector 4E, and calculates target drive amounts in the X direction and Y direction for the shake correction optical members (21, 22, 23), respectively. As a result, an image blur correction signal for vibrating the shake correction optical members (21, 22, 23) at the determined frequency is obtained. The target value calculation unit 4G further sends an image blur correction signal to the drive circuit 5.

The flow of the autocorrelation calculation process performed by the autocorrelation processing unit 4D will be described in detail with reference to the flowchart of FIG. The processing according to FIG. 7 is repeatedly executed in the unlocked state. In step S1 of FIG. 7, M pieces of sampling data are input to the autocorrelation processing unit 4D. Specifically, angular velocity data input from the A / D conversion unit 4A at predetermined intervals is sequentially stored in a memory (not shown), and when the number of stored data reaches M, M data is subjected to autocorrelation processing. Read to unit 4D. Here, the number M of sample data is a value calculated by the following equation (1).
M = 2 × 1 / F × 1 / t (1)
Here, F is the lowest value (Hz) of the assumed oscillation frequency, and t is the A / D conversion interval (= sampling interval (sec)) by the A / D converter 4A. In the present embodiment, it is assumed that the oscillation frequency is 2 to 20 Hz, and F = 2 (Hz).

In step S2, the autocorrelation processing unit 4D calculates a DC (direct current) component included in the sampling data (angular velocity data) x [n], and proceeds to step S3. The DC component rdc is calculated by the following equation (2).

In step S3, the autocorrelation processing unit 4D removes the DC component rdc calculated from the sampling data (angular velocity data) x [n], and proceeds to step S4. The angular velocity data x ′ [n] after removing rdc is expressed by the following equation (3). In actual processing, the DC component rdc can be removed by shifting the sampling data (angular velocity data) x [n] stored in the memory by the number of bits corresponding to the DC component rdc. it can.
x ′ [n] = x [n] −rdc (3)

In step S4, the autocorrelation processing unit 4D calculates an autocorrelation value r [m] according to the following equation (4) for the M pieces of angular velocity data x ′ [n] after removing rdc.

  The autocorrelation processing unit 4D further performs normalization processing so that r [0] = 100 (%), and proceeds to step S5. As a result of normalization processing, the value calculated when the correlation value includes many low-frequency components and noise components does not become a large value.

  In step S5, the autocorrelation processing unit 4D detects the peak of the calculated correlation value. The peak is the sign change point (positive → negative) of the differential value of the positive part of the correlation value. When a plurality of peaks are detected, a peak having a high correlation value (that is, a high periodicity) is preferentially detected. The autocorrelation processing unit 4D further obtains the cycle of the angular velocity data from the peak interval of the correlation value. When a plurality of peak intervals are detected, a cycle is obtained with priority given to those having a narrow peak interval (that is, a high frequency region).

  In step S6, the autocorrelation processing unit 4D uses the data indicating the correlation value, the presence / absence of periodicity (change in peak interval), and the period (frequency) as filter selection information. The autocorrelation processing unit 4D transmits the filter selection information to the determination processing unit 4F, and ends the process of FIG. The determination processing unit 4F that has received the filter selection information determines a swing frequency to be subjected to image blur correction as shown in FIG. 8, and outputs a control signal for selecting an angular velocity signal corresponding to the determined swing frequency. Send to selector 4E.

  In FIG. 8, the determination processing unit 4F displays an image when the correlation value is 80% or more (step S50), the peak interval is periodic (step S51), and the frequency is higher than 10 (Hz) (step S52). The oscillation frequency to be subjected to blur correction is determined on the high frequency side. The determination processing unit 4F further sends a control signal to the selector 4E so as to select the angular velocity signal Sig.C that has passed through the high-pass filter (HPF) 4C (step S54). As a result, the target value calculation unit 4G calculates a target drive amount suitable for image blur correction when the telescope is attached to a tripod.

  When the correlation value is 80% or more (step S50), the peak interval is periodic (step S51), and the frequency is 10 (Hz) or less (step S53), the determination processing unit 4F performs image blur correction. The target oscillation frequency is determined on the low frequency side. In this case, the determination processing unit 4F sends a control signal to the selector 4E so as to select the angular velocity signal Sig.B that has passed through the low-pass filter (LPF) 4B (step S55). As a result, the target value calculation unit 4G calculates a target drive amount suitable for image blur correction when the telescope is held by an observer.

  Further, the determination processing unit 4F performs image blur correction when the correlation value is less than 80% (step S60), the frequency characteristics change transient state in which the peak interval varies (step S61), and a plurality of peak intervals exist. The target oscillation frequency is determined on the frequency side different from the previous determination (step S62). In this case, the determination processing unit 4F sends a control signal to the selector 4E so as to select an angular velocity signal that has passed through a high-pass filter 4C (or low-pass filter 4B) different from the previously selected low-pass filter 4B (or high-pass filter 4C) ( Step S63 or Step S64).

  If a plurality of peak intervals do not exist, the oscillation frequency that is subject to image blur correction is preferentially set to those with a narrow peak interval (that is, a high frequency range) regardless of the previously determined frequency. The determination processing unit 4F sends a control signal to the selector 4E so as to select an angular velocity signal corresponding to the oscillation frequency (step S63 or step S64).

  Furthermore, if the correlation value is less than 20% or less (step S70), the determination processing unit 4F has a dominant noise in the angular velocity data, or a failure has occurred in the shake sensor 8, or the telescope is panned. Assuming that the image is corrected (step S71), a signal for stopping the image blur correction operation is output (step 72). In this case, the MPU 4 sends a lock instruction to the shake correction optical member and stops the shake correction operation.

The embodiment described above has the following effects.
(1) An auto-correlation process is performed on the angular velocity signal input from the shake sensor 8 to appropriately determine the oscillation frequency to be subjected to image blur correction. A non-periodic signal that is unrelated to the vibration of the telescope, such as a low-frequency drift component included in the angular velocity signal output from the shake sensor 8, an angular velocity signal due to panning of the telescope, and noise superimposed on the angular velocity signal, has a low correlation value. Therefore, it can be easily distinguished from a signal having a frequency that requires image blur correction. On the other hand, the periodicity signal included in the angular velocity signal (that is, a signal indicating fluctuation (vibration) that causes image blur) is calculated with a high correlation value, so it is reliably detected as a signal having a frequency that requires image blur correction. can do. Further, unlike a method using a filter as in the prior art, erroneous determination due to the relationship between the frequency of an aperiodic signal and the passband of the filter is not caused. Furthermore, the autocorrelation process has a smaller amount of computation than the fast Fourier transform process (FFT), and can reduce the burden on the MPU 4.

(2) When a plurality of autocorrelation value peak intervals are detected, the one with a narrow peak interval (that is, a high frequency region) is preferentially set as a target for image blur correction. By making the high frequency range that is difficult to follow with the observer's eyes, the image that is observed after image blur correction is better compared to the case where the low frequency range is prioritized for image blur correction. Can do well.

(3) When the correlation value is 80% or higher (step S50), the peak interval is periodic (step S51), and the frequency is higher than 10 (Hz) (step S52), the image blur correction target Since the dynamic frequency is determined on the high frequency side and the target driving amount based on the angular velocity signal Sig.C that has passed through the high-pass filter (HPF) 4C is calculated by the target value calculation unit 4G (step S54), the telescope is tripod. It is possible to perform image blur correction suitable for the case where it is attached to the camera.

(4) When the correlation value is 80% or more (step S50), the peak interval is periodic (step S51), and the frequency is 10 (Hz) or less (step S53), the image is subjected to image blur correction. Since the oscillation frequency is determined on the low frequency side and the target driving amount based on the angular velocity signal Sig.B that has passed through the low-pass filter (LPF) 4B is calculated by the target value calculation unit 4G (step S55), the telescope Image blur correction suitable for a hand held by an observer can be performed.

(5) When the correlation value is less than 80% (step S60), the frequency characteristics change transient state in which the peak interval varies (step S61), and there are a plurality of peak intervals, the image blur correction target The target frequency is determined based on the angular velocity signal that has passed through the high-pass filter 4C (or low-pass filter 4B) that is different from the previously selected low-pass filter 4B (or high-pass filter 4C). The calculation unit 4G is operated (step S62). Thereby, when the period of the periodic signal is changing, image blur correction can be performed in correspondence with the periodic signal included in the sampling data that is temporally new.

(6) When the correlation value is less than 20% or less (step S70), the angular velocity data is considered to be dominant in noise, a failure has occurred in the shake sensor 8 or the like, or the telescope has been panned (step S71). Since a signal for stopping the image blur correction operation is output (step 72), useless image blur correction processing can be stopped.

  In the above description, one of two types of filters (that is, two types of angular velocity signals) is alternatively selected according to the oscillation frequency determined by the determination processing unit 4F, but there are two types of filters. However, the present invention is not limited to this, and may be configured to include three or more types of filters. In this case, a low pass filter (LPF), a band pass filter (BPF), and a high pass filter (HPF) are provided so that the pass bands of the respective filters are different. In the case of using four or more types of filters, it is preferable to provide a plurality of bandpass filters so that the passbands of the bandpass filters do not overlap. The determination processing unit 4F alternatively selects one filter (that is, one angular velocity signal) so that an angular velocity signal including the determined oscillation frequency is input to the target value calculation unit 4G.

  The band pass filter may be selected in the following cases. If the correlation value is 50% or less, the determination processing unit 4F assumes that high-frequency noise or low-frequency noise is superimposed on the angular velocity data, and selects a bandpass filter that cuts both the high-frequency component and the low-frequency component. Then, a control signal is sent to the selector 4E. As a result, the target drive amount based on the angular velocity signal that has passed through the bandpass filter can be calculated by the target value calculation unit 4G, so that image blur correction can be performed while suppressing the influence of noise.

  The correspondence between each component in the claims and each component in the best mode for carrying out the invention will be described. The optical observation apparatus includes, for example, telescopes 10 and 30 including monoculars and binoculars. The image blur correction signal corresponds to a signal indicating the target drive amount. The image blur correction signal processing circuit, the autocorrelation calculation means, and the frequency determination means are constituted by, for example, the MPU 4. The signal indicating vibration corresponds to, for example, an angular velocity signal. The vibration detecting means is constituted by, for example, a vibration sensor 8. The image blur correction optical system includes, for example, shake correction optical members 21, 22, and 23. A drive means is comprised by the drive members 24, 25, 26, 27, 28, 29, 40, 41, for example. In addition, the above description is an example to the last, and when interpreting invention, it is not limited to the correspondence of the component of said embodiment and the component of this invention at all.

1 is a perspective view of a telescope provided with an image blur correction device according to an embodiment of the present invention. It is a front view explaining the internal structure of a telescope main body. It is sectional drawing explaining the internal structure of a telescope main body. It is sectional drawing explaining the internal structure of a telescope main body. It is a block diagram explaining the circuit structure of an image blur correction apparatus. It is a block diagram explaining the determination system of a rocking | fluctuation frequency. It is a flowchart explaining the flow of an autocorrelation calculation process. It is a figure explaining the relationship between a rocking | fluctuation frequency and a selection filter.

Explanation of symbols

2 (21, 22, 23) ... shake correction lens (shake correction optical member)
4 ... MPU
4A ... A / D conversion unit 4B ... Low pass filter 4C ... High pass filter 4D ... Autocorrelation processing unit 4E ... Selector 4F ... Determination processing unit 4G ... Target value calculation unit 5 ... Drive circuit 8 ... Shake sensor 10 ... Telescope body 24, 25 , 26, 27, 28, 29, 40, 41 ... drive member

Claims (7)

In an image blur correction signal processing circuit that calculates an image blur correction signal for correcting blur of an observation image by an optical observation device based on vibration of the optical observation device,
Autocorrelation calculating means for performing autocorrelation processing on the signal indicating the vibration;
An image blur correction signal processing circuit, comprising: a frequency determination unit that determines a frequency of the image blur correction signal according to a calculation result of the autocorrelation calculation unit. The image blur correction signal processing circuit according to claim 1,
The image blur correction signal processing circuit, wherein the frequency determination means determines a frequency of the image blur correction signal according to a peak interval of an autocorrelation value. In the image blur correction signal processing circuit according to claim 2,
The image blur correction signal processing circuit, wherein the frequency determination unit obtains the peak interval from a signal having a high autocorrelation value. In the image blur correction signal processing circuit according to claim 2 or 3,
The frequency determination means determines the frequency of the image blur correction signal from a narrower peak interval when there are a plurality of types of peak intervals. In the image blur correction signal processing circuit according to claim 2 or 3,
The frequency determining means is characterized in that, when the autocorrelation value is lower than a first predetermined value and there are a plurality of the peak intervals, the frequency of the image blur correction signal is different from the previously determined frequency. An image blur correction signal processing circuit. In the image blur correction signal processing circuit according to claim 5,
An image blur correction signal processing circuit that stops frequency determination of the image blur correction signal when the autocorrelation value is lower than a second predetermined value lower than the first predetermined value. . An image blur correction signal processing circuit according to any one of claims 1 to 6,
Vibration detecting means for detecting vibration of the optical observation device and outputting a detection signal;
An image blur correction optical system for correcting blur of an observation image by the optical observation device;
An optical observation apparatus comprising: drive means for driving the image blur correction optical system in accordance with an image blur correction signal calculated by the image blur correction signal processing circuit.

Images