JPH09105857A - Photoelectric converter for focus detection - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a photoelectric converter for focus detection which can have a wider area where pupil eclipse does not occur by arranging a photoelectric converting element for focus detection off an optical axis and along a straight line passing the center of the optical axis. SOLUTION: A standard part and a reference part are provided for respective islands IS1-IS7 and there is the center of the optical axis in the center between the standard part and reference part. Then the photoelectric converting device for focus detection which is used for TTL type automatic focusing has the photoelectric converting elements IS2 and IS3 for focus detection arranged off the optical axis along the straight line passing the optical axis center. The photoelectric converting elements IS2 and IS3 are thus arranged, so a larger area where there is no pupil eclipse can be obtained as compared with a case wherein the area is arranged off the axis vertically. Consequently, the photoelectric converting device which is advantageous to pupil eclipse can be obtained.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a photoelectric conversion device for focus detection of a camera, and more particularly to a photoelectric conversion device for focus detection of a camera having a plurality of monitors.

[0002]

2. Description of the Related Art Conventionally, line sensors have been arranged in parallel in an off-axis direction in a vertical direction in order to widen a region in which pupil eclipse does not occur.

[0003]

The line sensor for focus detection used in the conventional camera is provided as described above, but the conventional method has a problem in that the area where the pupil is not dilated cannot be taken too wide. there were.

The present invention has been made to solve the above problems, and an object of the present invention is to provide a photoelectric conversion device for focus detection which can take a wider area in which pupil fringing does not occur.

[0005]

SUMMARY OF THE INVENTION TTL according to the present invention
In the focus detection photoelectric conversion device used for the system autofocus, the focus detection photoelectric conversion element is not on the optical axis and is arranged along a straight line passing through the center of the optical axis.

Since the photoelectric conversion element for focus detection is arranged other than the optical axis and along the center of the optical axis, a larger area can be obtained as compared with the case where the area where pupil eclipse does not occur is arranged vertically off the axis. As a result, it is possible to provide a photoelectric conversion device for focus detection, which is more advantageous for pupil blurring.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A focus detecting optical system in a single-lens reflex camera with an automatic focus detecting function using a focus detecting photoelectric conversion device according to the present invention will be described with reference to FIGS. The main body of the single-lens reflex camera includes a taking lens 11 provided on the optical axis 10 and a main mirror 12 provided behind the taking lens 11 at an angle of 45 ° upward. A film exposure surface 13 is provided behind the main mirror 12. The photographing light flux that has passed through the photographing lens 11 is reflected upward by the main mirror 12, is focused on the focusing screen, and is guided to the finder optical system via the pentaprism.

A half mirror is formed on at least a part of the main mirror 12. Between the half mirror portion and the film exposure surface 13, a sub mirror 14 having a rotation shaft pivotally attached to the back surface of the main mirror 12 is provided at a downward angle of 45 °. The focus detection light flux that has passed through the half mirror portion of the main mirror 12 is reflected downward by the sub mirror 14 and is guided to the focus detection device 15 arranged below the mirror box of the camera body.

At the time of photographing, the main mirror 12 and the sub mirror 1
Reference numeral 4 is rotated forward and upward and is retracted from the optical axis 10, and the photographing light flux passing through the photographing lens 11 is exposed on the film exposure surface 1.
An image is formed on the film exposure surface 3, and an image is exposed on the film exposure surface 13.

The focus detection device 15 includes seven photoelectric conversion element arrays 16a, 16b, 16c, 16d, 16e, 16f,
It includes an AF sensor 17 with 16g. Of the photoelectric conversion element rows 16a to 16g, three photoelectric element rows 16a to 16
c is arranged in a horizontal position including the optical axis 10. The four photoelectric conversion element rows 16d to 16g are the photoelectric conversion element rows 16a to 16g.
It is oriented at about 90 ° with respect to 16c.

A separator lens plate 18 is provided in front of the AF sensor 17. The separator lens plate 18 includes
Separator lenses 18a to 18g corresponding to the photoelectric conversion element arrays 16a to 16g are integrally formed. A diaphragm mask 19 is provided immediately in front of the separator lens plate 18, and the diaphragm mask 19 includes separator lenses 18 a to 1 a.
Openings 19a to 19g corresponding to 8g are formed.
A reflection mirror 20 is provided to face the aperture mask 19 and the sub mirror 14. The focus detection light flux reflected downwardly by the reflection mirror 20 by the sub mirror 14 is passed through the aperture mask opening 19
a to 19 g and the separator lenses 18 a to 18 g to lead to the photoelectric conversion element arrays 16 a to 16 g. Reflection mirror 20
And the sub-mirror 14 between the aperture mask openings 19a-1.
Condenser lenses 21a to 21g facing 9g are provided. On the upper surface of the condenser lenses 21a to 21g,
A visual field mask 22 is provided. The field mask 22 is
Photoelectric conversion element array 1 in which the position and direction of the light beam for focus detection are different
Opening 2 for separating so as to correspond to 6a to 16g
2a to 22g.

The principle of focus detection is the TTL phase difference detection method. Different regions 1 on the exit pupil plane of the taking lens 11
Reference part light flux a passing through 1a and 11b and 11c and 11d
The reference portion light flux b (shown by the broken line in FIG. 2) and the reference portion light flux b (shown by the solid line in FIG. 2) are respectively received by the standard portion A and the reference portion B in each of the photoelectric conversion element arrays 16a to 16g. The light distribution pattern of the image formed by the received light is converted into an electric signal in each of the photoelectric conversion element arrays 16a to 16g, and their correlation is obtained by a correlator (not shown). Focus adjustment is performed by moving the taking lens 11 back and forth by the drive mechanism based on the shift signal of the correlator 7.

In the focus detection optical system of FIG. 1, in addition to the photoelectric conversion element arrays 16a to 16c in the horizontal position, photoelectric conversion element arrays 16d to 16g in the vertical position are provided. Therefore, the horizontal and vertical focus detections can be performed at the same time, so that the focus detection of a horizontal line or the like can be performed.

FIG. 3 is a diagram showing a display of a viewfinder of a camera using an AF sensor to which the photoelectric conversion device according to this embodiment is applied. Focus detection of an object in regions IS1 to IS7 (hereinafter, referred to as first to seventh islands) of a graph indicated by a solid line at the center of the shooting screen S is performed. A rectangular frame AF indicated by a dotted line is displayed in order to show the photographer the area where focus detection is performed. A display portion Lb shown outside the photographing screen S indicates a focus detection state, and is turned on when focusing.

FIG. 4A is a diagram showing the light receiving portion of the CCD image pickup device array on the AF sensor 17 used in this focus detection device. As shown in the figure, the CCD image sensor array includes seven line sensors, each of which includes a light receiving unit, a storage unit, and a transfer unit. A standard part and a reference part are provided for each of the islands IS1 to IS7 in FIG. 3, and the optical axis center is located at the center of the standard part and the reference part. A plurality of monitor light-receiving elements for controlling the integration time of the CCD in the storage portion is provided for each island on one side in the longitudinal direction of the reference portion of each island. I will describe.

The seventh island IS7 is a pixel with a red filter. As will be described later with reference to FIG. 29, it is used by switching to the sixth island IS6.

The seventh island is a pixel with a red filter in order to measure the distance only with the auxiliary light component (infrared light) without being affected by outside light when using the auxiliary light. Therefore, in the past, distance measurement could not be performed on an object having a low contrast such as a white wall under external light, but the present invention solves such a conventional problem. That is, in this system, only the auxiliary light component is extracted without being affected by external light and the camera is focused on the contrast pattern. Therefore, the distance measurement performance for a low-contrast subject is significantly improved. The specific content will be described later.

In the focus detection apparatus of this embodiment, the above-mentioned seven islands of the reference portion are divided into a plurality of blocks,
Focus detection is performed by comparing each of the divided blocks of the standard portion with the reference portion.

The divided range and the defocus range of the divided range are shown in FIGS. 5, 6 and 4B. FIG. 5 is an enlarged view of the focus detection area on the photographing screen shown in FIG. Each of the islands IS1 to IS7 for focus detection corresponds to the area of the reference portion shown in FIG. In FIG. 5, the numerical value shown in each island shows the data number obtained by the CCD shown in FIG. The number of data (X ', Y') in the standard portion and the reference portion in each island is equal to the number of islands IS1 and IS.
It is (30, 40) for S4, IS6, IS7, and (40, 48) for the islands IS2, IS3, IS5. Each island is divided. Island is
1, IS4, IS6 and IS7 are divided into two blocks. For example, the island IS1 includes two blocks BL1 and BL2, where BL1 is a data number (1
˜20), and BL2 includes (11-30). The islands IS2, IS3 and IS5 are each divided into three blocks. For example, island IS2 is BL3,
It is divided into BL4 and BL5. BL3 is the data (1-2
0), BL4 is (11 to 30), BL5 is (21 to 30).
40).

In the focus detection of this phase difference detection method, the optical axis center is present between the standard part and the reference part, and the image interval when the images of the standard part and the reference part match is larger than a predetermined interval. It is determined that the subject is in focus at a predetermined distance with a rear focus sometimes, and a front focus when the focus is small. Therefore, the defocusing range of the divided blocks is assigned to the rear pin side as the block is farther from the optical axis center in each island. The reason for this will be specifically described with reference to FIG. FIG.
(B) shows the standard part and the reference part of the island IS2.
Consider the defocus range of the fourth block BL4 divided into blocks. At this time, the focusing range includes the 15th to 34th (BL4 ′) images from the left end in the reference portion and the fourth
This is when the image of the block BL4 matches. As a result, if the images match on the left side of the reference portion, it is determined to be the front focus.
At this time, the maximum number of pixels in front pin deviation (called deviation pitch)
Is 14. When the image coincidence is on the right side of the reference portion with respect to the illustrated position, the rear pin is formed, and at this time, the maximum rear pin shift pitch is 14. The same applies to the defocus range divided into blocks on each of the other islands.

This state is shown in FIG. Third block BL3, ninth block BL9, seventeenth block BL17
Then, the front pin side displacement pitch is 4 and the rear pin side displacement pitch is 24. 4th block BL4, 10th block BL
In the tenth and eighteenth blocks BL18, both the front pin side displacement pitch and the rear pin side displacement pitch are 14. Fifth block BL5, eleventh block BL11, nineteenth block BL
In FIG. 19, the front pin side displacement pitch is 24 and the rear pin side displacement pitch is 4. First block BL1, 15th block BL15, 23rd block BL23, 25th block B
In L25, the front pin side displacement pitch is 5 and the rear pin side displacement pitch is 15. Second block BL2, 16th BL1
6, 24th block BL24, 26th block BL26
Then, the front pitch side deviation pitch is 15, and the rear pin side deviation pitch is 5. 6th block BL6, 12th block BL
In the 12th and 20th blocks BL20, both the rear pin side and the front pin side have 4 pitches. In the seventh block BL7, the thirteenth block BL13, and the twenty-first block BL21, the front pins have 4 pitches, the rear pins have 14 pitches, and the eighth block B has
L8, 14th block BL14, 22nd block BL2
In FIG. 2, the front pin has 14 pitches and the rear pin has 4 pitches. However, since calculation is performed while avoiding overlap with the sixth block BL6, the seventh block BL7 has 4 to 14 pitches on the rear pin side, and 8th pitch. The block BL8 has a pitch of 4 to 14 on the front pin side.

FIG. 7 shows the photoelectric conversion device of the present invention as a camera focus.
A rough outline of the AF system when used in a point detection device
It is a lock figure. Referring to FIG. 7, the controller is an AF sensor.
17 and the AF controller 30 and their peripheral circuits.
No. The AF controller 30 is a one-chip microcomputer
The AF sensor 17 has an analog output line.
V_outThe analog signal obtained from the
A / D conversion unit 31 for conversion and a memory formed by RAM
And part 32. The memory unit 32 includes a photographing lens (replacement lens).
Lens data output section 40 including a ROM
Defocus amount that differs for each lens-lens extension amount
Conversion coefficient K_L, Color temperature defocus amount dL _LData such as
Is input in advance and the digital data from the A / D conversion unit 31 is input.
Data is stored one by one. The AF controller 30 is
Correction calculation for calculating the correction amount based on the output of the memory unit 32
Based on the correction amount calculated by the correction unit 34 and the correction calculation unit 34.
The focus detection unit 33 that detects the focus and the detection output of the focus detection unit
Lens drive signal to drive the lens based on force
The lens drive controller 35 for sending to the circuit 42 and the AF sensor
Integral value at SA 17
Call) will reach a specified value within a specified time.
AF circuit 1 and timer circuit 36 for timekeeping for viewing
7 and an AF sensor control unit 37 that sends and receives signals.
No. The lens drive control unit 35 controls the lens movement status data.
Data from the encoder 44. Motor for driving the lens
43 is connected to the lens drive circuit 42, and the display circuit 41 is
It is controlled by the AF controller 30. AF sensor
17 and AF controller 30 each have one chip
It is formed separately. Therefore, the AF system is
It consists of a total of 2 chips. Analog reference voltage V_REF
Is the A / D converter 31 of the AF controller 30 and the AF sensor.
Given to SA17. The AF sensor 17 has a power supply line V
_CCAnd ground line GND.

Nine signal lines MD1, MD2, MD3, IS are provided between the AF sensor 17 and the AF controller 30.
They are connected by T, CBG, RST, CP, V _out and ADT. The mode terminal MD1 forming the signal line is a terminal for inputting a signal for switching the integration and reading modes. The mode terminal MD2 is a terminal for inputting a signal for selecting a read pixel column, a signal for selecting a monitor, and a signal for ST and PD integration selection described later and outputting integration end information 1. is there. The mode terminal MD3 sends out a signal for selecting a read pixel column, a signal for selecting a monitor, a read speed switching input and a signal for CP stop, and the integration end information 2
Is a terminal for outputting.

The integration start signal input terminal IST is a terminal for inputting an integration start signal and a signal for controlling the output mask.

In this embodiment, the read start signal input terminal RST
Is a terminal having an input / output switching function, and is a terminal for inputting a read start signal and outputting a gain information signal. The ADT output signal terminal ADT is a terminal for outputting a timing signal and integration end signal for pixel data output. The integration end terminal CBG is a terminal for inputting a forced end input signal, a barrier gate control signal, and a black level clamp signal to the CCD forming the photoelectric conversion device. In another embodiment described later, the terminal RST
And CBG have different functions. The signal output terminal V _out is a terminal for outputting a signal from each island.
The clock pulse input terminal CP is a terminal for inputting a clock pulse. Hereinafter, the operation of each terminal will be described more specifically.

The mode terminals MD1, MD2 and MD3 are AF
It is a signal line that outputs a logic signal from the controller 30 to the AF sensor 17, and sets the operation mode of the AF sensor 17. The operation mode of the AF sensor 17 includes three modes of a low brightness integration mode, a high brightness integration mode, and a data dump mode. Signal lines MD1, MD2, MD
The operation mode is set by a combination of 3 logic levels. The basic clock supplied from the signal line CP is outputting pixel data (when the signal line MD ₁ is High).
(when h and IST are High), the signal line MD3 is set to H
ON / OFF inside the AF sensor 17 by setting to high
F control is possible, and by turning off this basic clock, the operation of the AF sensor 17 is temporarily frozen. The AF controller 30 can also control other circuit parts, for example, the lens drive circuit 42 and the like. The signal line ADT indicates completion of output of one pixel data of the AF sensor 17 in the data dump mode,
An ADT signal for instructing the start of A / D conversion is supplied to the A / D conversion unit 31 in the AF controller 30. In other modes, the AF sensor 17 is used at the point where the charge is accumulated to an appropriate level in each island of the AF sensor 17.
Outputs an interrupt signal to the AF controller 30 to indicate the completion of integration. The signal output terminal V _out is an analog signal line, and after analog signal processing of the outputs of the photoelectric conversion element arrays 16a to 16g in the AF sensor 17,
It is supplied from the F sensor 17 to the A / D conversion unit 31 in the AF controller 30. The signal output terminal V _out signal is output for each pixel in synchronization with the above-mentioned ADT signal, A / D converted, and then taken into the AF controller 30 as subject image information obtained from the AF sensor 17.

Next, a specific structure of the AF sensor 17 will be described with reference to FIG. In the figure, the photoelectric conversion element array 16a to
16g shows the I / O part with the AF controller 30 on the right side. The photoelectric conversion element arrays 16a to 16g include seven islands IS1 to IS7 arranged in a shape as shown in the in-finder display of FIG. 3 and are controlled individually in principle.

The AF sensor incorporating the focus detecting photoelectric conversion device according to the present invention has two kinds of integration modes (high brightness integration mode and low brightness integration mode) according to the brightness of the object. FIG. 8 is a diagram showing a supply potential in each part of the photoelectric conversion element in the high brightness integration (hereinafter referred to as ST integration) mode.

(A) shows a state during integration operation (during charge accumulation). The potential of the barrier gate BG is set to High (hereinafter abbreviated as “H”) so that the charges generated by photoelectric conversion in the photodiode PD flow into the storage unit ST. The barrier gate BG is opened and the supply potential of the storage portion ST is set to H.
To make it easier for the electric charge to accumulate. At this time, the potential of the shift gate SH is set to Low (hereinafter abbreviated as “L”),
The charge to be stored in the storage unit ST is the shift register SR.
I try not to spill it.

(B) shows the operation of ending the integration. The potential of the barrier gate BG is set to L, the barrier gate BG is closed, and the charge generated in the photodiode PD is stored in the storage portion ST.
Is prevented from flowing into. The supply potential of the storage unit ST is set to L, and the generation of dark current in the storage unit ST is suppressed. In this state, the pixel data is held for some time.

(C) shows the state at the start of reading. (B)
From this state, the potential of the shift gate SH is set to H in response to a data read request from the system controller (microcomputer),
The shift gate SH is opened, and the accumulated charge is transferred to the SR section. The accumulated charges transferred to the SR section are transferred to the analog section by the transfer clocks φ ₁ and φ ₂ .

Note that (d) is a supply potential diagram in a state where the integration clear operation is performed. FIG. 9 is a supply potential diagram in the low luminance integration (hereinafter referred to as PD integration) mode.
(A) shows a state during integration after the integration clear operation is performed. Unlike the high brightness integration mode in which the generated charges are accumulated in the accumulation part ST, the potential of the barrier gate BG is set to L and the barrier gate BG is closed. Charge storage is performed on the photodiode PD.

(B) shows signals SH, RCG, and STIC that represent dark charges generated in the storage section ST during the integration period shown in (a).
G is discharged as H to the power source OD.

(C) shows the state after discharging the dark charges generated in the storage section ST in (b). Barrier gate B
The potential of G is set to H, the barrier gate BG is opened, and the accumulated charge is transferred to the accumulation part ST. At this time, the potential SH of the shift gate is set to L so that the charges transferred to the storage portion ST do not flow out to the SR portion.

(D) shows a state after the accumulated charges are transferred to the accumulating portion ST in (c). The potential of the barrier gate BG is set to L, the barrier gate BG is closed, and the electric charges generated later in the photodiode PD portion are prevented from flowing into the storage portion ST. Further, the supply potential of the storage unit ST is set to L, and the generation of dark current in the storage unit ST is suppressed.
From this point onward, it is the same as the high brightness integration mode. Note that the integration clear operation is the same as in (d) of FIG.

As described above, in the low brightness integration (PD integration) mode, the operation of ending the integration is more complicated than in the high brightness integration (ST integration) mode. Therefore, when the brightness is high, there is a case where the excessive integration is performed due to the delay of the completion of the integration. However, since the dark current in the photodiode PD unit is smaller than that in the storage unit ST, the influence of the integration end delay is small and advantageous when the subject brightness is low. For this,
The two types of integration modes described above are used properly.

FIG. 10 is a layout view of the color detection elements and the monitor pixels. As shown in the figure, the color detection element has pixels with green (G) and red (R) filters arranged at both ends of the reference pixel row. Pixels with blue (B) and infrared (IR) filters are arranged at both ends of the reference pixel row. Each extends so as to cover the upper side of the standard portion and the reference portion. Here, the pixel area of each color detection element is the same as other pixels. As a result, the color temperature of the light source viewed by the image sensor is detected, the focus movement at each wavelength is corrected, and the focus detection accuracy is improved. Further, it is possible to detect a flesh-colored subject (person) by the output of each color detection element. A method of detecting this flesh-colored subject (person) will also be described later.

As shown in FIG. 10, four kinds of monitor pixels M1 are provided in each of the standard portion and the reference portion of the image sensor.
(M1 ') to M4 (M4') are arranged. These can be selectively used by a microcomputer. Also this 4
The types of monitor pixels are configured to have the same area. The reason for this is that the pixel output of the portion covered by the monitor pixel becomes an appropriate level regardless of which monitor pixel is used.

The monitor pixel M1 (M1 ') is a normally used monitor pixel, and monitors the photoelectric conversion output of the entire pixel area. However, when there is an extremely bright subject around the main subject (for example, when the subject is backlit), when the monitor pixel M1 (M1 ') is used, the pixel output for the main subject reaches an appropriate level due to the effect of the bright portion. The integration of the image sensor ends before. Therefore, correct distance measurement cannot be performed. That is, there is a problem that the main subject is dark and the contrast is low, so that distance measurement is impossible and the camera focuses on a bright portion (a portion other than the main subject).

It is an object of the present invention to provide a photoelectric conversion device for focus detection which can prevent such a problem and can correctly focus on a main subject.
That is, when the microcomputer determines that the main subject is at the center, the microcomputer selects the monitor pixel M2 (M2 ′) and the monitor output is integrated. As a result, the pixel output for the main subject is set to an appropriate level. At this time, since the bright portion overflows, distance measurement is performed only in the main subject output waveform portion. Then, the main subject can be properly focused. Similarly, if the main subject is on the right side, monitor pixel M3 (M3 ') is used, and if it is on the left side, monitor pixel M2 (M2') is used. As a result, it is possible to provide a focus detection photoelectric conversion device capable of accurately focusing on the main subject.

The photodiode P shown in FIG.
D, barrier gate BG, storage portion ST, shift gate S
H, the shift register SR, and the signal terminals RCG and OD respectively correspond to the same reference numerals in FIG.

(1) First Embodiment (Two-Terminal Output Type) In the first embodiment, the sensor output of the image sensor is divided into two terminals and output. FIG. 11 is a time chart of a two-terminal output type in ST integration. First, a long H pulse is input to the integration start signal IST. As a result, the L level of the signal BG _i becomes
The H potential is applied to the shift gate SH, and the H potential is applied to the register clear gate RCG. During that time, a short read start signal RST pulse is input twice. As a result, the automatic gain circuit (hereinafter abbreviated as AGC circuit) is reset. Then, the potential of L is applied to the shift gate SH at the fall of the signal IST pulse, and the shift gate SH is closed to start the integration.

At this time, the monitor is selected using the mode terminals MD2 and MD3 at the fall of the signal RST.
When the signal IST falls, the terminal MD2 sets the integration mode, and the terminal MD3 switches between the sixth and seventh islands. Monitor selection means that the microcomputer sets which of the monitor pixels M1 (M1 ') to M4 (M4') shown in FIG. 4 is to be used. The integration mode setting means that the microcomputer sets ST integration and PD integration according to the previous integration time.
That is, when the previous integration time is shorter than the predetermined value (when the subject has high brightness), ST integration is performed, and when it is long (low brightness), PD integration is set. Switching between the sixth and seventh islands means switching between the normal pixel row (sixth island) and the red filter pixel row (seventh island) depending on whether auxiliary light is used or not.

Next, the termination of integration will be described. Signal IS
The integration is started by the fall of T, and the monitor level drops according to the amount of incident light. When this level falls below a certain threshold value within a certain period of time, it is assumed that the amount of light incident on the image sensor is sufficient, and the AGC integration end signal FLG-D _i is inverted and the signal BG _i is changed. It is set to H, the potential of the barrier gate BG _i is set to L, and the integration ends. Then, the signal ADT falls. When the integration of the other islands has not been completed yet, the signal ADT is raised after the elapse of a certain period of time to prepare for the completion of the next integration.

Next, the forced termination will be described. In the AF system, the lens must be driven to the in-focus position within a time period that does not make the photographer uncomfortable. Therefore, this system determines the maximum integration time. Therefore, when the integration time reaches the maximum integration time, the microcomputer forcibly ends the integration. First, the microcomputer sets the signal CBG to H in order to forcibly terminate the integration. As a result, the end of integration is transmitted to the CCD. Then, the CCD sets the signal BG _i to H, and the barrier gate BG
The potential of _i is set to L to prevent the charge from flowing from the photodiode part PD to the storage part ST, whereby the integration is ended and the signal ADT is set to L. In this time chart,
One of the four islands is naturally terminated and the remaining three are killed. After the CCD finishes the integration end operation of all islands, the microcomputer inputs the signal RST pulse and shifts to the read operation.

Next, PD integration will be described with reference to FIG. The integration start operation is almost the same as ST integration,
Since it is necessary to operate the signal BG as shown in FIG. 9, the signal BG _i changes to H at the same time when the signal IST pulse falls.
It is said. After the integration is cleared, the generated charges are accumulated in the photodiode PD section. At this time, the shift gate signal SH _i is kept at H, and the dark charge generated in the storage portion ST is discharged to the shift register portion SR.

In the natural termination, when the signal FLG-D _i becomes H, the signal BG _i becomes H in the ST integration, and the signal ADT
Is L. On the other hand, in PD integration, the shift gate SH is opened before the barrier gate BG is opened, and the storage unit ST is opened.
Unnecessary charges generated in step S are discharged, and then the shift gate S
After closing H, the barrier gate BG is opened. As a result, the generated charges are transferred from the photodiode part PD to the storage part ST, the barrier gate BG is closed, and the integration is completed.
Therefore, the microcomputer sets the signal SH to L and the signal ADT to L after the signal FLG-D _i becomes H. After that, the signal BG _i is set to L, and after a fixed time (100 μs)
The signal BG _i is set to H. During this 100 μs, the charges are transferred from the photodiode part PD to the storage part ST.

Forced termination is performed in the same manner as ST integration.
The signal CBG is set to H, and the operation thereafter is the same as the natural end. After completion of integration of all islands, a read operation is started by inputting a signal RST pulse.

A two-terminal input type read operation will be described with reference to FIGS. As shown in the block diagram of the 2-terminal output type in FIG. 15, analog data from the CCD is output from the terminals V _out1 and V _out2 . On the other hand, the number of inputs to the A / D converter is one. Terminal V
The outputs of _out1 and V _out2 are A / D converted by switching the inputs alternately pixel by pixel. If there are two A / D inputs, they are simultaneously input to the A / D converter.

As shown in FIG. 13, in the 2-terminal output type, after the signal RST pulse is input, the mode terminal MD1
Is set to H, and the AF sensor is set to the data dump mode. The data of the fifth island is output from the terminal V _out1 and the data of the sixth or seventh island is output from the terminal V _out2 in synchronization with the ADT signal (ADS signal) which is the timing signal of the pixel data output. At this time, the signal CBG is set to L at the time of the pixel output of the black reference portion, so that the black reference output is sampled and held. The black reference output is sampled and held in this way for the following reason. The dark current of the CCD increases in proportion to the temperature. Therefore, the black reference portion pixels shielded by aluminum are subtracted from the photoelectric conversion output as a representative of all the pixels. As a result, only the actual photoelectric conversion output excluding the dark current portion is output. When the standard pixel output or the reference pixel output is output, the signal IST is set to H, so that the output from the terminal V _out is switched from the temperature data output to the pixel data output.

The read start signal input RST terminal is an input / output switching terminal, and the AGC5 signal is generated by the signals MD2 and IST.
1, AGC52, AGC61, AGC62 are switched and output. In addition, if an interrupt is received from another during the data dump period and the process is prioritized, the signal IS
When T is H, the input clock pulse is stopped by setting the signal MD3 to H. Then the data dump is interrupted. Here, AGC51 and AGC52 are 2-bit data of the AGC signal from the fifth island IS5, and 1 and 2 are the first data and the second data, respectively.
Represents the second data.

When the reading of the pixel outputs of the standard portion and the reference portion is completed, the signal MD1 is set to L and the signal RST pulse is input, so that the next pixel output, that is, the second and fourth pixels is output.
The pixel data of the islands IS2 and IS4 is read.
The reading method is the same as in the case of the fifth and sixth islands IS5 and IS6.

Next, the first and third islands IS1 and IS3
Output is read. In this embodiment, the order of islands read out as described above is predetermined, and the signal R
It changes every time the ST pulse is input. The relationship among the pixel output signals V _out1 , V _out2 , the external output signal ADT and the internal signals φ ₁ , φ ₂ , RS, ARS, S / H is shown in FIG.
Is shown in Here, φ ₁ and φ ₂ are transfer clocks of the CCD, RS is a reset signal of the output section of the CCD, ARS
Is the reset signal of the output amplifier, S / H is the output signal V _out
Is a signal for sampling and holding.

FIG. 16 is a block diagram showing details of a two-terminal output type AF sensor and its periphery. A 2-terminal output type device configuration will be described with reference to FIG. The photoelectric conversion device for focus detection includes an AF sensor 17 for detecting a signal from a subject and an AF sensor 17
And an DC / DC converter 51 that supplies an electric potential for driving the AF sensor 17 and the AF controller 30. The AF sensor 17 includes the seven islands IS1 to IS7 described in FIG. 5, which are CCDs. The CCD forming each island is provided with a monitor pixel for detecting the accumulation of electric charge in the CCD. The output from each monitor pixel is input to the brightness determination circuit 53. The output signal from each island is input to the switching control circuit 54. However, 6th island IS6 and 7th
The circuit is configured so that it is used by being switched as described above with respect to the island IS7. The output from each luminance determination circuit 53 is input to the control circuit 52. A signal from the control circuit is input to the switching control circuit 54. From each switching control circuit 54, the integration result is output to the AF controller 30 via the terminals V _out1 and V _out2 via the CCD output processing circuits 55 and 56. The CCD output processing circuit 55
The correction signal from the temperature detection circuit 57 is applied to the output signal from the.

The shift register S of the CCD is connected from the DC / DC converter 51 to the control circuit of the AF sensor 17.
A power supply voltage V _CC (13V) for driving R is applied. From the DC / DC converter 51 to the control circuit 52 of the AF sensor 17 and the AF controller 30,
V _REF power supply (5V) and GND are supplied. A clock pulse CP and a mode switching signal MD1 are sent from the AF controller 30 to the control circuit 52 of the AF sensor 17.
~ MD3, integration start signal input IST, read start signal RS
T, a forced termination signal CBG for integration, and a timing signal ADT for pixel data output are input. The photoelectrically converted output signals V _out1 and V _out2 are input to the A / D converter 31 of the AF controller 30 in synchronization with the signal ADT.
The AF controller switches the input pixel by pixel and alternately performs A / D conversion.

Next, the operation of the camera to which the focus detecting photoelectric conversion device according to the present invention is applied will be described. FIG.
7 is a flowchart showing the main flow of a camera to which the focus detecting photoelectric conversion device according to the present invention is applied. When a main switch S _M (not shown) provided on the camera body is turned on (step # 101), the camera system is turned on. Next, the release button (not shown) is 1
It is determined whether or not the button has been pushed (step # 102). When the release button is not pushed down one step in step # 102 (hereinafter abbreviated as S1 switch), the main switch S _M
Is determined to be off (step # 103,
(# Is omitted below). If the main switch S _M is off in step 103, the camera stops operating (step 104). If the main switch S _M is not off in step 103, the program returns to step 102.

If S1 is turned on in step 102, lens data is input to the camera body (step 10
5) The AF subroutine described below is executed (step 106), and photometry and exposure calculation (hereinafter abbreviated as AE) are performed (step 107). Next, it is determined whether or not a release button (not shown) has been pressed in two steps (hereinafter abbreviated as S2 switch) (step 108). If the switch S2 is turned on in step 108, exposure control (shooting) is performed (step 109), and the switch S1
Is turned off (step 110) and the program returns to step 102. If the switch S2 is not on in step 108, the program returns to step 105.

FIG. 18 is a flow chart showing the contents of the AF subroutine shown in step 106 of FIG.
First, the first monitor (M1, M that covers the entire area of the island
1 ') is selected (step 1000) and integration is performed (step 1005). After the integration is completed, the data of the entire island is dumped (step 1010), and the distance measurement calculation is performed (step 1015). Next, the main subject is discriminated between the blocks in the island (step 102).
0), the block used for distance measurement is determined.
On the other hand, it is determined whether the monitor selection is correct (step 1025), and if it is correct, the lens is driven as it is (step 1030). Here, whether or not the selection of the monitor is correct is judged from the detected luminance distribution data. If the monitor selection is incorrect, the program proceeds to step 1030 to select the monitor corresponding to the block containing the main subject. That is, step 102
If the monitor selection is wrong in step 5, the block in which the main subject exists is determined and the monitor corresponding to the block is selected again (steps 1030, 1035, 104).
0, 1045). Then, the integration is redone (step 1050). After the integration is completed, a data dump is performed (step 1055), and the distance measurement calculation is performed only on the block in which the main subject exists (steps 1060 and 1065).
1070). After that, the program proceeds to step 1030 to drive the lens and move the lens to the in-focus position.

That is, according to the present invention, in the CCD composed of one line for measuring the luminance of the object, the CCD shown in FIG.
As shown in 0, the CCD is divided into a plurality of blocks, and a plurality of monitors corresponding to the blocks are arranged near the positions corresponding to the blocks. When the output signal from the CCD is monitored and the main subject is not clear from the signal from the first monitor that covers the entire area of the island, the block in which the main subject exists is determined and the monitor corresponding to that block is selected. That is, since integration control is performed according to the main subject, even if the main subject has low brightness during backlighting, it is possible to correctly focus on a dark subject without being pulled by the high-brightness contrast subject in the background. Next, another embodiment of the AF subroutine will be described. FIG. 19 is a flow chart showing another embodiment of the AF subroutine. First, the first monitor (M1, M1 ') covering the entire area of the island is selected (step 110).
0), integration is performed (step 1105). After the integration is complete, the data of the entire island is dumped (Step 1
110) and the brightness of each block is examined (step 11
15) It is checked whether there is an extremely low level block with respect to other blocks (step 1120). If there are no low-level blocks (step 1120 returns N
O), distance measurement calculation is performed for each block (step 1125), and the main subject is determined (step 113).
0), the lens is driven (step 1135). As a result, the lens is moved to the in-focus position (step 119).
8). If there is a block whose level is extremely lower than the other blocks in step 1120, the pixel data of blocks other than the low level block are stored in the memory (step 1145) and the monitor corresponding to the low level block reselects. (Step 1150, step 1155,
Steps 1160 and 1165). Then, reintegration is performed (step 1170) and data dump is performed (step 1175). Next, the distance measurement calculation is performed only on the low-level blocks (step 1180, step 1185, step 1190), and the defocus amount is calculated (step 1195). Further, the distance measurement calculation is performed by the pixel data stored in the memory other than the low level block (step 1195), and the defocus amount is calculated.
The main subject is discriminated from the obtained defocus amounts (step 1140), the lens is driven (step 1135), and the lens is driven to the in-focus position (step 1198).

In another embodiment of the AF subroutine shown in FIG. 19, when the brightness is judged for three blocks and there is a low level block in which the main subject is black due to backlight or the like, The main subject is discriminated from the distance measurement calculation results of blocks other than the low level block, and the lens is driven to the in-focus position based on the discrimination.

Next, a modification of the arrangement of the monitors will be described with reference to FIG. FIG. 20 is a diagram corresponding to FIG. 10. The difference between FIG. 10 and FIG. 20 is that the first island covers the entire island.
This is that the monitors M1 and M1 'are arranged on the side close to the pixels of the CCD. Further, FIG. 20 further shows an example of division of pixel columns into blocks. Referring to FIG. 20, the reference part and the reference part are each divided into three blocks A, B, and C, and a monitor is arranged at a position corresponding to each block. That is, the monitors M3 and M3 'are provided for the block A, and the monitors M2 and M3 are provided for the block B.
M2 'and block C are provided with monitors M4 and M4', respectively. The same monitor is arranged at the position corresponding to each of the standard portion and the reference portion, and each monitor is connected by wiring. The reason for this is to obtain an appropriate gain at least when the focus is achieved.

(2) Second Embodiment (One-terminal Output Type) Next, a one-terminal output type photoelectric conversion device for focus detection will be described. 1 terminal output type photoelectric conversion device is AF
The sensor 17 has only one output terminal.
Therefore, in the terminal output type photoelectric conversion device described in the first embodiment, the output signal V from the AF sensor 17 is
_This corresponds to the case where there is no _out2 (see FIG. 16).

A time chart of the one-terminal output type will be described with reference to FIGS. Basically CC
The driving method of D is the same as that of the two-terminal output type, but if all pixel outputs for six islands are output from one terminal, the data dump time becomes extremely long, which is unsuitable for an AF system. Therefore, during the first distance measurement, as shown in FIG. 28B, the transfer clocks φ ₁ , φ ₂
Is set to double the normal speed (FIG. 28 (A)), and the shift gate signal SH is output once for two pixel outputs. Therefore,
The data for two pixels is added as hardware for one pixel and output. As a result, the number of pixels is apparently equivalent to half the number, and the data dump time is ½ of the normal time.

When this mode is used, as will be described later with reference to FIG. 21, after the island having the main subject is obtained by the first distance measurement, the distance measurement from the second time is performed only on the island having the main subject. Alternatively, data is read at normal speed by designating only a plurality of islands including the island. As a result, it is possible to shorten the data dump time and measure the distance with high accuracy.

The mode setting at the start of integration is set by the fall of the integration start signal IST using the mode signals MD2, MD3 and MD4 as shown in FIG. This mode setting may be performed by the method shown in the first embodiment. The setting of the read island is also the mode signal MD.
2, MD3, MD4 (see FIG. 26).

Referring to FIG. 23, the ST of 1-terminal output type
A time chart of integration will be described. As described above, the mode signal MD is generated by the fall of the integration start signal IST.
2, MD3, MD4 are used to select a monitor, an integration mode, and sixth and seventh islands (see FIG. 25). The reading speed is set by the mode signal MD2 and the rise of the integration forced termination signal CBG. Read start signal RST and mode signals MD2, MD3,
A read island is selected by MD4.

The signal setting of the barrier gate signal BG _i and the shift gate signal SH _i and the accompanying operation are the same as in the case of the first embodiment. The method of natural termination is the same. Similarly to the case of the first embodiment, the forced termination operation is also performed by setting the integration forced termination signal CBG to H. At this time, the reading speed (normal speed reading or double speed reading) is switched by the mode signal MD2. Is set. Next, the read start signal RST pulse is input, the read island is selected as described above, and the island to be read is designated by the microcomputer. PD integration (see Figure 23)
Also applies to the case of ST integration. The data dump (see FIG. 24) is basically the same as the two-terminal output type. By setting the read island when the read start signal RST pulse is input, the problem that the data dump time becomes long is solved.

FIG. 27 is a diagram showing a high-speed read circuit in the case of performing high-speed read at a double pitch. When this high-speed read circuit is used, the auto-gain signal AGC is X1, X2, X4, and X8 when the read mode is the normal read mode, but is X0.5, when the high-speed read mode is set.
X1, X2, X4.

Next, the flow chart of the one-terminal output type AF subroutine will be described. FIG. 29 is a flow chart of the 1-terminal output type AF subroutine.
It corresponds to FIG. 18 of the first embodiment. First, monitor pixels M1 and M1 ′ covering the entire pixel row are designated (step 2000), and PD integration is set as an integration mode (step 2005). Then, the sixth island of the sixth and seventh islands is designated (step 20).
10), charge accumulation (integration) is started (step 20).
15). After the integration is completed, the integration time is checked (step 2020). If the integration mode is inappropriate (NO in step 2020), the integration is redone, and if appropriate, the read speed is set to the double read speed (step). 20
25), all islands are sequentially designated and data dump is performed (step 2030, step 2035).
If the luminance distribution is confirmed and the monitor setting is appropriate (step 2045), the program shifts to the distance measurement calculation (step 2050). If the monitor setting is incorrect (NO in step 2040), the monitor pixel corresponding to the pixel block that is considered to include the main subject is reset and the integration is performed again. Distance measurement calculation (step 2050),
Average processing (step 2055), most recent priority (step 2)
The amount of focus shift is obtained by an algorithm such as (060) (step 2065), and the lens is driven to the in-focus position (step 2070). When the distance measurement is continuously performed on the main subject, only the pixel column including the main subject is read, the island is designated, and the distance measurement is performed.

In the device configuration diagram corresponding to FIG. 16 of the first embodiment in the second embodiment, the output terminal of V _out2 is omitted in FIG. 16, so that the illustration is omitted.

Next, an embodiment in the double speed mode will be described. Here, the double speed mode means that data is read out faster than in normal reading. FIG. 21 is a flowchart showing an AF subroutine when the double speed mode is used. In step 1200, it is determined whether or not it is the first distance measurement, and if it is the first distance measurement, multipoint distance measurement using all islands is performed (step 120).
5). In step 1205, integration of all islands of the CCD is performed, the data output mode is set to the high speed mode (double speed mode) (step 1210), and data dump of all islands is performed (step 1215).
Next, distance measurement calculation for each island is performed (step 12
20) Then, it is determined which island the main subject is in from the distance measurement result of each island (step 12).
25) The islands to use are determined (step 1230). The lens drive amount is calculated from the amount of focus shift of the island to be used next (step 1235), and the lens is driven by that amount (step 1240).

When it is determined in step 1205 that the distance measurement is the second time or later, the distance measurement of the designated island determined last time is performed. That is, integration of the designated island is performed (step 1245), the data output mode is switched from the high speed mode to the normal speed mode (step 1250),
Data for the specified island is dumped (step 12)
55). Next, distance measurement calculation is performed (step 126).
0), the lens drive amount is obtained from the focus shift amount (step 1265), and the program proceeds to step 1240.

It should be noted that the integration of only the designated island is performed at the second and subsequent distance measurements.
Integration may be performed for all islands, as indicated by 0.

(3) Third Embodiment (One-terminal Output Type and Four Islands) Next, a third embodiment of the present invention will be described. In the third embodiment, the output terminals are the same as in the second embodiment, but the number of islands for focus detection is reduced from seven to four.

The four islands in this embodiment are the 11th to 14th islands IS11 to IS14, respectively.
And

FIG. 30 is a device configuration diagram according to the third embodiment of the present invention, FIG. 31 is a layout diagram of monitor pixels in the third embodiment of the present invention, and FIG. 32 is the third embodiment. 6 is a timing chart of each terminal of the device according to the example.

FIG. 30 corresponds to FIG. 16 of the first embodiment, FIG. 31 corresponds to FIG. 4A, and FIG. 32 corresponds to FIG. 11 and FIG.
Corresponds to 2.

Referring to FIG. 32, mode signal MD1 is L
, The integration mode is set, the read start signal RST pulse is input twice while the integration start signal IST is H, and A
The GC circuit is reset. At this time, the read start signal R
At the trailing edge of the ST pulse, monitor setting is performed using the mode signals MD2 and MD3. At the fall of the integration start signal IST pulse, the integration mode is set by the mode signal MD2, the reading speed is set by the mode signal MD3, and the integration is started. After the integration start signal IST pulse falls and a certain period of time has elapsed, the input and output of the mode signals MD2 and MD3 are switched, and their output terminals become output terminals. The integration information of each island is output in parallel using these output terminals TAT, MD2, MD3, and ADT. An AF controller, not shown, waits for each of the four parallel outputs of TINT to go low by allowing an interrupt. Therefore, the AF controller can check the order in which the integration end signal TINT of each island becomes L and the integration time. The first read start signal RST is input when all the TINT signals are input or when the integration is forcibly terminated after a certain period of time. By inputting this signal, the mode terminals MD2 and MD3
Is designated as an input pin, and the subsequent second read start signal R
The input of the ST pulse sets the read island by these pins and starts reading the data. Next, the mode signal MD1 is set to H, and the data dump mode is set. In synchronization with the ADT output signal ADT pulse,
Pixel data is output from the signal output terminal V _out . When the data represents unnecessary image data, the integration start signal IS
T is set to L, and the temperature data is output from the signal output terminal V _out . When the black reference pixel output is output, the read start signal RST is set to H, and the black reference output data is sampled and held. Next, when the reference pixel is output, the integration start signal IST is set to H, and pixel data is output from the signal output terminal V _out . Further, during the data dump, it may be impossible to take in the data because the AF controller does other work. In such a case, the signal TAT is set to H and the data dump is interrupted.

(4) Auxiliary Light Mode Next, description will be made on the case of distance measurement with the auxiliary light component (infrared light) described in the first embodiment. FIG. 33 is a diagram showing the relationship between the emission wavelength and the emission intensity of an auxiliary light LED that generates auxiliary light. 33, the auxiliary light has a spectral characteristic having a peak at a wavelength of 695 nm.

FIG. 34 is a diagram showing the relationship between the wavelength and the spectral sensitivity of the CCD sensor constituting the AF sensor with respect to the wavelength. When the distance measurement is performed with the constant light, the distance measurement is performed only with the auxiliary light so that the surrounding white light does not affect the distance measurement. At this time, a red filter for measuring the distance of only the auxiliary light component is provided in the detector of the CCD. A red filter is also provided in the monitor element that controls the integration. This is to match the amount of light received by the sensor with the amount of light received by the monitor. In FIG. 34, the emission wavelength of the auxiliary light LED shown in FIG. 33 is indicated by a dotted line, and the CCD formed by the bandpass filter formed by the infrared cut filter and the red filter in this way.
Since the auxiliary light is detected in the high-sensitivity area, distance measurement is performed without being affected by external light. It is more effective to change the transmission band of the red filter according to the spectral distribution of the auxiliary light.

Next, the AF subroutine in the auxiliary light mode will be described. FIG. 35 shows A in auxiliary light mode.
7 is a flowchart showing an F subroutine. First,
Through 6th islands IS1 to IS6 are used (step 3000) and integration is performed (step 300).
5). The result is data dumped (Step 301
0), it is determined whether or not all the first to sixth islands are low contrast (step 3015). If all the islands are not low contrast (NO in step 3015), the distance measurement calculation is performed based on the data of the island for which the contrast is obtained (step 3020).
An island with a subject is selected (step 302
5) and the lens is driven to that position (step 303).
0), the lens is moved to the in-focus position (step 303).
5). When it is determined in step 3015 that all the islands are low contrast, it is determined whether or not the output signals from all the islands have low brightness (step 3045). When it is determined that the output signals from all the islands have low brightness (Y in step 3045).
When ES), the auxiliary light mode is set (step 3
050), the auxiliary light circuit (not shown) is driven. Then, the seventh island IS7 with a red filter for auxiliary light is selected from the sixth island and the seventh island (3055), and integration is performed (step 306).
0), the data is dumped (step 3065). After that, it is judged again whether or not all the islands are low-con (step 3070), and when the data from all the islands are not low-con (when NO in step 3070), the seventh
Distance measurement calculation is performed based on the data obtained by the island IS7 (step 3075), and lens is moved to the in-focus position based on the focus shift amount (step 303).
0, step 3035). If it is determined in step 3070 that all islands are low contrast, low contrast scanning is performed (3080) and the program returns to step 3000 again.

(5) Discrimination of Color Detection Element and Island in which Main Subject is Present Using the Color Detection Element Next, the color detection element described in the explanation of FIG. 10 of the first embodiment and the island in which the main subject is present are used. A method for determining the above will be described. FIG. 36 is a diagram for explaining a method of forming a dye filter used in the color detection element. This color detecting element is formed by an on-air method with a dye filter composed of three primary colors of R, G, and B. The on-air method is formed by applying a transparent register called a base layer on a wafer on which a CCD wafer process has been completed. This layer serves to improve the unevenness of the wafer generated during the wafer process and to improve the adhesion with the subsequent dye resist. After this, a normal PEP (Photo Engraving Proc)
The exposure and development are performed by ess).

A dye filter is formed through the processes of (a) to (g) of FIG. After the exposure and development, as shown in (b), a dyeing resist is applied on the base layer to form a pattern on each pixel. Next, this wafer is dipped in a green G dyeing solution and dried to form a G layer. Next, as shown in (c), a transparent resist is applied and patterned again in order to prevent color mixture in the next dyeing step. This transparent layer is called the intermediate layer. Next, as shown in (d), a dyeing resist is applied and patterned in the same manner as that for forming the G layer. Next, this wafer is dipped in a red R dyeing solution and then dried. This layer is called the R layer. Next, as shown in (g), a transparent resist is applied again in order to improve the concave state generated on each pixel. Next, as shown in (f), a dyeing resist is applied and patterned on this transparent resist (second intermediate layer). This layer is called the W (white) layer and is not dyed. This layer is used for improving the surface condition of the R layer and G layer and improving the color reproducibility. Finally, as shown in (g), an overcoat layer for protecting the surface is formed to complete the color filter of the color CCD image sensor.

The spectral transmittance characteristics of the color filters of the color CCD image sensor thus formed are shown in FIG.

Next, a method for detecting the island where the main subject exists will be described. Since the main subject is generally a person, it is considered that the main subject exists in the area where the skin color is detected. Therefore, the algorithm for detecting the skin color will be described below. Color information at each part is detected from the output ratio of the color detection elements R, G, and B elements, the type of light source is determined from the overall color output ratio, and image magnification information is obtained from the defocus amount of each part. From these three pieces of information, the part considered to be a person is watched. The position of the main subject can be selected by recognizing that part as the main subject and applying autofocus.

First, the detection of color information will be described. 12 on the AF sensor, which is located at the same position
The light-receiving elements provided at the places are covered with red, green, blue, and infrared (hereinafter abbreviated as R, G, B, and IR) dye filters. Here, the output of the light receiving element R is V _r , the output of G is V _g , the output of B is V _b , and the output of the infrared IR is V _ir . The output ratio V _r / V _{b of} the above-mentioned 12 light receiving elements,
V _g / V _b and V _ir / V _b are obtained. These are color information for each of the six islands. Next, the detection of the light source will be described. The skin color of a person has a different reflection wavelength distribution depending on the light source. Therefore, it is not enough to detect the color information of the subject. Therefore, it is necessary to detect the state of the light source. The light source is estimated from the color information described above, and the correction by the light source is added to the skin color threshold of the color information.

Next, the method of estimating the light source will be described. The light source is estimated by the ratio of the output of the IR filter and the output of the B filter, which are arranged similarly to the color information detecting element. 12
The output of the IR filter and the output of the B filter among the outputs of the color information elements provided at the respective places is calculated, and the light source is estimated by the ratio of the sums.

[0088] R (ir / b) = ΣV iri / ΣV bi (i = 1~4) relationship of the light source is determined to detectable levels and ratios and the skin color obtained by this calculation is shown in Figure 38. Here, R _{g / b at a} level where skin color can be detected represents the coverage of V _g which is the output of G and V _b which is the output of B, and R _{r / b} is the output of V _r and B which is the output of R _Shows the ratio of V _{b to} the output.

Based on FIG. 38, the light source is first estimated by the value of R _{ir / b} , and it is checked whether or not the color detection information of each block is included in the skin color detection level in the estimated light source state. The skin color block is extracted.

FIG. 39 is a flow chart showing the AF subroutine when the pixels with color filters are used. FIG.
When a pixel with a color filter is used, first, integration of all islands is performed (step 400).
0), based on the result, the main subject, that is, the island where the person exists is determined (step 400).
5). Next, distance measurement calculation is performed using the data of the island in which the main subject exists (step 4010), the focus shift amount due to the difference in the light source is corrected (step 4013), and the focus shift amount is obtained. The lens is driven (step 4015), and the lens is moved to the in-focus position (step 4020).

Next, with reference to FIG. 40, a subroutine for discriminating an island in which a person as a main subject exists will be described. In the person island determination subroutine, first, the island number and the loop number n
Is set to 6 (step 4110). Next, the skin color of the person is detected from the nth output data of R, G, B, and IR by the method described above (step 411).
5). Next, it is determined whether or not the output data of the n-th island represents a skin color (step 4120),
If the skin color is represented, the n is stored in a predetermined memory (step 4140). If not, 1 is subtracted from n and the above determination is repeated until n becomes 0 (steps 4125 and 4130). ). After that, an island close to the camera is selected from the stored n (step 4135). As described above, the main subject, that is, the island where the person is present is discriminated, and the AF sensor performs integration so that the output data of the desired island is a proper value.

[Brief description of the drawings]

FIG. 1 is a perspective view showing a focus detection optical system of a camera using a focus detection photoelectric conversion device according to the present invention.

FIG. 2 is a principle explanatory view of the focus detection optical system shown in FIG.

FIG. 3 is a diagram showing a display in a viewfinder in a camera to which the present invention has been applied.

FIG. 4 is a diagram showing details of a CCD chip used in the photoelectric conversion device according to the present invention.

5 is an explanatory diagram showing divided areas of a reference portion in the CCD chip shown in FIG.

FIG. 6 is an explanatory diagram showing a shift amount for each divided area in a CCD chip.

FIG. 7 is a block circuit diagram of an AF sensor and an AF controller that realize the photoelectric conversion device according to the present invention.

FIG. 8 is an explanatory diagram showing different integration modes of the photoelectric conversion device to which the present invention is applied.

FIG. 9 is an explanatory diagram showing different integration modes of the photoelectric conversion device to which the present invention is applied.

FIG. 10 is a diagram showing an arrangement of a color detection element and a monitor according to the present invention.

FIG. 11 is a time chart according to the first embodiment of the present invention.

FIG. 12 is a time chart according to the first embodiment of the present invention.

FIG. 13 is a time chart according to the first embodiment of the present invention.

FIG. 14 is a time chart according to the first embodiment of the present invention.

FIG. 15 is a block diagram showing a main part of a two-terminal output type according to the first embodiment of the present invention.

FIG. 16 is a principal block diagram of the photoelectric conversion device according to the first embodiment of the present invention.

FIG. 17 is a flowchart showing an operation of a main part of the photoelectric conversion device according to the first embodiment of the present invention.

FIG. 18 is a flowchart showing an operation of a main part of the photoelectric conversion device according to the first embodiment of the present invention.

FIG. 19 is a flowchart showing an operation of a main part of the photoelectric conversion device according to the first embodiment of the present invention.

FIG. 20 is a diagram showing a modification of the monitor pixel arrangement of the photoelectric conversion device according to the present invention.

FIG. 21 is a flowchart of an AF subroutine in double speed mode.

FIG. 22 is a time chart of the photoelectric conversion device according to the second embodiment of the present invention.

FIG. 23 is a time chart of the photoelectric conversion device according to the second embodiment of the present invention.

FIG. 24 is a time chart of the photoelectric conversion device according to the second embodiment of the present invention.

FIG. 25 is a diagram for explaining a method of setting an integration mode and a read island in the photoelectric conversion device according to the present invention.

FIG. 26 is a diagram for explaining a method of setting an integration mode and a read island in the photoelectric conversion device according to the present invention.

FIG. 27 is a diagram for explaining a high speed read mode.

FIG. 28 is a diagram for explaining a high speed read mode.

FIG. 29 is a flowchart showing an AF sub routine of the second embodiment according to the present invention.

FIG. 30 is a diagram for explaining the third embodiment according to the present invention.

FIG. 31 is a diagram for explaining the third embodiment according to the present invention.

FIG. 32 is a view for explaining the third embodiment according to the present invention.

FIG. 33 is a diagram for explaining an auxiliary light mode of the photoelectric conversion device according to the present invention.

FIG. 34 is a diagram for explaining an auxiliary light mode of the photoelectric conversion device according to the present invention.

FIG. 35 is a diagram for explaining an auxiliary light mode of the photoelectric conversion device according to the present invention.

FIG. 36 is a diagram for explaining a pixel with a color filter in a photoelectric conversion device according to the present invention and a method of using the pixel.

FIG. 37 is a diagram for explaining a pixel with a color filter in a photoelectric conversion device according to the present invention and a method of using the pixel.

FIG. 38 is a diagram for explaining a pixel with a color filter in a photoelectric conversion device according to the present invention and a method of using the pixel.

FIG. 39 is a flowchart showing an AF subroutine when using pixels with color filters.

FIG. 40 is a flowchart showing a subroutine for discriminating a person island.

[Explanation of symbols]

16a to 16g are photoelectric conversion element arrays, 17 is an AF sensor,
Reference numeral 30 is an AF controller, 31 is an A / D conversion unit, 32 is a memory, 33 is a focus detection unit, 34 is a correction calculation unit, 35 is a lens drive control unit, 36 is a timer circuit, and 37 is an AF sensor control unit.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Tokuji Ishida 2-3-13 Azuchi-cho, Chuo-ku, Osaka, Osaka International Building Minolta Co., Ltd. (72) Inventor Toshio Morita 2-chome, Azuchi-cho, Chuo-ku, Osaka 3-13 Osaka International Building Minolta Co., Ltd.

Claims (1)

[Claims] 1. A photoelectric conversion device for focus detection used for TTL autofocus, wherein the photoelectric conversion device for focus detection is arranged other than the optical axis and is arranged along a straight line passing through the center of the optical axis. , A photoelectric conversion device for focus detection.