JP2540801B2 - Automatic focus adjustment device - Google Patents

Description

Detailed Description of the Invention

[0001]

[Industrial applications]

The present invention relates to an automatic focus adjustment device that detects a focus state of a photographing lens by receiving a subject image using a line image sensor and drives the photographing lens based on the detected focus state to perform focus adjustment.

[0002]

[Prior art]

Conventionally, the line image sensor as described above has been used for a focus detection device of a camera, for example. However, this line image sensor has the following problems because it can detect only a linear light amount distribution in a single direction. In a focus detection device of a type that analyzes a video signal obtained by an image sensor, reliable focus detection cannot be performed unless the image on the image sensor has a certain degree of contrast. When a line image sensor is used as the image sensor, focus detection cannot be performed if the cotton last in the line direction of the image on the image sensor is low, but even then the contrast of the image in a direction different from the sensor line is often sufficient. . When taking a picture of a person or an outside scene, the contrast in the horizontal direction is often higher than the contrast in the vertical direction.Therefore, when using a line image sensor for focus detection, the image sensor should be placed horizontally. It is rational. However, in this case, focus detection cannot be performed for a subject that happens to have a low horizontal contrast but a high vertical contrast. The same problem also occurs when the camera is used in the vertical position. This problem is solved by using a two-dimensional image sensor. As a focus detecting device of this type, a proposal has been made by JP-A-59-174807. The point of this proposal is to arrange the light-receiving elements two-dimensionally and form a subject image on it, and read the output of one row of light-receiving elements from this array of light-receiving elements to find if the contrast is insufficient. , Change the direction and read the output from the light receiving element array. In this way, focus detection calculation is performed by searching for a direction in which sufficient contrast is obtained, but it is expensive because a two-dimensional image sensor is used. Therefore, a light receiving device that can detect the light amount distribution in a plurality of directions by using an inexpensive line image sensor is desirable. Also, in particular, when the subject light that has passed through the two regions on both sides of the optical axis across the lens is used to form the subject images separately on the planned focal plane, the subject in the two regions on the lens Since the respective images are displaced due to the parallax with respect to, the focus adjusting means of the method of calculating the amount of defocus of the lens by detecting the amount of this deviation is reliable for a subject with poor contrast in the alignment direction of the above two areas. It is difficult to obtain a proper differential focus amount, and this difficulty cannot be eliminated even if the image sensor has a two-dimensional resolution.

[0003]

[Problems to be Solved by the Invention]

An object of the present invention is to obtain an automatic focus adjustment device capable of reliable focus detection for any subject using a line image sensor, and particularly to solve the problem of the directionality of the contrast distribution of the subject. It is about to be resolved.

[0004]

[Means for solving problems]

The present invention relates to an automatic focus adjusting device for automatically adjusting the focus of a taking lens, which is formed by the optical means for forming a subject image by light from the subject and the optical means arranged in the horizontal and vertical directions. A plurality of line image sensors that receive the subject image, and detect the focus state for each array direction of each line image sensor based on the video signal output from each of these line image sensors,
A focus adjusting means for driving the taking lens based on the focus state and a control means for controlling the focus adjusting operation by giving priority to a line image sensor arranged in the horizontal direction among the line image sensors. And by
The automatic focus adjusting device is configured such that the control means is arranged vertically when the reliability of the focus state detected by the output of the line image sensor arranged horizontally is low. Based on the focus state detected by the output of the line image sensor, the shooting lens is controlled to be driven, or the focus state is first detected by the output of the line image sensor arranged in the horizontal direction, and the focus state of this focus state is detected. When the reliability is high, the focus state is not detected by the output of the line image sensor arranged in the vertical direction.

[0005]

[Action]

According to the present invention, since a plurality of line image sensors are arranged in the horizontal direction and the vertical direction, even if the contrast distribution of the subject has directionality, reliable focus detection can always be performed. Further, since the line image sensor is used, the cost of the device is lower than that when a sensor having a two-dimensional resolution is used. In addition, there are more changes in the horizontal direction than in the vertical direction in the subject actually shot,
Since the human eye is sensitive to changes in the horizontal direction, if the line image sensor arranged in the horizontal direction is given priority to perform the focus adjustment operation, the better focus adjustment operation will be performed.

[0006]

【Example】

FIG. 1 shows an optical system and an image sensor according to an embodiment of the present invention.
FIG. 2 is a perspective view of the arrangement of the support, and FIG. 2 is an exploded perspective view of the apparatus of the embodiment.
FIG. In FIG. 1, 1 is a taking lens of a camera, and 6 is a lens.
Densa lens, 8 is four reprojection lenses, and 10 is an image
It is a sensor. The condenser lens 6 is four reprojection lenses.
The image of the pupil mask placed on the front of the camera 8 is placed on the taking lens 1.
Form. The dotted circle drawn on the shooting lens in the figure is this
It is a projected image. The front of the condenser lens 6 has a cross shape
A field mask 2 having an opening is arranged, and this mask surface is
Planned focal plane of taking lens 1 equivalent to film plane of camera
It is in a good position. The reprojection lens 8 is an image of the surface of the field mask 2.
Are formed on the image sensor 10. With this configuration
One of the projection lenses 8 a'condenser lens 6
Therefore, the image on the lens 1 is a, and similarly the image of b ′ is b.
Then, I passed the area surrounded by the circle of A of the taking lens 1.
The image formed on the field mask 2 by the subject light is
Formed on the image sensor 10 by the projection lens
You. Similarly, the object that has passed through the area B on the shooting lens
The image formed on the field mask 2 by the body light is re-projected.
It is formed on the image sensor by the shadow lens B '.
In the cross-shaped rectangles on the image sensor 10, X1
Horizontal of the cross-shaped opening of the field mask 2 by the projection lens
X2 is the image of the area, and X2 is the field
It is an image of the horizontal part of the cross-shaped opening of Ku. Similarly for Y1 and Y2
And the field of view mass by a pair of reprojection lenses arranged in
It is an image of the vertical part of the cross-shaped opening of Ku. On X1 and X2
An image of the same part of the object is formed, but the subject is photographed
The image formed by the lens 1 is just formed on the visual field mask 2.
You. In other words, when the subject is in focus
Based on the position of the reprojected image on X1 and X2 of
Is closer to the shooting lens than the field mask (before
Pin), the re-projected images on X1, X2 approach each other,
When they are pins, they move away from each other. Therefore, those who connect X1 and X2
The line image sensor is placed in the direction to process on the video signal.
Operate the X signal for the image signal of the subject image on X1. ₂ Top cover
The image signals of the image are superimposed on each other by slightly shifting the image signals.
By detecting the amount of image shift that maximizes the correlation of
Which side of the subject image is from the correct focus position
It can be calculated whether or not it is approaching. The above is the present invention
It is the principle of focus detection, but regarding the processing operation of the video signal
Japanese Patent Application Laid-Open No. 60-24721
It is described in No. 0. In accordance with the above principle, X1, X
Along the direction of 2 and the direction of Y1 and Y2 orthogonal to it
And a line image sensor is arranged in each of them. In Fig. 2, the taking lens is not shown
After the transparent part in the center of the smirror, tilt it 45 ° and place it downward.
Infrared cut of the light transmitted through the shooting lens by the mirror placed
Towards filter 3, field mask 2 and condenser lens 6
And then turned horizontally by the 45 ° mirror 4.
Faced through the pupil mask 7 and re-projection lens 8 (2 to 4)
And is projected on the image sensor 10. 5 is all above
It is a frame that connects the elements of to one unit. Image sen
Sa is horizontal as described above (direction connecting X1 and X2)
Line image sensors are arranged in the vertical direction and
However, as a line image sensor, a CCD image
A di-sensor is used. CCD image sensor is a photodiode and its output
The capacitor that integrates the photocurrent is an element for one pixel.
The elements are arranged in an array, and each element
When the photocurrent integration is performed for an appropriate time all at once, the integration control is performed.
By applying a shift pulse to the gate,
A light amount signal based on the accumulated charge of each pixel is stored in the shift register.
To the shift register and then the transfer clock to the shift register.
By applying, the charge signals in the shift register are sequentially
Obtain a video signal by reading it as a voltage signal
It has become. FIG. 3 shows the CCD image in the embodiment of the present invention.
The circuit configuration around the charge sensor is shown. PD array I
~ PD array IV is the above-mentioned requirement for CCD image sensor.
PD array I is a prime array
Position, PD Array III is also in position X2, PD array
AII is located at Y1 and PD array IV is located at Y2
I have. The average brightness of the image projected on the PD array
Therefore, in order to determine the integration time of the photocurrent, the
At the same time, connect the monitor photodiode (PD) M1 to the PD
Along with Ray II, the monitor photodiode (PD) M2
It is arranged. G1 to G4 correspond to PD array I to PD array IV
An array of integration control gates to connect each element of the PD array.
There is a one-to-one correspondence. R1 and R2 are shift registers. Shi
The soft register R1 corresponds to PD arrays I and IV, and
Minute shift gates are applied to control gates G1 and G4.
The photocurrent integrated charge of each element in PD arrays I and IV.
Are transferred in parallel to the shift register R1. Integral control
Timing of shift pulse applied to roll gates G1 and G4
Gu is different. Shift register R2 is PD array II, III
And the charge signals in PD arrays II and III are transferred.
Be done. These shift registers are two-phase transfer clock
It is driven by the pulses φ1 and φ2 and stored in the same register.
Information is sequentially output. For convenience of explanation below, some words are decided. Rye
The direction of the image sensor is as shown in Fig. 1.
The x direction (horizontal) and the y direction (vertical) are determined. These two directions
Is the direction indicated by the arrows x and y in FIG. Total Contrast
Is the difference between adjacent difference data in the video signal.
Is the sum of the absolute values of the
The finer the darkness, the greater the total contrast.
Become. "LowCon" is an abbreviation for Low Confidence, which is the focus detection signal.
This means that the reliability is low. Below is the configuration of each part of the device and
The operation will be described in detail. (CCD image sensor circuit) In FIG. 3, CCD analog shift registers R1 and R2 are
Charge transfer is performed with two-phase clocks φ1 and φ2. Its output is
A voltage converter and a buffer are provided to store the PD arrays I and IV.
The product charge is transferred from the OSI pin via the analog shift register R1.
It is output and the accumulated charge of PD arrays II and III is analog.
It is output from the OSII pin via the soft register R2. See you
The output side of the Nita PD has the same configuration as the accumulated charge of the PD array.
And monitor (PD) via voltage converter and buffer
The accumulated charge of M1 is accumulated in the monitor (PD) M2 from the AGCOSI pin.
The electric charge is output from the AGCOSII pin. Also this monitor PD
PD is not connected due to the reference voltage output of
There is a voltage converter connected to the PD that is shielded from light.
The reference voltage DPS is output. This output is
It is used to control the timing of pulse generation. Integration control gates G1 and G3 are PD arrays in the x direction
A common shift pulse corresponding to I and III via the terminal SH1.
Space SH1 is applied. In the same way, the integration control game
G2 and G4 correspond to the PD arrays II and IV in the y direction, and the terminals SH2
The common shift pulse SH2 is applied via
ing. Also, each integration control gate G1 to G4 has no
It is also possible to apply shift pulses SH all at once via the child SH.
Is ready. Obtained from CCD image sensor
The video signal is suitable for focus detection regardless of subject brightness.
The integration time
Is the output of the monitor photodiodes (PD) M1, (PD) M2
Controlled by. Here, the x and y bands of the subject
The average brightness of the
The switches SH1 and SH2 can be applied separately. PD arrays I to IV are driven by the integration clear signal ICG pulse
Cleared all at once, and photocurrent integration started from that point.
You. Here, for example, the strip portion of the subject in the x direction is in the y direction.
When the average brightness is higher, shift pulse SH1 is output first
Then, the photocurrent integration signals of PD arrays I and III are integrated.
Intermediately held by the roll gates G1 and G3. Then PD
When the video signals of Rays II and IV reach appropriate values, the shift pulse
SH2 is emitted and the photocurrent integrated signals of PD arrays II and IV are multiplied.
Intermediate control gates G2 and G4 hold. So
After that, the shift pulse SH is applied to all the gates G1 to G4 all at once.
As a result, all video signals in the x and y directions are shifted
Data is transferred to R1 and R2. As described above, the integration controls G1 to G4 are PD array I
~ Hold the output of IV temporarily and store it in shift register R
It has a function to transfer to R1 and R2 in parallel.
The road structure is shown in FIG. Figure 4 shows the structure of one pixel
The charge photoelectrically converted by the PD array is passed through the barrier gate
Charged to almost the power supply level by the integral clear pulse ICG
Is stored in the first storage unit C1 via the barrier gate. This
The average brightness of the PD array columns of the
SH when the integrated signal reaches the proper integration level ₁ Or SH2 pal
Applied to each pixel, the charge of each pixel is transferred from C1 to C2 in parallel.
Will be sent. At this time, charge transfer due to the capacitance difference between V1, V2, C1 and C2
Is almost completely done. Thus, from the application of ICG pulse to S
Accumulated before application of Hn (n = 1 or 2) pulse
The charge is transferred from the storage unit C1 to C2, and even in this state
The accumulation of charge on the PD array on which the other image is projected is complete.
Wait for it to finish. No photocurrent is generated in this second storage unit C2.
And the amount of electric charge is substantially maintained. The other PD
When the array also completes the charge accumulation, the CCD image sensor
The charges of all pixels are stored in the second storage unit C2, which is suitable for focus detection calculation.
The bells are aligned. Next, the SH gate
Image to the analog shift register
The raw information is transferred in parallel at an appropriate level and then transferred.
This charge is output sequentially from the OSI and OSII pins in synchronization with the clock.
Be done. (Circuit for performing focus detection and focus adjustment) Next, in FIG. 5, the image sensor is driven to perform focus detection and focus adjustment.
The circuit configuration which performs adjustment is shown. 20 drives the image sensor 10 and inputs the information.
Performs focus detection calculation and drives the lens through the motor drive circuit 90.
Control is performed through the focus state display circuit 100.
It is a microcomputer for AF. Microcomputer for SF
The computer starts operation when the AF start switch SAFS is turned on.
You. 30 detects the monitor output AGCOS1 in the x direction,
A shift pattern that causes PD arrays I and III to complete integration.
Shift pulse generator circuit for generating the loose SH1, 31 is in the y direction
Monitor output of AGCOS2 is detected, PD array II, IV in y direction
Generates a shift pulse SH2 to complete integration
Shift pulse generating circuit. This circuit is shown in Figure 6.
It is composed of the circuit as shown. Reference voltage DOS is buffer
It is input to the circuit Buf1 and the output of that circuit is a resistor R31 and a constant current I31.
The voltage dropped by the constant voltage ΔV1 by
Applied to the (+) input of Com1. Of this comparator
The monitor output AGCOSn is applied to the (-) input. product
By applying minute clear pulse ICG, both outputs DOS and AGCOSn are equal
After that, the potential of AGCOSn becomes the electric charge in the monitor PD.
It decreases in proportion to the generated amount, that is, the amount of incident light. COMPA
Looking at the input level of the rectifier Com1, when ICG is applied
The (−) input is high by ΔV1, but decreases with charge accumulation,
Output of comparator when (-) input falls below (+) input
Is reversed. Focus at the average level of the video signal during this inversion
R31, I3 so that the proper focus detection result can be obtained by detecting
1 or ΔV1 is set. This time
The inverted signal of the COM1 switch is the flip-flop reset by the pulse ICG.
Flop FF31 is set and FF31 output inversion is AND31, INV3
1, converted to a pulse by the delay circuit 32, SHn (n = 1 or
2) is output as a signal. In addition, pulse ICG application
The time until this SHn signal is output becomes low brightness
Since a long time is required, set the maximum integration time
Microcomputer shift pulse SH when time passes
It is also possible to limit the integration time by generating
Has become. Special handling is required for these low brightness conditions.
It is equivalent to that explained in Kaisho 60-125817. Circuit 40 is a transfer clock generation circuit.
Dividing the basic clock supplied from the computer, φ1,
Generate φ2 pulse. The transfer clock frequency is applied to the Sφ terminal.
Is the signal for switching the wave number a microcomputer 20?
This signal is supplied to the output of both x and y directions.
Is high, and only the output in one of the x and y directions is output.
When inputting, set Sφ signal to Low and transfer clock frequency
By inputting the number as a double of the above-mentioned two-way output input
We are trying to reduce the charge transfer time. In addition, the above-mentioned second
Of the charge transfer from the storage means C2 to the analog shift register
The SH signal is not being input because it is necessary to synchronize
You. 50 and 51 are analog processing circuits for each pixel output OSI and OSII
The basic structure is shown in FIG. Each pixel output is a differential amplifier Am
At p51, it is output as the difference from the reference voltage V52. This
Is the maximum output of each photodiode array PDI to IV.
The dark output signal of the aluminum light-shielded pixel provided in the front is output.
SP1 output by the microcomputer 20 when
Is sampled by the signal of SP2 and held by C51.
The difference between the optical output and the optical output that will be output thereafter is determined by the differential amplifier Amp52.
By extracting the light output component only. Here, sample and hold the dark output with each PD array I-IV
The PD arrays I, III and II, IV have different integration times.
This is because it is controlled and a difference occurs in the dark output voltage.
The pixel output from which only the light component is extracted is sampled next.
After being input to the hold circuits 60 and 61, it is input to the multiplexer 70.
Is entered. Where the multiplexer 70 is the sample hole
Input data selection of one of the output pixel outputs I1 and I2
Select by zone signal SZ and select D1 terminal in A / D conversion circuit 80
Output from As mentioned above, the microcomputer x
y When inputting data in both directions, output Sφ signal Hi
The transfer clock is generated at normal speed and AND2 and OR1 are set.
Output switching of multiplexer 70 via transfer clock φ
Switch in synchronization with 1. This result timing chart No.
8 CCD shift register R1, CCD shifter
The output signal of the register R2 is output alternately, and the A / D conversion
Digitized on channel 80 and input to the microcomputer
You. On the other hand, when inputting only the x direction or the y direction,
By setting the Sφ signal to Low and one input of AND2 to Low,
The output switching of the multiplexer is selected by the microcomputer.
It depends on the selection signal SZ. Also at this time CCD image set
The transfer lock frequency of the sensor becomes double speed. Micro computer
The computer outputs the standard part in the x direction and the reference part in the y direction.
Input, then the SZ signal is inverted and the CCD shift register I, I
Output of I Input of reference part in x direction Input of standard part in y direction
To do. By fully utilizing the A / D conversion time like this
At this time, the data transfer time can be shortened.
The imming is shown in FIG. 8 (b). (Automatic Focus Detection Operation) In the present invention, there are several operations in automatic focus detection.
Production mode is possible. Of operation in these modes
Specific examples are shown in FIGS. 9 to 10. Figure 9 shows x
The focus detection calculation is performed in the y-direction and the y-direction.
Focus detection results for directions determined to be closer to the camera
This is a flow of driving the lens based on the result. Figure 10 shows x,
y Compare the total contrast in both directions, and
The lens is driven by performing focus detection calculation preferentially in the direction
Only when it becomes LowCon, the other focus detection calculation is performed.
This is a flow for driving a lens. Where the contrast value
Is a large LowCon as a LowCon discrimination criterion.
As proposed by the applicant in Kaisho 60-247210,
The evaluation function YM (XN) / CN by the arithmetic operation is less than a predetermined value.
Because it is also a condition, like a perspective competing subject
In such cases, the evaluation function may be significantly deteriorated.
In Fig. 11, focus adjustment is performed with priority given to the focus detection function in the x direction.
Perform focus check in y direction only when x direction becomes LowCon
An example of utilizing the output function will be shown. First, FIG. 9 will be described. When the AF switch SAFS turns on, the microcomputer
20 is activated. First of all, the microcomputer
Reset the image sensor. This can be done before
Pre-stored in register and photoelectric conversion unit while transmission clock is stopped
It is necessary to perform this once at startup because it is necessary to discharge unnecessary charges.
There is a point. Next, the microcomputer 20 is the CCD image sensor 10
ICG pulse is supplied to and the integration starts. Mark of this ICG pulse
In addition, the image sensor discharges the accumulated charge of each pixel.
At the same time, the accumulated charge of the monitor output is also discharged,
As the charge disappears, both of them start to accumulate the generated charge.
You. After that, the microcomputer has both TINT1 and TINT2 pins.
Inversion, PD array I and III, PD array II and IV
When the average of the elementary accumulated charges reaches a preset level, the shift
Pulse SH1 and SH2 are generated and PD is applied to the second storage unit C2 of each pixel.
The accumulated charge of the arrays I and III is transferred to the PD of the second accumulation part C2 of each pixel.
Wait for the charges accumulated in arrays II and IV to be completely transferred. this
When the microcomputer detects completion, the microcomputer
The computer generates an SH pulse and stores the PD array I and IV
Load the analog shift register (CCD register) R1 and PD
The charges accumulated in arrays II and III are stored in an analog shift register (CC
D register) Transfer to R2 in parallel. After that, in synchronization with the transfer clock,
The pixel signal is output, and the microcomputer is
Count A / D conversion completion signal A / DEOC of one pixel signal
By knowing the number of output pixels, it is also installed in each PD array I to IV.
The dark output sample signals SP1 and SP2 of the aluminum shielded pixels
A / D of each light output pixel that is output and subsequently output
Image information is obtained by sequentially inputting conversion values. This Taimi
Will be described later. In this way, the focus detection calculation required
Digital information of all pixel output in a microcomputer
Once stored, the microcomputer will perform focus detection calculation.
To start. First, the microcomputer is the correlation in the x direction.
Calculate. First, x-direction difference data is created. this
Difference data Ux (k) = Sx (k) −Sx (k + 4), Vx
(K) = Tx-Tx (k + 4) Every fourth raw data
Data difference. This is a low frequency component where focus detection cannot be calculated.
This is to cut the minutes. Difference data of standard part and reference part
When 27 pieces and 35 pieces are gathered, the microcomputer slides into an image.
While increasing the pitch amount by one pitch, each image shift
The correlation value YM (l) is obtained by the amount. Also the total contrast
A value, that is, the sum of adjacent data of the difference data is calculated. This
The highest correlation among the correlation values YM (l) obtained in this way
The image shift amount, that is, the value of the correlation value YM (l) is minimized.
Find lx. This lx is related to the lens defocus amount.
The system used here is more precise.
In order to obtain the degree, the correlation value and
An interpolation calculation is performed using the correlation value in quantity. For this interpolation calculation
This is explained in detail in JP-A-60-247211 by the applicant.
Please refer to it. In this way, a precise and detailed image
The shift amount XM and the correlation evaluation function YM (XM) / CX are obtained. this
Correlation evaluation function YM (XM) / CX, total contrast value CX, and total
LowCon is discriminated at three points of the output raw data value. This LowCon
Discrimination is also explained in JP-A-60-247210.
Therefore, it is omitted here. The phase at both ends of lx = 1 or 9
If the function is the minimum, use LowCon. When it is judged that it is not LowCon, XM-5 as PX value,
That is, the difference from the image shift amount at the time of focusing is calculated and stored.
Also, the image of the correlation calculation in the y direction is stored by setting lmin = lx-1.
Limit the range of displacement. On the other hand, if it is determined to be LowCon
In this case, the image shift range of the correlation calculation in the y direction is not set,
Correlation calculation is performed over the circle. Thus the correlation calculation in the y direction
After memorizing the range limits, the difference data in the y direction is calculated in the x direction.
Create it in the same way as for Muko. Uy (k) = Sy (k) −
Sy (k + 4), Vy (k) = Ty (k) −Ty (k + 4)
Same as the one obtained in the x direction based on the difference data created by
Similarly, the correlation value is calculated. However, here the correlation in the x direction
Only for image shift amount of lmin or more for which the calculation result is obtained
Perform correlation calculation. This is because the closer the subject is to
Since the image spacing of is large, the correlation in both x and y directions
The one with the larger image shift amount of the calculation result is selected. Therefore, first
Correlation calculation is performed only on a portion larger than the obtained image shift amount in the x direction.
Is sufficient, and the correlation calculation can be shortened. like this
Then, the correlation value is obtained also in the y direction, and the highest correlation value is obtained.
Calculate the part that is Next, perform the correlation calculation as in the x direction.
Then, as the result of the correlation calculation in the y direction, the image shift amount xN, the correlation evaluation
Calculate the function YM (XN) / Cy. This evaluation function YM (XN) / C
Total contrast values in y and y directions Cy and y direction raw data peak
And the both ends of the correlation calculation image shift amount, lmin,
9 points of whether there is no calculated minimum correlation value image interval lx1
If it is not LowCon, it is judged as PY.
The image shift amount xN-5 from the focus state is stored. In this way, when the correlation calculation is completed for both x and y directions,
The icrocomputer is the result of both correlations in both x and y directions.
The lens is driven by. First, if it is determined to be LowCon in both directions,
Microcomputer drives the lens and contrast
To find the lens position that can be detected (LowCon Scan)
I do. This operation will make the entire lens drive range at least once.
Stop the lens drive after scanning and
Only the focus detection calculation is repeated at
In this mode, the lens drive is restarted in the detected state. Required x if at least one is not LowCon
The magnitude of the image shift amount in the direction is compared.
This is used as the image interval amount used for driving the lens later. here
At LowCon, the value of Px to Py is set to Min value (-4)
It has been done. Despite the limitation when calculating in the y direction,
The amount of image deviation is quite large within one pitch.
The differential focus has a differential focus amount and a value obtained by interpolation calculation.
This is because there is a large difference in quantity. This calculated image shift amount P is calculated as the differential focus amount, and the
The lens drive amount conversion coefficient
The lens drive amount is calculated by, and focus determination is performed. Lens drive
When the amount is extremely small and there is no need to drive the lens,
Indicates the focus, and if not, the lens drive amount
Follow the steps below to drive the lens and perform focus detection again.
The image sensor is re-integrated for this purpose. Next, using the flowchart in Figure 10, the total contrast
Priority is given to the correlation calculation for the first direction of quantity.
Shadows such as when the lens is driven and its direction is a perspective competing subject
Only when the sound enters the LowCon state due to the sound
When the output correlation calculation is performed and the lens is driven according to the result.
The operation of the focus detection device will be described. It starts before the operation and completes the data input.
As in the case of Fig. 9 described above, the image in both x and y directions
The sensor data is stored in the microcomputer
You. The microcomputer first obtains the difference data in the x direction.
Created in the same way as in the case of FIG. 9, and sum of adjacent differences of difference data
Calculate the x-direction total contrast value Cx by obtaining
You. Subsequently, the difference data in the y direction is created, and similarly, the y direction
The total contrast value Cy is calculated. Thus in both x and y directions
After calculating the total contrast values Cx and Cy, the magnitude ratio of the two
Make a comparison. Here, the phase in the direction in which the total contrast value becomes large
The result of the function calculation is for the direction in which the total contrast value decreases.
It is usually considered to be more reliable than all correlation calculation results. Therefore, priority is given to the direction Z in which the total contrast value is large.
Correlation calculation is performed. This correlation calculation is shown in Fig. 9.
The same method as shown is used. Of highly correlated parts
Image shift amount xM, correlation evaluation function YM (X
Calculate M) / CZ. LowCon discrimination is performed using this result.
If it is determined that it is not LowCon, this image shift amount XM
The image shift amount from the time of focusing is calculated by
And the amount of lens drive are calculated. If it is determined whether or not it is in focus,
When focusing, the focus is displayed, and when out of focus, the lens drive amount
Therefore, the lens is driven. On the other hand, if it is determined to be LowCon, this time in the opposite direction.
Correlation calculation is performed. This result shows that there is a high correlation
Minute calculation and interpolation calculation to calculate the image shift amount XM, YM (XM) / CZ
Calculate. Using this result again, LowCon discrimination is performed and LoCon
If it is determined that it is not wCon, this image shift amount XM
The image shift amount from the time of focusing is calculated, and the differential focus amount,
The lens drive amount is calculated. When in focus by determining whether it is in focus
Is displayed in focus, and when out of focus, it follows the lens drive amount.
Drive the lens. Also, when it is determined to be LowCon, this time both x and y directions
Is determined to be LowCon, and the above-mentioned LowCon Scan is performed.
U. Frequent placement of high contrast areas at the end
Focusing calculation is performed by giving priority to the x direction (horizontal direction)
Focus detection in the y direction only when the direction is determined to be LowCon
Focus detection device that performs calculations Flowchart in Fig. 11
It will be described with reference to FIG. In this focus detection device, the above-mentioned FIG.
And unlike the two examples in Fig. 10, only the data in the x direction has priority.
Type in. Measures reduction of integration time and data transfer time.
In order to improve the responsiveness of the system, CC after AF operation is started.
When the initialization of D is completed,
The transfer clock frequency is set to double speed in the above example
Output Sφ = Low and output at high speed in one direction only.
It is carried out in the form as shown in (b). Microphone after applying an integration clear pulse to start integration
(B) The computer uses the TINT1 signal that indicates the completion of integration in the x direction.
Has an inverted signal. When detecting the inversion of the TINT1 signal, the microphone
The computer does not depend on whether the integration in the y direction is completed or not.
Shift pulse is generated and the pixel output data in the x direction
Start typing. First, OS1 outputs PD array I, x direction reference pixel output
Output, SZ = Hi is output, and the multiplexer 70 D1 output
By setting the force as I1 signal, that is, OS1 processed signal,
A / D conversion of the reference pixel output to a microcomputer
input. During this time, the output of the y-direction reference part output from OS1
Is ignored. When the input of this x direction reference part is completed,
Microcomputer outputs Sz = Low and multiplexer 7
D1 output I2 signal of 0, that is, OS2 processing signal of image sensor
By this, the output of the reference part in the x direction is A / D converted and input
To do. In this way, the input of the standard part and the reference part in the x direction is completed.
Then, the difference between these data is the same as the above two examples.
Data creation, correlation value calculation, extraction of the highest correlation, interpolation calculation,
Perform LowCon judgment. LowCon similar to that shown in the above two examples
As a result of judgment, the amount of image deviation obtained is reliable instead of LowCon
If it is determined that the data is high,
Calculate the difference from the focus image shift amount = XM-5 Defocus amount PF,
The lens drive amount is calculated and the lens drive amount is extremely small.
When it is out of focus, the focus is displayed and calculated otherwise.
The lens is driven according to the specified lens drive amount, and the x direction
The re-integration and refocus detection calculation is repeated for each pixel of. Meanwhile, the focus detection calculation result in the x direction is determined to be LowCon.
If so, the microcomputer then moves in the y direction.
The integration of the image sensor and the focus detection calculation are started. My
The black computer uses S to input data in one direction.
Generates an integration clear signal ICG while outputting φ = Low
After that, it waits for the inversion of the integration completion signal TINT2 in the y direction. This TIN
When the inversion of T2 is detected, this time the integration in the x direction is completed or not completed.
SH pulse is generated and input of y direction data is started regardless
To do. At this time, contrary to the above-mentioned x direction input, first SZ = Low
Is output and the x-direction reference part output is ignored, while the y-direction reference is output.
I2 is passed through the multiplexer only for the partial output and A / D converted.
Data input, and when this is completed, SZ = Hi is output and x direction
I1 only reference output in the y direction while ignoring reference output
Through a multiplexer to input data after A / D conversion.
U. After entering all the data in the y direction,
Ikuro computer creates difference data in y direction, correlation value
Calculation, extraction of highest correlation, interpolation calculation, LowCon discrimination in x direction
The procedure is the same as that for. LowCon judgment
As a result, the required image shift amount is not reliable for LowCon.
If it is determined that the focus image is
Calculation of difference P = xM-5, differential focus amount DF, lens
The drive amount is calculated, and when the lens drive amount is extremely small,
The focus is determined and the focus is displayed. Otherwise, it is calculated.
The lens is driven according to the lens drive amount, and the
Focus detection operation is not performed at all for each pixel in the y direction.
The re-integration and re-focus detection calculation are repeated. On the other hand, if it is determined to be LowCon in the y direction as well,
In this case, it is judged as LowCon in both x and y directions.
While performing Con Scan, move the image sensor in both x and y directions.
Repeated re-integration and refocus detection are repeated, and LowCon disappears
Wait for the situation.

[0007]

【effect】

According to the present invention, since the plurality of line image sensors are arranged in the horizontal direction and the vertical direction, it is possible to always perform reliable focus detection even if the contrast distribution of the subject has directionality. Further, since the line image sensor is used, the cost of the device is lower than that when a sensor having a two-dimensional resolution is used. Furthermore, since the subject actually photographed changes more in the horizontal direction than in the vertical direction, and the human eye is more sensitive to the change in the horizontal direction, the focus adjustment operation is performed with priority on the line image sensor arranged in the horizontal direction. By doing so, the better focus adjustment operation is performed. Therefore, the present invention enables highly accurate and reliable focus adjustment.

[Brief description of drawings]

FIG. 1 is a perspective view showing an optical configuration of the present invention.

FIG. 2 is an exploded perspective view of an optical portion according to an embodiment.

FIG. 3 is a circuit diagram around the image sensor in the same embodiment.

FIG. 4 is a circuit diagram of an integration control gate in the same embodiment.

FIG. 5 is a circuit diagram of a portion for performing image sensor driving, focus detection, and focus adjustment in the embodiment.

FIG. 6 is a partial detailed circuit diagram of the above circuit.

FIG. 7 is a detailed circuit diagram of another part of the same.

FIG. 8 is a time chart of the operation of the above circuit

FIG. 9 is a flowchart of one operation mode in the above embodiment.

FIG. 10 is a flowchart of another operation mode of the same.

FIG. 11 is a flowchart of another operation mode of the same.

[Explanation of symbols]

 1 …… Shooting lens 2 …… Field mask 6 …… Condenser lens 8 …… Reprojection lens (2 to 4 pieces) 10 …… Image sensor

Front page continuation (72) Inventor Toshio Morita 2-30 Azuchi-cho, Higashi-ku, Osaka City Osaka International Building Minolta Camera Co., Ltd. (72) Nobuyuki Taniguchi 2-30 Azuchi-cho, Higashi-ku, Osaka Osaka International Building Minolta Camera Co., Ltd. (56) Reference JP-A-62-95511 (JP, A) JP-A-52-55658 (JP, A)

Claims (3)

(57) [Claims] 1. An automatic focus adjusting device for automatically adjusting the focus of a taking lens, wherein optical means for forming a subject image by light from a subject and horizontal and vertical arrangements are formed by said optical means. A plurality of line image sensors that receive the subject image and a focus state for each array direction of each line image sensor based on the video signal output from each of these line image sensors,
Focusing means for driving the photographing lens based on the focus state and control means for controlling the focus adjusting operation by giving priority to the line image sensor arranged horizontally in the line image sensors. An automatic focusing device characterized by comprising: 2. The control means detects the output of a line image sensor vertically arranged when the reliability of the focus state detected by the output of a line image sensor horizontally arranged is low. The automatic focusing apparatus according to claim 1, wherein the photographing lens is controlled to be driven based on a focus state. 3. The control means first detects the focus state by the output of the line image sensor arranged in the horizontal direction, and when the reliability of the focus state is high, the line image sensor arranged in the vertical direction. Claim 1 wherein the focus state is not detected by the output of
An automatic focusing device according to the item.