JP2000013670A - Image pickup device and controlling method for image pickup device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a device which uses a controlling method that is optimum to each operation state of a vibration proof photographic mode and a pixel shifting photographic mode by changing drive control of an image blurring correcting means according to a selected photographic mode between a photographic mode that corrects image blurring and a photographic mode that synthesizes a high resolution image. SOLUTION: A CPU 1 is in charge of control over the entire camera. A photographic mode setting means 2 consists of a switch which switches vibration proof photographic mode that is for eliminating image blurring that is caused by hand shake and a pixel shifting photographic mode which is for producing a high-definition image, etc. The sensitivity of a correction system position detecting part 19 is changed according to whether a photographic mode is a photographic mode that premises normal blurring correction or the pixel shifting photographic mode. And the control is carried out in methods which are optimum to the respective photographic modes in such manners that one makes it possible to perform photographing with a camera in hand by giving priority to stroke and the other drives a correcting lens to an accurate position by giving priority to control resolution.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image pickup apparatus having a high-definition image pickup function by shifting pixels and an image blur correction function.

[0002]

2. Description of the Related Art Conventionally, this type of digital still camera employs an optical axis deflecting means such as a variable apex angle prism attached to the front surface of a taking lens as disclosed in Japanese Patent Application Laid-Open No. 7-240932. There is a so-called pixel shifting method in which a subject image projected on an image sensor is shifted spatially and time-sequentially, and each shooting data is later synthesized to finally obtain a high-resolution image.

In this method, first, in the state where the variable apex angle prism is at a predetermined angle, the first photographing is performed, the subject image at that time is photographed by the image pickup device, and each pixel data is sequentially obtained. Each is read out, converted into digital data via an A / D converter, and stored in a memory.

On the other hand, during the reading, the next photographing is performed continuously. In this case, before the photographing is started (for example, during the V blanking period), the variable apex angle prism is moved with respect to the first photographing. By tilting by a predetermined amount, the image forming position of the subject image projected on the image sensor becomes different from that in the first shooting.

Therefore, if the predetermined amount of displacement of the variable apex angle prism is appropriately selected, the subject images on the image pickup device in the first photographing and the second photographing are, for example, 1 / pixel of each pixel interval of the image pickup device. You can create a state that is shifted by two. With such a method, the variable apex angle prism is displaced by a predetermined amount for each photographing, and the same photographing data at a plurality of different spatial positions is separately stored in the memory.

Normally, R (red), G (green), B
In the case of a three-panel system using a separate CCD for each color of (blue), the next approach is to use 1 in the X direction only for the first approach image.
The variable apex angle prism is driven so as to be shifted by 1/2 pixel, and then the variable apex angle prism is driven so as to be shifted by 1/2 pixel only in the Y direction. Finally, X opposite to the second drive direction is driven.
The variable apex angle prism is driven so that only the direction is shifted by 画素 pixel.

[0007] By synthesizing the four photographing data obtained in this way in post-processing, it is possible to obtain photographing data having twice the resolving power both horizontally and vertically with respect to the photographing data obtained from the original image sensor. It becomes possible.

On the other hand, as a so-called anti-shake device for preventing image shake due to camera shake of a photographer, in addition to the above-mentioned correction means using a variable apex angle prism, a shift correction means whose specific configuration is shown in FIG. Is also used.

The detailed operation of this shift correction means will be described later. This is to make it possible to move some lens groups of the optical system of the photographing optical means freely on a plane perpendicular to the optical axis. If this lens group is moved in the predetermined X and Y directions, the same effect as in the above-described pixel shift shooting using the variable apex angle prism can be obtained.

[0010]

However, there is a problem in the coexistence of the pixel shift function and the image blur correction function, and the correction system in the case where the camera shake prevention is actually executed and the case where the pixel shift is actually executed. No consideration is given to the difference in the control method, and nothing is disclosed.

In consideration of actual use, when the photographer selects the image stabilization shooting mode, a stroke that mainly compensates for the camera shake amount of the photographer in hand-held shooting is necessary. Control of the correction system is indispensable.

On the other hand, when the photographer selects the above-described pixel shift photographing mode, a plurality of exposures are required. Therefore, the photographer is premised on tripod photographing. It is indispensable to control the position at such a fine pitch.

SUMMARY OF THE INVENTION An object of the present invention is to provide an image pickup apparatus using a control method which is optimal for the respective operating states of an image stabilizing shooting mode and a pixel shift shooting mode. It is an object of the present invention to provide an image pickup apparatus in which the drive control itself of the correction means is changed between a first photographing mode in which the correction is performed and a second photographing mode in which the pixel shift photographing is performed.

Further, a second object of the present invention is to change the shake sensor processing itself between a first shooting mode for executing image stabilization shooting and a second shooting mode for executing pixel shift shooting. It is to provide an imaging device.

[0015]

According to the first aspect of the present invention, there is provided an image pickup means, a shake detecting means for detecting a shake, and an output from the shake detecting means. Image blur correction means for correcting image blur based on the image position, an image position on the image pickup means, a pixel shift means for minute displacement using the image shake correction means, and an image on the image pickup means by the pixel shift means Image synthesizing means for synthesizing a high-resolution image based on a plurality of image data imaged at different positions, a first imaging mode for correcting image blur, and synthesizing the high-resolution image And a control unit for changing the drive control of the image blur correcting unit according to the selected shooting mode.

According to the invention described in claim 2 of the present application, in the invention described in claim 1, the image blur correction means includes: a position detection means for detecting a current position of the correction unit; And at least two or more amplifying units having different amplification factors, and the control unit is configured to select an output of the amplifying unit according to the selected shooting mode. Features the device.

According to the invention described in claim 3 of the present application, in the invention described in claim 1, the image blur correction means includes a correction optical system, and the control means controls the image blur correction means. In the first photographing mode, control is performed to give priority to the movable range over the resolution of the correction optical system, and in the second photographing mode, the movable range of the correction optical system is narrowed to increase the resolution. The imaging apparatus is configured to perform control giving priority to.

According to the invention described in claim 4 of the present application, in the invention described in claim 1, the control means includes:
An image pickup apparatus includes a frequency characteristic changing unit that changes a frequency characteristic of the image blur correcting unit according to the shooting mode.

According to a fifth aspect of the present invention, in the fourth aspect of the present invention, the frequency characteristic changing unit is a correction target of the image blur correcting unit in the first photographing mode. A frequency characteristic is set so as to reduce a phase delay with respect to a shake frequency range, and in the second shooting mode, a frequency characteristic is set by giving priority to a responsiveness of the image shake correction unit when the image shift correction unit is minutely driven by the pixel shift unit. An imaging device configured to perform the above operation is characterized.

According to the invention described in claim 6 of the present application, in the invention described in claim 5, the frequency characteristic in the second photographing mode is changed by the pixel shift means so that the image shake correction means is a static friction. The imaging device is characterized in that it is configured to be determined on the basis of the responsiveness capable of minute driving to a target position against the target position.

According to the invention described in claim 7 of the present application, in the invention described in claim 1, the control means includes:
It is characterized by an imaging device configured to be arbitrarily switchable by an operator.

According to the invention described in claim 8 of the present application, an image pickup means, a shake detection means for detecting a shake, an image shake correction means for correcting an image shake based on an output of the shake detection means, A pixel shifting unit for minutely displacing the position of the image on the imaging unit using the image shake correcting unit, and a plurality of image data obtained by shifting the position of the image on the imaging unit by the pixel shifting unit. Means for synthesizing a high-resolution image, a first shooting mode for the purpose of correcting image blur, and a second shooting mode for the purpose of synthesizing a high-resolution image. And a control unit for changing a signal processing method from the shake detecting unit according to the selected photographing mode.

According to the ninth aspect of the present invention, in the eighth aspect of the present invention, the control means includes:
When the second shooting mode is selected, the imaging apparatus is configured to perform control so as not to perform shake detection.

According to a tenth aspect of the present invention, in the ninth aspect of the invention, when the second photographing mode is selected, the control means supplies a power to the shake detecting means. It is characterized by an imaging device configured to control so as not to supply.

According to the invention described in claim 11 of the present application, in the invention described in claim 10, when the second photographing mode is selected, power is not supplied to the shake detecting means. An image pickup apparatus is configured to output a target position of the image blur correction unit from the control unit.

According to a twelfth aspect of the present invention, in the eighth aspect of the invention, when the second photographing mode is selected, the control means outputs the output of the shake detecting means. An image pickup apparatus configured to perform control so as to prohibit the operation of the image blur correction unit based on the image blur correction means.

According to the thirteenth aspect of the present invention, in the eighth aspect, when the first photographing mode is selected, the control means responds to a power-on of the imaging device. An image pickup apparatus configured to start energization to the shake detecting unit and to drive and control the image shake correcting unit based on an output of the shake detecting unit.

According to the fourteenth aspect of the present invention, the image pickup means, the shake detection means for detecting shake, the image shake correction means for correcting image shake based on the output of the shake detection means, The image blur correction unit is configured to shift the position of the image on the imaging unit by a small amount using the image blur correction unit, and the image shift unit shifts the position of the image on the imaging unit by the pixel shift unit. A method for controlling an imaging apparatus, comprising: an image synthesizing unit that synthesizes a high-resolution image based on a plurality of pieces of image data, wherein the first imaging is performed to correct image blur by the image blur correcting unit. Mode and a second photographing mode for synthesizing a high-resolution image can be selected, and the drive control of the image blur correcting means is changed according to the selected photographing mode. Wherein the manufacturing method of an imaging device.

According to a fifteenth aspect of the present invention, in the first aspect of the invention, in the first photographing mode, the movable range has priority over the resolution of the correction optical system of the image blur correction means. In the second shooting mode, the movable range of the correction optical system is narrowed,
The present invention is characterized by a control method of an imaging device in which control is performed with priority given to increasing the resolution.

According to the invention described in claim 16 of the present application,
The invention according to claim 15, wherein in the first shooting mode, a frequency characteristic is set so as to reduce a phase delay with respect to a shake frequency range to be corrected by the image shake correcting means, and the second shooting mode is set. The present invention is characterized by a method of controlling an image pickup apparatus in which the frequency characteristic is set with priority given to the response at the time of minute driving of the image blur correcting means by the pixel shifting means.

According to the invention described in claim 17 of the present application,
An image pickup unit, a shake detection unit that detects a shake, an image shake correction unit that corrects an image shake based on an output of the shake detection unit, and a position of an image on the imaging unit, using the image shake correction unit An image synthesizing means for synthesizing a high-resolution image based on a plurality of image data obtained by displacing the position of the image on the imaging means by the pixel shifting means. A control method for an apparatus, wherein a first shooting mode for correcting image blur and a second shooting mode for synthesizing a high-resolution image can be selected and selected. A method of manufacturing an image pickup apparatus, wherein a method of processing a signal from the shake detecting means is changed depending on the taken photographing mode.

According to the invention described in claim 18 of the present application, in the invention described in claim 17, when the second photographing mode is selected, the image pickup is controlled so as not to perform shake detection. It is characterized by a control method of the device.

According to a nineteenth aspect of the present invention, in the seventeenth aspect, when the second photographing mode is selected, no power is supplied to the shake detecting means. An image pickup apparatus configured to output a target position of the image blur correction means from the control means.

[0034]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) FIG. 1 is a block diagram showing the overall hardware configuration of the present invention. In this figure, reference numeral 1 denotes a CPU as control means for controlling the entire camera. Numeral 2 denotes a photographing mode setting means for setting a photographing mode of the camera itself. For example, an anti-shake photographing mode for removing image shake caused by camera shake of a photographer himself when performing normal hand-held photographing, and a tripod It is possible to take multiple shots in a state where the camera is mounted on the camera, shift the imaging position of the subject image by an extremely small amount, such as the pixel pitch of the image sensor, for each shooting, and combine these multiple images later. It is composed of a switch and the like for switching a pixel shift photographing mode for producing a fine image.

In addition, reference numeral 18 denotes a camera operation switch for starting photographing of the camera, which represents a main switch for starting power supply to all circuit systems and a release switch for starting actual photographing. ing.

Reference numeral 3 denotes a main photographing optical system of the present camera. Reference numeral 4 denotes a so-called pixel which shifts a subject image formed on the image pickup means 6 spatially parallel as described later. It shows an optical means for performing a shift. Here, as this optical means, a so-called shift correction optical system as shown in FIG. 3, for example, is used.

The specific configuration of the shift optical system will be described with reference to FIG. In this figure, reference numeral 50 corresponds to the correction lens group 4 shown in FIG. 1, and this correction lens group 50 is moved in the X-axis direction in FIG. In the magnetic circuit unit constituted by 52, the operation can be freely performed by changing the amount and direction of current supplied to the winding coil 52. Similarly, with respect to the movement in the Y-axis direction in FIG. 3, in the magnetic circuit unit including the magnetic member 53 including the magnet and the yoke and the winding coil 54, the current amount and the current It is possible to operate freely by changing.

The actual movement of the correction lens group is represented by an IRED 56 (for X-direction detection) and an IRED 57 (for Y-direction detection) that move integrally with the lens group as a relative movement amount with respect to the lens barrel support frame 55. The combination of a PSD 58 (for X-direction detection) and a PSD 59 (for Y-direction detection) fixedly attached to the lens barrel support frame 55,
The detection is performed by an optically non-contact method.

In addition, reference numeral 60 denotes a mechanical lock mechanism for fixing the position of the lens group to a predetermined position in a state in which the drive to the correction optical system is stopped. By changing the position, the protrusion 61 at the tip of the lever of the mechanical lock mechanism is adjusted to
Locked state (state in which the correction lens group is mechanically fixed), unlocked state (state in which the correction lens group is free) depending on whether it jumps into or out of the dent that moves together with the zero.
To produce Incidentally, reference numeral 63 denotes a support sphere as an anti-tilt for regulating the inclination of the correction lens group 50 with respect to the optical axis.

The shift optical system itself is configured as described above, but the actual position of the shift optical system is detected by the correction system position detecting means 19 including the combination of the PSD and the IRED as described above.

The specific circuit configuration of the correction system position detecting means will be described with reference to FIG. Among them, the signal light from the IRED 71 that emits infrared light when a certain current flows is incident on the PSD 70 through the slit 72.

The two photocurrents Ia generated by the PSD 70
And Ib, the ratio of which varies depending on the position where the signal light from the IRED 71 is incident (actually, the position of the center of gravity of the projected image of the signal light on the PSD), and the sum (I
a + Ib) is proportional to the incident light level.

This current output Ia becomes a voltage output -Va through a current-voltage conversion circuit composed of an OP amplifier 73 and a resistor 74, and the other current output Ib is likewise obtained.
The voltage output becomes -Vb through the current-voltage conversion circuit including the OP amplifier 75 and the resistor 76.

Next, the two voltage outputs -Va and -Vb are
One is an OP amplifier 77 and resistors 78, 79, 80, 81
, And subtracts both outputs to generate an output Va-Vb. This output is, of course, the position where the signal light from the IRED 71 is incident on the PSD7.
0, it increases to the + side as it approaches the a side, and the incident position increases to one side as it approaches the b side of the PSD 70. Therefore, the IRED of the IRED that moves together with the movement of the shift lens group 50 as shown in FIG. Output the movement as it is.

The outputs -Va and -Vb are also input to an addition circuit comprising an OP amplifier 82 and resistors 83, 84 and 85, where the outputs are added to generate an output Va + Vb. This output is the sum of the reference voltage VC of each OP amplifier and the signal voltage due to the incidence of the signal light. This voltage is applied to the next-stage OP amplifier 86, transistor 87, resistors 88, 89, and 91. , A capacitor 90, and Va + V
An operation of automatically adjusting the IRED current is performed so that the output of b becomes equal to the reference voltage KVC (> VC). In this way, the sum of the signal outputs from the PSD is always constant regardless of the temperature and the individual differences in the IRED emission power.
If the ED current is adjusted, one Va-Vb output will always accurately represent the position of the shift lens group.

Subsequently, the Va-Vb output is input to an inverting amplifier circuit composed of an OP amplifier 92 and resistors 93 and 94 surrounded by a dotted line A, where the output is subjected to a predetermined amplification, and the output is converted to an A / D signal. Connect to the AN-A input of converter 98.

The Va-Vb output is input to an inverting amplifier circuit composed of an OP amplifier 95 and resistors 96 and 97 surrounded by a dotted line B, where a predetermined amplification is performed, and the output is output to A.
/ D converter 98 is connected to the AN-B input.

Here, the amplification factor of the amplifying portion in the portion surrounded by the dotted line B is set to be larger than the amplification factor of the amplifying portion in the portion surrounded by the dotted line A, and Voltage output is large.

With the above arrangement, the absolute position of the correction system is extracted. However, the output can be extracted in exactly the same manner for the movement in the Y direction, so that the description is omitted here.

Normally, this shift correction optical system is used for a camera shake preventing mechanism for the photographer himself / herself of the entire camera. In this case, a shake sensor 17 (normally called an angular velocity called a vibration gyro) detects the amount of shake of the entire camera. Using two sensors and separately detecting angular blur around two different axes (yaw, pitch)).

The configuration of the shake sensor and the processing circuit is specifically as shown in FIG. The output from the vibrator 40 for actually detecting the angular velocity is taken out via the synchronous detection circuit 41, and the output is resonated through the drive circuit 42 near the resonance frequency of the vibrator itself.

Therefore, the output from the vibrator appears as a signal that is AM-modulated at its resonance frequency, and the Coriolis force detected by the vibrator 40 is detected through the periodic detection circuit 41, which corresponds to a normal angular velocity signal. Get the output to

The output through the periodic detection circuit 41 is
There is a predetermined offset voltage (null voltage) even when the angular velocity input is 0. To remove this unnecessary DC voltage component, an OP amplifier 43, a capacitor 44, a resistor 45,
Signal components below a predetermined frequency are cut through an active high-pass filter circuit composed of 46 and 47, and only necessary shake signal components are input to the A / D converter.

Therefore, in FIG. 1, the output of the shake sensor 17 is
The output of the correction system position detection means 19 for detecting the current position of the shift correction optical system is also input to the correction system control means 20. Here, data for driving the shift correction optical system 4 according to specific control described later. Then, by moving the lens through the correction system driving means 5, the subject image is always stabilized on a predetermined image plane without shaking.

On the other hand, the subject image signal formed on the image pickup means 6 is converted into predetermined digital data by a series of video signal processing circuits composed of 6 to 16.

First, the charges accumulated by the image pickup means 6 (generally using an image pickup device such as a CCD) over a predetermined period of time are sequentially read out for each pixel, and at the same time, the charges are stored by the A / D conversion means 7. The object brightness information corresponding to the amount is converted into digital data.

Here, since an optical color filter for producing each color signal such as RGB is attached on the developing means 6, the output signal from the imaging means appears as a signal indicating each color alternately. . Normally, the output value from the A / D conversion means 7 is input to a process processing circuit 8 before actual photographing, where dark level correction, γ conversion, and the like are performed, and the result is converted to an image. The signal is input to the synthesizing circuit 9.

Here, the actual processing in the image synthesizing circuit will be described with reference to FIG. The color filter array of the image sensor used in this figure is a general Bayer array, and is a G (green) checkered, R (red), and B (blue) line sequential array. Therefore, in the case of a single-chip image sensor, not all pixels have RGB information. For example, all pixel points on the image sensor are obtained by an interpolation operation using a 3 × 3 matrix matrix shown in the center of the figure. It is common to create RGB color information in.

In FIG. 5, although the G interpolation filter and the R / B interpolation filter are different, for example, when a G signal at the position a is generated, the luminance data of a surrounded by a dotted line and eight pixels surrounding it are represented by a dotted line. , G by multiplying the coefficients of the interpolation filters.

In this case, the coefficient for the G output at the position a is 1 and its top, bottom, left and right are 0.25. However, since the G output at this position is 0, the output at the position a is substantially output. G data is determined only by the value. On the other hand, when the G signal at the position b is generated, the luminance signal of the pixel a surrounded by a dotted line and eight pixels surrounding the luminance signal are similarly multiplied by the coefficients of the G interpolation filter. Since the G output is 0, the G data at this position is determined using the average value of the up, down, left, and right G outputs.

Similarly, for R / B, R
Determine the R / B data for all pixel points using the / B interpolation filter. In this manner, RGB outputs for all pixel points can be finally generated as shown on the right end of FIG.

The RGB data calculated by the above method is transferred to the video memory 15 for each frame, and the monitor display means 1 is operated based on the data on the video memory.
6 is used to display a finder image on the photographing screen.

On the other hand, at the time of actual shooting, each output value passed through the process processing circuit 8 is first directly transferred to the frame memories 11 and 12, and the entire screen data is temporarily stored therein.

Next, the image synthesizing circuit 9 synthesizes the contents of the frame memory by the method described above, and transfers the RGB data corresponding to each pixel to the work memory 13 this time.

Further, the memory control means 10 compresses the contents of the work memory based on a predetermined compression format, and stores the result in the external memory 14 (usually constituted by a nonvolatile memory such as a flash memory).

On the other hand, when observing the image data stored in the external memory 14, the data is temporarily transferred to the memory control means, and the same decompression processing as the previously set compression format is performed. After that, the result is transferred to the work memory 13. Further, by transferring the data on the work memory to the video memory via the image synthesizing means 9, the already photographed image is displayed on the finder or the like through the monitor display means 16.

Next, the actual sequence operation of the camera will be described with reference to the flowchart of FIG.
First, in step 100, the MAINSW of the camera
It is determined whether or not the main switch (corresponding to a part of the camera operation switch 18 in FIG. 1) is turned on. If the MAINSW is turned on by the operation of the photographer, it is immediately determined in step 101. Then, power is supplied to all the circuit blocks shown in FIG.

Next, at step 102, the signal from the imaging means 6 is converted into a video signal through the respective circuits of the A / D conversion means 7, the process processing means 8 and the image synthesizing means 9 as in the method described above. The operation is started, and in step 103, a monitor display operation for the video signal is started. Therefore, after steps 102 and 103, the video signal processing operation is repeated for each frame.

Subsequently, at step 104, an interrupt process for executing an actual image stabilizing operation or a pixel shifting operation is permitted through a shake detection / correction interrupt process described later, and the shift is performed via the lens driving means 5 as described above. The driving of the correction optical system 4 is started. Therefore, thereafter, during the execution of the main operation, the shake detection / correction interrupt processing shown in FIG. 12 is executed at predetermined time intervals.

After the interrupt permitting operation has been performed, it is determined in step 105 what the photographing mode of the camera is. The setting of the photographing mode setting means 2 in FIG.
In the case of normal photographing for photographing by a general photographer, the process proceeds to step 106, where the internal flag PM
ODE is set to 0, and the routine proceeds to step 108.

On the other hand, if the setting of the photographing mode setting means 2 is set to the pixel shift photographing in step 105, the internal flag PMODE is set to 1 in step 107, and the routine proceeds to step 108.

After the above operation, in step 108, it is determined whether or not a release operation has been performed by the photographer.
The release switch in the camera operation switch 18 of FIG. 1 is turned on.
This switch is still OFF
In this case, the flow returns to step 105 again, and the determination of the shooting mode is repeated.

On the other hand, if the switch is ON, the flow proceeds to step 109, where the state of the internal flag PMODE set in the above steps 106 and 107 is determined. Execute the storage mode 1.

The operation in the photographing / recording mode 1 will be described with reference to the flowchart of FIG. First, in step 200, 1 is assigned to a parameter K for selecting a frame memory for temporarily storing the output of the process processing circuit 8, and the frame memory 1 is designated.

Next, in step 201, it is determined whether or not the image data accumulating operation of the image pickup means 6 has been completed.
It waits here until accumulation is completed.

Here, in the image pickup means such as a normal CCD, when the accumulation operation for a predetermined time is completed, the charges generated by the photoelectric conversion operation are immediately transferred to the transfer section, and the generated charges are sequentially read out. It is assumed that the next charge accumulation operation is being performed during the operation.

Therefore, in the next step 202, the results of the process processing for each pixel data as described above are sequentially stored in the frame memory K (in this case, the frame memory 1 indicated by 11 in FIG. 1). When it is detected in step 203 that all pixel data in one frame has been stored in the frame memory K, the process proceeds to the next step 204.

In step 204, the contents of the frame memory are first transferred to the pixel synthesizing circuit 9, where the interpolation operation for the missing RGB information for each pixel is executed as in the method described above. The result is temporarily transferred to the work memory in step 205. This operation is continuously performed for one frame. When it is detected in step 206 that the processing for one frame is completed, the process proceeds to step 207.

In steps 207 to 211, a method of compressing an actual photographed image and a method of storing data will be described. First, in step 207, execution of lossless compression is set for the memory control circuit 10 as a method of compressing an actual image.

The reversible compression type is a DPCM (Differential PC) as a specific compression method in the JPEG format that defines the compression standard for still images.
M) etc. are used. The DPCM method is based on the idea that only the difference between adjacent pixels among the pixels included in the image data is transmission-coded. According to this method, the compression ratio (the generated (Image size / original image size x 100) can be compressed only up to about 50%, but the original image can be completely restored from any photographed subject, so it is suitable for use when the original image does not need to be further degraded. ing.

Therefore, in step 208, the DPCM
Lossless compression such as the compression method is performed for each block of the original image (in this case, it is not always necessary to use the block). In step 209, the actually compressed image data is subjected to Huffman coding (the probability of occurrence is calculated). (A short code length is assigned to a high code, and a long code length is assigned to a code with a low probability of occurrence).

Next, the encoded image data is sequentially stored in the external memory 14 as shown in step 210, and in step 211 all the images (all blocks) are compressed and stored in the external memory. It detects that the saving is completed, and ends.

The shooting / storage mode 1, which is a normal shooting, ends after a series of operations as described above. Next, during this operation, an interruption operation is performed at predetermined time intervals to perform processing. The processing will be described with reference to the flowchart in FIG.

This flowchart mainly describes the internal operation of the correction system control means 20 shown in FIG. 1, and comprises a periodic interruption and a data transfer operation for the above-described overall control operation.

First, in step 300, the operation of converting the output of the shake sensor 17 into digital data via the A / D conversion circuit in the correction system control means 20 is started.
At 01, a predetermined time is waited until the A / D conversion operation is completed. When the completion of the A / D conversion is detected, the result of the conversion is transferred to the internal register U in step 302.

Next, at step 303, this register U
Is input, and a high-pass filter operation for removing unnecessary DC components (here, mainly the DC offset in the amplifier section including the OP amplifier 43 shown in FIG. 2) included in the shake sensor 17 is executed. However, this operation will be described with reference to the flowchart of FIG. Here, as a simple high-pass filter circuit, a first-order advance circuit surrounded by a dotted line portion C in FIG. 13 is used, and this input / output transfer characteristic is expressed using a used resistance value R ₁ and a used capacitance value C _1. Then, H (S) = VOUT / VIN = S · C ₁ · R ₁ / (1+
S · C ₁ · R ₁ ).

When this transfer characteristic is replaced on a Z plane for representing discrete characteristics, H (Z) = (a ₀ + a ₁ · Z ^-1 ) / using a known SZ transformation. (1 + b ₁ · Z ⁻¹ ).

Here, when each coefficient value a ₀ , a ₁ , b ₁ is expressed by using a data sampling time interval T _s , a ₀ = (2 / T _s ) / (1 / C ₁ / R ₁ +2 / T _s ) a ₁ = (− 2 / T _s ) / (1 / C ₁ / R ₁ + 2 / T _s ) b ₁ = (1 / C ₁ / R ₁ -2 / T _s ) / (1 / C ₁ /
R ₁ + 2T _s ).

A predetermined coefficient value is obtained in advance by the above conversion method, and this value is set in the internal registers A0, A1, and B1 in steps 350 to 352.

Next, at step 353, one of the calculation results calculated by the same processing at the previous sampling timing is transferred from the internal memory M (WH) storing it to the internal register W1. In 354, the internal register U in which the input data is set as the first operation
, The result of the multiplication of the registers B1 and W1 is subtracted, and the result is transferred to another internal register W0.

In step 355, the result of multiplication of the internal registers A0 and W0 is added to the result of multiplication of the internal registers A1 and W1, and the result is set in the internal register V. By storing the value of the register W0 calculated in the above in the internal memory M (WH), all the operations of the high-pass filter are completed.

Referring again to the flowchart of FIG. 12, first in step 304, the value of the internal register V storing the result of the high-pass operation is stored in the internal register U for the next operation.
Transfer to Then, in the next step 305, the angular velocity signal after removing the unnecessary DC component by the above calculation is
An integration operation for converting into an angular displacement signal is executed.

This integration operation will be described with reference to the flowchart of FIG. Here, as a simple integration circuit, a first-order delay circuit surrounded by a dotted line portion D in FIG. 14 is used, and this input / output transfer characteristic is expressed using a used resistance value R ₁ and a used capacitance value C _1. , H (S) = VOUT / VIN = 1 / (1 + S · C 1 · R
₁ )

When this transfer characteristic is replaced on a Z plane for expressing discrete characteristics, H (Z) = (a ₀ + a) by using a well-known SZ conversion, similarly to the high-pass filter operation. ₁ · Z ⁻¹ ) / (1 + b ₁ · Z ⁻¹ ).

Here, when each coefficient value a ₀ , a ₁ , b ₁ is expressed by using a data sampling time interval T _s , a ₀ = (2 / T ₂ ) / (1 / C ₁ / R ₁ +2 / T _s ) a ₁ = (− 2 / T _s ) / (1 / C ₁ / R ₁ + 2 / T _s ) b ₁ = (1 / C ₁ / R ₁ -2 / T _s ) / (1 / C ₁ / R
₁ + 2 / T _S ).

A predetermined coefficient value is obtained in advance by the above conversion method, and this value is set in the internal registers A0, A1, and B1 in steps 400 to 402.

Next, at step 403, one of the calculation results calculated by the same processing at the previous sampling timing is transferred from the internal memory M (WI) storing it to the internal register W1, and At 404, the internal register U in which the input data is set as the first operation
, The result of the multiplication of the registers B1 and W1 is subtracted, and the result is transferred to another internal register W0.

Further, in step 405, the multiplication result of the internal registers A1 and W1 is added to the multiplication result of the internal registers A0 and W0, and the result is set in the internal register V. By storing the value of the register W0 calculated in (1) in the internal memory M (W1), all the integration operations are completed.

After transferring the value of the internal register V obtained as a result of the integration operation output calculated by the above operation to the internal register U in step 306, in step 307, the zoom position (z) of the photographing lens 3 in FIG. Sensitivity based on the focus position (f) (a value that sets how much the shake correction system is moved with respect to the actual shake signal) is represented by a function k
It is set in the internal register K according to (z, f). In step 308, the value of the register K is multiplied by the internal register U storing the result of the integration operation to convert the result into a drive amount necessary for the actual shift correction driving. Set to DR.

Subsequently, in step 309, the state of the internal flag PMODE uniquely set in the photographing mode in the entire sequence shown in FIG. 9 is determined. Here, since the current shooting mode is the normal shooting mode, the value of PMODE is set to 0.
Therefore, step 310 is executed next.

In step 310, the output from the correction system position detection circuit shown in FIG.
8 input AN-A is selected. Here, since the amplifier circuit section surrounded by the dotted line A input to AN-A has a voltage setting that covers the entire stroke of the correction system,
The entire stroke range can be detected by the A / D converter 98.

After the above settings are made, the actual A / D conversion operation is started in step 311, and the process waits until this conversion is completed in step 312, and the result of this A / D conversion is stored in step 313. Set in register U.

At step 314, the internal register U
Is multiplied by a predetermined gain value H ₀ to set the sensitivity gain (in this case, for adjusting the actual movement amount to a predetermined digital value) to an appropriate value. Reset to the register PS.

Next, at step 322, the value of the internal register DR storing the sensor driving amount detected from the shake sensor output and the value of the internal register PS storing the current correction system position output value detected by the above method are stored. And the result DR-PS is set in the internal register U.
The value set in the internal register U is the difference between the actual shake amount at the present time and the correction amount in the correction optical system at that time. The difference should be zero. In practice, step 32
As indicated by 3, the difference between the two is multiplied by a predetermined gain value LPG (normal feedback system gain), and the difference is amplified and set in the internal register U again.

In step 324, a phase compensation operation is performed on the value of the internal register U to stably operate the feedback of the overall control system. This phase compensation calculation will be described with reference to the flowchart of FIG.

Here, as a standard phase compensation circuit, a phase lead compensation circuit surrounded by a dotted line portion E in FIG. 15 is used.
Expressing this input / output transfer characteristic using the used resistance values R ₁ , R ₂ and the used capacitance value C ₁ , H (S) = VOUT / VIN = (R ₂ + S · C ₁ · R ₁
· R ₂₎ / become _{(R 1 + R 2 + S} · C 1 · R 1 · R 2).

When this transfer characteristic is replaced on a Z plane for representing discrete characteristics, H (Z) = (a ₀ + a ₁ ) using a known SZ conversion similar to that described above. · Z ^-1 ) / (1 + b ₁ · Z ^-1 ).

Here, when each coefficient value a ₀ , a ₁ , b ₁ is expressed by using a data sampling time interval T _S , a ₀ = (1 / C ₁ / R ₁ + 2 / T _s ) / (( R ₁ + R
_{2) / C 1 / R 1} / R 2 + 2T s) a 1 = (1 / C 1 / R 1 -2 / Ts) / ((R 1 + R
_{2) / C 1 / R 1} / R 2 + 2T s) b 1 = ((R 1 + R 2) / C 1 / R 1 / R 2 -2T s)
/ (R ₁ + R ₂ ) / C ₁ / R ₁ / R ₂ + 2 / T _s ).

A predetermined coefficient value is obtained in advance by the above conversion method, and this value is set in the internal registers A0, A1, and B1 in steps 450 to 452.

Next, at step 453, one of the calculation results calculated by the same processing at the previous sampling timing is transferred from the internal memory M (WS) storing it to the internal register W1. At 454, the internal register U in which the input data is set as the first operation
, The result of the multiplication of the registers B1 and W1 is subtracted, and the result is transferred to another internal register W0.

Further, in step 455, the result of multiplication of the internal registers A0 and W0 is added to the result of multiplication of the internal registers A1 and W1 and the result is set in the internal register V. By storing the value of the register W0 calculated in (1) in the internal memory M (WH), all the phase compensation calculations are completed.

Next, at step 325, the value of the internal register V storing the result of the phase compensation calculation is reset to the internal register U. At step 326, the result of the calculation is passed through a D / A converter (not shown). The data is converted into analog data and used as input data to the correction system driving means 5. And
Finally, it was described in the description of the shift correction unit in FIG.
The correction system is driven in a predetermined direction via the magnetic circuit.

As described above, the difference between the actual shake amount and the position amount of the correction system with respect to the actual shake amount is obtained at every predetermined time interval, and the correction system is always subjected to feedback control using the amplified current amount, so that friction and the like can be obtained. In this way, it is possible to always achieve accurate shake correction without being affected by the above. In this operation, only the shake correction around one axis direction has been described, but the operation is the same for the other axis, so that the description is omitted here.

As described above, the shake detection / correction interrupt processing during the photographing / recording mode 1 in FIG. 10 is completed, but finally, in step 112 of the overall sequence in FIG. It is determined whether or not the release switch is off. If the release switch remains on, the process stays at step 112 as it is. When the release switch is detected, the process returns to step 108 again.

On the other hand, if the state of the internal flag PMODE set by the state of the photographing mode setting means 2 in FIG. 1 is 1 in step 109 of the camera sequence in FIG. The shooting / storage mode 2, which is the shift shooting mode, is executed.

Here, what kind of the pixel shift shooting is will be described with reference to FIG. The upper diagram schematically shows the RGB arrangements of the original image, and forms the Bayer arrangement described above.

The original optical image data is decentered by a predetermined amount in the X direction in the correction optical means 4 in FIG. 1 during the next one frame period, as shown in the lower left end in FIG.
Image data shifted from the original image by one pixel pitch in the horizontal direction can be obtained.

Therefore, by the first pixel shift, it is possible in principle to double the spatial frequency of a horizontal image for each color.

In the second pixel shift, the correction optical unit 4 is decentered by a predetermined amount in the X direction and the Y direction in the first pixel shift state, as shown in the lower center of FIG. It is possible to obtain image data that is shifted from the original image by a half pixel pitch in the oblique direction.

Further, in the third pixel shift, the correction optical unit 4 is again decentered only in the X direction in the second pixel shift state, so that the original image as shown at the lower right end in FIG. 6 is obtained. On the other hand, it is possible to obtain image data shifted by a half pixel pitch in the oblique direction. By shifting the original image by a predetermined pixel pitch for each frame and combining and combining a total of four shots of image data, the spatial frequency of the image in both the horizontal and vertical directions is improved to about twice It is possible to do.

Next, the actual pixel shift shooting will be described with reference to the shooting / storage mode 2 shown in the flowchart of FIG. First, in step 250, 1 is assigned to a parameter K for temporarily storing the output of the process processing circuit 8 and for selecting a frame memory, and specifies the frame memory 1.

Next, in step 251, it is determined whether or not the image data accumulating operation in the photographing means 6 has been completed.
It waits here until accumulation is completed. Here, in the imaging means such as a normal CCD, when the accumulation operation for a predetermined time is completed, the charge generated by the photoelectric conversion operation is immediately transferred to the transfer unit, so that the generated charge is sequentially read out. Assume that the next charge accumulation operation is being performed.

When the image storing operation of the original image as shown in FIG. 6 is completed, next, in steps 252 and 253, the eccentric data amount of the correction optical means for realizing the first pixel shift, ΔX ( K) and ΔY (K) are set, and the correction optical unit 4 is actually driven eccentrically via the lens driving unit 5.

In this case, the initial amount of eccentricity ΔX (1) is such that the subject is shifted on the image pickup plane by one pixel pitch with respect to the original image, and ΔY (1) is 0 since it is not eccentric in the Y direction. It is.

Therefore, in the next step 254, the result of processing the original image for each pixel data is stored in the frame memory K (in this case, the frame memory 1 shown in FIG. 1 in order). Then step 2
At 55, when it is detected that all the pixel data in one frame has been stored in the frame memory K, the next step 2
Proceed to 56.

At step 256, it is determined whether or not the value of the frame memory setting parameter K is equal to N (in this case, 4).
Is incremented by one, and it is determined again in step 251 whether or not the accumulation of the next one frame is completed.

When the completion of image storage is detected in step 251, this time, in steps 252 and 253, ΔX
(2) and ΔY (2) are set such that the values are shifted by a half pixel pitch in the oblique direction with respect to the original image, and then the operations of steps 254 to 257 are repeated.

When steps 252 and 253 are executed again, ΔX (3) is set to a value which is shifted by one pixel pitch in the horizontal direction with respect to the second pixel shift, and ΔY (3) Is 0.

As described above, the process is repeated at step 256 until the value of K becomes equal to N (4 in this case), and as shown in FIG. 6, a predetermined pixel pitch is set in the X and Y directions for each frame. It is possible to obtain four frames of images that are shifted from one another.

FIG. 7 shows the state of the above-mentioned pixel shift photographing along the movement of the correction optical means 4 a little more. FIG. 7 shows the actual movement of the correction optical means 4 in the X and Y directions with respect to the time axis t. First, the correction system is driven based on the output of the shake sensor, and the first time After photographing (end of image storage), the correction optical unit 4 sets the Δ
The eccentric movement is performed in parallel by X (1), and the second photographing is performed in this state.

After the end of the second shooting, the eccentric movement is performed in the X and Y directions by ΔX (2) and ΔY (2), respectively, and the third shooting is performed. After the end of the third imaging, the eccentric movement is performed only in the X direction by ΔX (3), and the fourth imaging is performed to complete the entire operation.

Next, after step 258, the high-density image data actually obtained by the pixel shift is
An operation of converting into GB information is performed. First, in step 258, the value of a parameter K that specifies the frame memory that stores the image data captured in the first shooting in the pixel shift shooting is set to 1.

Subsequently, the contents of the frame memory are first transferred to the image synthesizing circuit 9, and here, unlike in the case of the above-described shooting / storage mode 1, interpolation for the missing RGB information for each pixel is immediately performed. The operation is not executed, and only the determination as to whether or not the transfer of one frame has been completed is performed in step 260.

If it is detected in step 260 that the transfer of one frame has been completed, the process proceeds to step 261 and
Here, it is determined whether or not the value of K is equal to N (4 in this case) in order to detect that the transfer of all captured image data has been completed. If the transfer of all captured image data has not been completed yet, the value of K is incremented by one in step 262, and the process proceeds to step 259 again to start transferring the contents of the next frame memory.

When the transfer of all photographed data is finally completed, the value of K becomes equal to N in step 261 and then the process proceeds to step 263, where the actual synthesis of all photographed image data is performed for the first time.

The state of this image composition will be described with reference to FIG. The left end of this figure is obtained by spatially rearranging the array of pixel data obtained after the pixel shift. Compared to the image data of the original Bayer array image sensor shown in FIG. It is an image data array with nearly double the spatial frequency.

However, in this case, in order to obtain twice as much RGB information in both the horizontal and vertical directions, it is necessary to apply an interpolation filter composed of a matrix matrix shown in the center of this figure to this image data. .

First, regarding the G component, in this case, the same 3 × 3 matrix matrix as before is sufficient,
For example, when generating a G signal at the position of a, it can be obtained by multiplying each luminance data of a surrounded by a dotted line and eight pixels around the a by a coefficient of a G interpolation filter.

In this case, the coefficient with respect to the G output at the position a is 1 and its upper, lower, left and right are 0.25. However, since the G output at this position is 0, only the output value is substantially at the position a. Determines the G data.

On the other hand, when producing a G signal at the position b,
Similarly, it can be obtained by multiplying each of the luminance data of b surrounded by the dotted line and the eight surrounding pixels by the coefficient of the G interpolation filter. In this case, there is no G signal at the position of b, so The G data at this position is determined by using the average value of the G signal.

Next, the R / B is a little more complicated, and as can be seen from the arrangement at the left end of this figure, the pixel data can be interpolated from the immediately adjacent pixel data in the horizontal direction, but can be interpolated in the vertical direction. It is necessary to interpolate using pixel data at positions slightly apart from each other, so use a 5 × 5 matrix matrix,
Moreover, the coefficient array is not point-symmetrical as viewed from the center of the matrix matrix as before.

By performing the above operation for each pixel array for each RGB, RGB information for the entire pixel array as shown at the right end of FIG. 8 can be finally calculated. .

Next, at step 264, first, all the data is temporarily transferred to the work memory 13 in order to actually compress and store the image data obtained by synthesizing the images from the four shootings. Subsequently, at step 265, the memory control circuit 10 is set to execute irreversible compression as the compression type (when the decompression operation is performed, the same image as the actual original image cannot be obtained).

As the irreversible compression method, for example, in the JPEG format which defines the compression standard for still images, for example, 8
A so-called DCT (Disc) which converts the image into two-dimensional frequency data of each image after dividing the image into blocks each having × 8 pixels.
For example, there is a Cosine Transform (rete Cosine Transform) conversion. According to this method, the data amount of the original image can be considerably reduced.

Accordingly, the actual execution of the compression operation is performed in step 266 by performing the irreversible compression such as the DCT method on the synthesized image after the pixel shift for each block unit (8 × 8 pixels is one block). Then, in step 267, the actually compressed image data is converted into actual compressed code data using Huffman coding or the like, as in the case of the shooting / storage mode 1. The encoded image data is sequentially stored in the external memory 14 as shown in step 268, and in step 269, the fact that the compression of all images (all blocks) and the storage in the external memory have been completed is completed. Detect and end.

Next, the shake detection / correction interrupt processing during the execution of the photographing / storage mode 2 will be described again with reference to FIG. In FIG. 12, the operations of steps 300 to 308 are as already described, and the description is omitted here.

In step 309, the value of the internal flag PMODE set in the camera shooting mode determination is determined. In this case, the pixel shift shooting mode for outputting a high-definition image has been set. Proceeding to step 315, AN-B is selected as an input for performing A / D conversion.

As shown in FIG. 4, the output of the inverting amplifier B surrounded by a dotted line is connected to the AN-B input as the final output of the shift correction optical system. The amplification factor is larger than that of the amplification unit A, and only the vicinity of the center of the stroke of the entire correction system is enlarged. Therefore, if this output is read by the A / D converter, it is possible to realize higher-definition position control than when reading the output of the amplifier A.

In step 316, an A / D conversion operation of the actual correction system position output is started.
It is determined whether or not the D / A conversion operation has been completed. When it is detected that the A / D conversion operation has been completed, the process proceeds to step 318, and the result of the A / D conversion is transferred to the internal register U.

In step 319, the internal register U
Is multiplied by a predetermined gain value H ₁ to adjust the sensitivity gain (in this case, to adjust the actual movement amount to a predetermined digital value.
Is set to a value smaller than the gain H ₀ ), and the result is stored in the internal register PS.
Reset to.

Next, in step 320, the value of the pixel shift amount DRX (or DRY) set in steps 252 and 253 in the above-described pixel shift shooting mode is uniquely determined by the zoom or focus state of the shooting optical system. The value of the variable value K to be determined is multiplied, and the result is stored in the internal register DRS.
Set to.

The values of DRX and DRY are both 0 at the time of the first photographing and DRX = ΔX at the time of the second photographing.
(1), DRY = 0, DRX = ΔX at the time of the third shooting
(2), DRY = ΔY (2), DRX for the fourth shooting
= ΔX (3) and DRY = 0 are set before the actual start of each shooting.

Subsequently, in step 321, step 30
The value of the correction system drive amount DR based on the shake sensor output determined in step 8 and the value of the correction system drive amount DRS due to this pixel shift are added, and the result is set in the internal register DR again.

Thereafter, steps 322 to 32 described above are performed.
By executing the operation of step 6, the difference between the actual correction system drive amount and the correction system position detection system is calculated, an appropriate loop gain is set and phase compensation is performed, and the data is converted into correction system drive data. The detection / correction interrupt processing ends.

In the case of the pixel shift photographing mode as described above, the correction stroke for sufficiently correcting the actual camera shake becomes difficult in the detection range, but the entire position detection resolution in a specific range is increased by that amount. The configuration has been changed.

As described above, the photographing / storage mode 2 in FIG. 11 is completed. Finally, it is determined in step 112 in FIG. 9 whether or not the release switch of the camera is off.
If the release SW remains on, step 112 is left as it is.
And the process returns to step 108 again when the power is turned off.

As described above, in this embodiment, the correction system position detection is performed depending on whether the camera shooting mode is a shooting mode based on normal camera shake correction or a pixel shift shooting mode for outputting a high-definition image. By changing the sensitivity of the unit, one side allows sufficient handheld shooting by giving priority to stroke, and the other drives the correction lens to the exact position by giving priority to control resolution. The control is executed in a method most suitable for the mode.

More specifically, when the shooting mode of the camera is set to the anti-vibration shooting mode, the stroke range of the correction optical system is prioritized, and the resolution is read in a coarse state.
On the other hand, when the shooting mode is set to the pixel shift shooting mode, the stroke of the correction optical system is narrowed, and the resolution is made fine.

In the control of the correction optical system, the position of the lens itself is read and the feedback control is performed so that the position coincides with the target signal as in the above method. , Depending on the shooting mode.

For example, in the image stabilization mode, a signal from 1 Hz to 20 Hz corresponding to camera shake is detected and corrected, so that the entire frequency characteristic is adjusted so that the phase lag of the correction system in that range is minimized. However, in the case of the pixel shift photographing, the correction optical system is driven with only a small displacement, and therefore, how to accurately move to the next target position against the static friction is an important point in determining the characteristics.

(Second Embodiment) Next, a second embodiment of the present invention will be described with reference to the flowchart of FIG. This flowchart is similar to the shake detection / correction interrupt processing of FIG. 12, in which the interrupt operation is periodically performed during the execution of the shooting / storage mode 1 of FIG. 11 and the shooting / storage mode 2 of FIG. The processing is performed.

First, the operations of steps 500 to 508 are exactly the same as the operations of steps 300 to 308, and the detailed description is omitted here.
After converting to digital data via the / D converter,
Unnecessary DC components are removed through a high-pass filter, converted into an angular displacement amount through an integration operation, and then subjected to shift correction sensitivity processing based on the zoom state and the focus state of the photographing optical system to detect actual shake. The target drive amount for the amount is calculated.

Next, at step 509, AN-A is selected as the input of the A / D converter, and the inverting amplifier A is selected as the position detection processing of the correction optical system shown in FIG. Therefore, in this case, the total stroke of the shift correction optical system is set to A /
It will be taken in via a D converter.

At step 510, the A / D conversion operation of the correction system position output is actually started. At step 511, after waiting for the A / D conversion operation to be completed, the process proceeds to step 512 when the conversion is completed. Then, the conversion result is set in the internal register U.

In step 513, this internal register U
Is multiplied by a predetermined gain value H ₀ to set the sensitivity gain (in this case, for adjusting the actual movement amount to a predetermined digital value) to an appropriate value. Reset to the register PS.

Next, at step 514, the internal flag PM set uniquely according to the photographing mode of the camera is set.
The state of the ODE is determined. If the shooting mode of the camera is the normal shooting mode, the value of PMODE is 0, so the process proceeds directly to step 517. However, if the shooting mode of the camera is the pixel shift mode for outputting a high-definition image, the PMODE The value is 1, and in this case, step 515 and subsequent steps are executed.

First, in step 515, as described above, the value of the pixel shift amount DRX (or DRY) set in steps 252 and 253 in the pixel shift shooting mode is uniquely determined by the zoom or focus state of the shooting optical system. Is multiplied by the value of the variable value K to be determined, and the result is stored in the internal register D.
Set to RS. The values of DRX and DRY are the same as the values described above.

Subsequently, at step 516, at step 50
The value of the correction system drive amount DR based on the shake sensor output determined in step 8 and the value of the correction system drive amount DRS due to this pixel shift are added, and the result is set in the internal register DR again.

Next, at step 517, the value of the internal register DR that stores the sensor drive amount detected from the shake sensor output and the internal register PS that stores the current correction system position output value detected by the above method. And the result is set in the internal register U. The value set in the internal register U is the difference between the actual shake amount at the present time and the correction amount in the correction optical system at that time. The difference should be zero.

At step 518, the internal flag PMO is again set.
The state of DE is determined, and if the value of PMODE is 0, the process proceeds to step 519, where the value of this register U is multiplied by a predetermined gain value LPG ₁ (gain of a normal feedback system), and the internal Set in register U.

In step 520, a phase compensation operation -1 is executed on the value of the internal register U in order to stably operate the feedback of the overall control system. This phase compensation calculation-1 is realized by performing a predetermined calculation according to the flowchart of FIG. 15 described above, and each constant value is set to each value of R1, R2, and C1 in the right diagram. Can be uniquely determined.

Here, by setting the constant to an appropriate value, the frequency characteristics shown in FIGS. 17A and 17B are obtained. The characteristic shown in FIG. 17 shows the closed-loop characteristic of the shift correction optical system when the above-described phase compensation operation -1 is executed, and the frequency band of camera shake (20 Hz)
The gain is set to 1 and the phase lag is reduced as much as possible up to about 100 Hz so that the phase delay can be covered.

On the other hand, as a result of determining the state of the internal flag PMODE in step 518, the value of PMODE becomes 1
In the case of (1), the routine proceeds to step 521, where the value of the register U is multiplied by a predetermined gain value LPG ₂ (normal feedback system gain), and is set again in the internal register U.

At step 522, a phase compensation operation-2 for stably operating the feedback of the overall control system is performed on the value of the internal register U. Unlike the phase compensation calculation-1, the phase compensation calculation-2 is for executing control suitable for performing pixel shift photographing.
The frequency characteristics of the shift correction optical system when this calculation is performed are as shown in FIGS. 17 (c) and (d). In the case of this characteristic, emphasis is placed on the positional accuracy of the correction optical system for the pixel shift photographing by removing the actual camera shake of the photographer, and the closed loop gain near DC is as close to 1 as possible. The phase delay up to several Hz is set as small as possible.

Next, at step 523, the value of the internal register V storing this phase compensation calculation result is reset to the internal register U, and this calculation result is set at step 524 via a D / A converter (not shown). The data is converted into analog data and used as input data to the correction system driving means 5. And
Finally, it was described in the description of the shift correction unit in FIG.
The correction system is driven in a predetermined direction via the magnetic circuit.

As described above, in this embodiment, the actual correction optics depends on whether the camera shooting mode is a mode premised on normal image stabilization shooting or a pixel shift shooting mode for outputting a high-definition image. The frequency characteristic of the system is changed, and control suitable for both shooting modes is executed.

(Third Embodiment) Next, a third embodiment of the present invention will be described with reference to the flowchart of FIG. In the present embodiment, in the entire sequence shown in the flowchart of FIG. 9, the shake detection / correction interrupt processing of FIG. 18 is executed, and the control operation is changed according to the set photographing mode of the camera.

In step 550, first, the internal flag PMO uniquely set according to the shooting mode of the camera is set.
The state of the DE is determined, and if this value is 0, step 5
Step 51 and subsequent steps are executed. Steps 551 to 559
This is exactly the same as steps 300 to 308 in FIG. 12. After the output from the shake sensor is converted into digital data via an A / D converter, unnecessary DC components are removed via a high-pass filter, and an integration operation is performed. Is converted to an angular displacement. Accordingly, the target drive amount DR for shake correction is finally calculated in step 559.

On the other hand, in step 550, the internal flag PMOD is
When the value of E is 1, the shooting mode of the camera is set to the pixel shift shooting mode for outputting a high-definition image. In this case, the shake sensor processing from the above steps 551 to 559 is executed. Instead, step 560 and subsequent steps are directly executed.

Next, at step 560, AN-A is selected as the input of the A / D converter, and the inverting amplifier A is selected as the position detection processing of the correction optical system of FIG. Therefore, in this case, the total stroke of the shift correction optical system is set to A /
It will be taken in via a D converter.

In step 561, the A / D conversion operation of the correction system position output is actually started. In step 562, the process waits until the A / D conversion operation is completed, and proceeds to step 563 when the conversion is completed. Then, the conversion result is set in the internal register U.

At step 564, the internal register U
Is multiplied by a predetermined gain value H ₀ to set the sensitivity gain (in this case, for adjusting the actual movement amount to a predetermined digital value) to an appropriate value. Reset to the register PS.

Next, at step 565, the internal flag PM set uniquely according to the shooting mode of the camera is set.
The state of the ODE is determined again. If the shooting mode of the camera is the normal shooting mode, the value of PMODE is 0, so the process proceeds directly to step 569. However, if the shooting mode of the camera is the pixel shift mode for outputting a high-definition image, the PMODE The value is 1, in this case step 566.
Execute the following.

First, at step 566, the value of the internal register DR in which the value of the target drive amount of the correction system is set is cleared to zero. Therefore, in the case of the pixel shift photographing mode, the output from the shake sensor is not used at all.

Next, in step 567, the value of the pixel shift amount DRX (or DRY) set in steps 252 and 253 in the pixel shift shooting mode as described above is univocally determined by the zoom or focus state of the shooting optical system. Is multiplied by the value of a variable value K to be determined
Set to RS. The values of DRX and DRY are the same as the values described above.

Subsequently, at step 568, the value of the correction system drive amount DRS cleared to 0 at step 566 and the value of the correction system drive amount DRS at this pixel shift are added, and the result is again stored in the internal register. Set to DR.

Next, at step 569, the value of the internal register DR that stores the sensor drive amount detected from the shake sensor output and the internal register PS that stores the current correction system position output value detected by the above method. And the result is set in the internal register U. The value set in the internal register U is the difference between the actual shake amount at the present time and the correction amount in the correction optical system at that time. The difference should be zero.

At step 570, the value of the register U is multiplied by a predetermined gain value LPG (normal feedback system gain), and the result is set in the internal register U again.

In step 571, a phase compensation operation is performed on the value of the internal register U to stably operate the feedback of the overall control system. This phase compensation calculation is realized by performing a predetermined calculation according to the flowchart of FIG. 15 described above, and each constant value is uniquely defined by setting each value of R1, R2, and C1 in the right diagram. Can be determined.

Next, at step 572, the value of the internal register V storing this phase compensation calculation result is reset to the internal register U, and this calculation result is set at step 573 via a D / A converter (not shown). The data is converted into analog data and used as input data to the correction system driving means 5. And
Finally, it was described in the description of the shift correction unit in FIG.
The correction system is driven in a predetermined direction via the magnetic circuit.

As described above, this embodiment depends on whether the camera shooting mode is a mode premised on normal image stabilization shooting or a pixel shift shooting mode for outputting a high-definition image. The signal processing itself is changed, and control suitable for both shooting modes is executed.

In this embodiment, as described above, no output from the shake sensor is used in the pixel shift photographing mode, so that the shake sensor 17 is controlled via the overall control means shown in FIG.
It is also possible to stop the power supply to the system.

More specifically, when the shooting mode of the camera is set to the image stabilization shooting mode, the power supply to the shake sensor is started at the same time as the main switch of the camera is turned ON.
The output is subjected to signal processing, and drive control of the correction optical system is started based on the result. In addition, when the camera shooting mode is set to the pixel shift shooting mode, even if the main switch of the camera is turned on, power is not supplied to the shake sensor, and the control system is controlled in accordance with a target position signal generated inside the control system. Drive.

As described above, the setting photographing mode of the camera itself is the image stabilizing photographing mode for removing the camera shake of the photographer,
Efficient control is possible by switching the processing of the shake sensor itself depending on the pixel shift shooting mode for capturing a high-definition image.

[0195]

As described above, according to the present application, when the photographing mode of the camera itself set by the photographer is the normal image stabilizing photographing mode, the correction system is used to eliminate the influence of camera shake of the photographer. The sensitivity of the processing circuit system is set so that the correction stroke is prioritized as the position detection, while in the pixel shift shooting mode to output a high-definition image, the processing circuit is set so that the detection resolution is prioritized over the correction stroke By setting the sensitivity of the system, the correction system can be accurately driven to a predetermined position, so that there is an effect that the optimum accuracy of the correction system is maintained in two different shooting modes.

Further, according to the present invention, when the photographing mode is the normal photographing, the frequency characteristic of the correction optical system is reduced to a certain degree with respect to the entire camera shake frequency (for example, up to several tens of Hz) (the ability to suppress the image shake amount). (Represented on each frequency axis), while if the shooting mode is a pixel-shifted shooting mode, the frequency characteristics of the correction optical system are set so that sufficient performance can be obtained near DC. Therefore, 2
In any of the three different photographing modes, there is an effect that a sufficient dynamic characteristic of the correction optical system can be brought out.

According to the present invention, when the photographing mode is the normal photographing, the calculation for the shake sensor output is executed as usual, and the shake correction optical system is driven using the output as the target signal. In the case of the photographing mode, no operation is performed on the shake sensor at all, and the correction system is driven based on a drive signal for simply switching the imaging position of the subject image for each photographing. The calculation time can be shortened, and the erroneous signal of the shake sensor or the like during the tripod shooting (the shake sensor itself generates a signal other than the original shake output due to mechanical vibration or the like generated inside the camera) does not have an adverse effect, Also, if the power supply to the shake sensor itself is stopped, unnecessary current does not need to flow, which leads to energy saving.

[Brief description of the drawings]

FIG. 1 is an overall configuration diagram of a camera according to the present invention.

FIG. 2 is a configuration diagram of a shake sensor and a processing circuit system.

FIG. 3 is a configuration diagram of a shift correction system.

FIG. 4 is a configuration diagram of a position detection processing circuit of a shift correction system.

FIG. 5 is a diagram illustrating a method of color synthesis when a normal image sensor is used.

FIG. 6 is a diagram for explaining the principle of pixel shifting according to the present invention.

FIG. 7 is a diagram for explaining an actual operation of pixel shifting according to the whole of the present invention.

FIG. 8 is a diagram illustrating color synthesis when pixel shifting is performed according to the entirety of the present invention.

FIG. 9 is a diagram showing an entire sequence of a camera according to the whole of the present invention.

FIG. 10 is a view for explaining the photographing / storing operation of the camera according to the whole of the present invention.

FIG. 11 is a diagram illustrating a photographing / storage operation of the camera according to the whole of the present invention.

FIG. 12 is a diagram illustrating an operation of shake detection / correction according to the first embodiment of the present invention.

FIG. 13 is a diagram illustrating an operation of a shake sensor high-pass calculation according to the entirety of the present invention.

FIG. 14 is a diagram illustrating an operation of a shake sensor integral operation according to the entirety of the present invention.

FIG. 15 is a diagram illustrating an operation of a correction system phase compensation calculation according to the whole of the present invention.

FIG. 16 is a diagram illustrating an operation of shake detection / correction according to a second embodiment of the present invention.

FIG. 17 is a diagram illustrating frequency characteristics of a correction system according to a second example of the present invention.

FIG. 18 is a diagram illustrating a shake detection / correction operation according to a third embodiment of the present invention.

Claims (19)

[Claims] 1. An image pickup means, a shake detection means for detecting a shake, an image shake correction means for correcting an image shake based on an output of the shake detection means, and a position of an image on the image pickup means, Pixel shifting means for minutely displacing using a shake correcting means; and image synthesizing means for synthesizing a high-resolution image based on a plurality of image data obtained by shifting the position of the image on the imaging means by the pixel shifting means. A first shooting mode for correcting image shake and a second shooting mode for synthesizing a high-resolution image can be selected, and the image shake is selected according to the selected shooting mode. An imaging apparatus, comprising: control means for changing drive control of a correction means. 2. The image blur correction device according to claim 1, wherein the image blur correction device includes a position detection device that detects a current position of the correction unit, and at least two or more images having different amplification factors in outputs from the position detection device. An imaging device comprising: a width section; and wherein the control means is configured to select an output of the amplification section in accordance with the selected shooting mode. 3. The image blur correction device according to claim 1, wherein the image blur correction device includes a correction optical system, and the control device determines a resolution of the correction optical system in the first photographing mode. Control in which the movable range is prioritized over the movable range, and in the second photographing mode, control in which the movable range of the correction optical system is narrowed and priority is given to enhancing resolution is performed. Characteristic imaging device. 4. The image processing apparatus according to claim 1, wherein the control unit adjusts a frequency characteristic of the image blur correcting unit.
An imaging apparatus comprising: a frequency characteristic changing unit that changes according to the shooting mode. 5. The frequency characteristic changing unit according to claim 4, wherein the frequency characteristic changing unit sets a frequency characteristic in the first shooting mode so as to reduce a phase delay with respect to a shake frequency range to be corrected by the image shake correcting unit. In the second imaging mode, the imaging apparatus is configured to set a frequency characteristic by giving priority to a response at the time of minute driving of the image blur correcting means by the pixel shifting means. 6. The responsiveness according to claim 5, wherein the frequency characteristic in the second photographing mode is such that the image shift correcting means can be minutely driven to a target position by the pixel shift means against static friction. An imaging apparatus characterized in that the imaging apparatus is configured to be determined based on the information. 7. The imaging apparatus according to claim 1, wherein the control unit is configured to be arbitrarily switchable by an operator. 8. An image pickup means, a shake detection means for detecting a shake, an image shake correction means for correcting an image shake based on an output of the shake detection means, and a position of the image on the image pickup means, Pixel shifting means for minutely displacing using a shake correcting means; and image synthesizing means for synthesizing a high-resolution image based on a plurality of image data taken by shifting the position of the image on the imaging means by the pixel shifting means. A first shooting mode for correcting image shake and a second shooting mode for synthesizing a high-resolution image can be selected, and the image shake is selected according to the selected shooting mode. An imaging apparatus comprising: a control unit that changes a method of processing a signal from a detection unit. 9. The imaging apparatus according to claim 8, wherein the control unit performs control so that shake detection is not performed when the second shooting mode is selected. 10. The imaging apparatus according to claim 9, wherein the control unit controls so as not to supply power to the shake detection unit when the second shooting mode is selected. . 11. The apparatus according to claim 10, wherein when the second photographing mode is selected, the control section outputs a target position of the image blur correcting section without energizing the blur detecting section. An imaging device characterized by being configured as described above. 12. The control unit according to claim 8, wherein the control unit prohibits the operation of the image shake correction unit based on the output of the shake detection unit when the second shooting mode is selected. An imaging device characterized by the above-mentioned. 13. The control unit according to claim 8, wherein, when the first photographing mode is selected, the control unit starts energizing the shake detecting unit in accordance with power-on of the imaging device, An image pickup apparatus characterized in that it is configured to drive and control the image shake correcting means based on an output of the shake detecting means. 14. An image pickup unit, a shake detection unit for detecting a shake, an image shake correction unit for correcting an image shake based on an output of the shake detection unit, and an image on the image pickup unit, A pixel shifting unit for minutely displacing the position of the image using the image blur correcting unit; and a high-resolution image based on a plurality of image data obtained by shifting the position of the image on the imaging unit by the pixel shifting unit. A method for controlling an imaging apparatus, comprising: an image synthesizing unit that synthesizes a high-resolution image with a first shooting mode for correcting image blur by the image blur correcting unit. A second photographing mode for the purpose of selecting a photographing mode, and driving control of the image blur correcting means is changed according to the selected photographing mode. . 15. The method according to claim 14, wherein in the first photographing mode, control is performed such that a movable range is prioritized over a resolution of a correction optical system of the image blur correcting means, and in the second photographing mode, A control method for an image pickup apparatus, wherein control is performed such that a movable range of a correction optical system is narrowed and priority is given to increasing resolution. 16. The second image capturing method according to claim 15, wherein in the first image capturing mode, a frequency characteristic is set so as to reduce a phase delay with respect to a shake frequency range to be corrected by the image shake correcting means. In the mode, a frequency characteristic is set with priority given to responsiveness at the time of minute driving of the image blur correcting means by the pixel shifting means. 17. An image pickup means, a shake detection means for detecting a shake, an image shake correction means for correcting an image shake based on an output of the shake detection means, and a position of an image on the image pickup means, Pixel shifting means for minutely displacing using a shake correcting means; and image synthesizing means for synthesizing a high-resolution image based on a plurality of image data taken by shifting the position of the image on the imaging means by the pixel shifting means. A method for controlling an image pickup apparatus, comprising: selecting a first shooting mode for correcting image blur and a second shooting mode for synthesizing a high-resolution image. A method of controlling an imaging apparatus, wherein a method of processing a signal from the shake detecting means is changed according to a possible and selected shooting mode. 18. The method according to claim 17, wherein when the second shooting mode is selected, control is performed so that shake detection is not performed. 19. The apparatus according to claim 17, wherein when the second photographing mode is selected, the control section outputs a target position of the image blur correcting section without energizing the blur detecting section. Control method for an imaging device, characterized in that the control is performed in the following manner.