JPH08286103A - Focus detector - Google Patents

Abstract

PURPOSE: To detect a focusing state with reference to plural different areas without obstructing the miniaturization of a detector by constituting a separator lens so that the image of an object may be formed on plural photodetecting elements by a pair of separator lenses. CONSTITUTION: A photodetecting means is provided with plural photodetecting elements so as to receive luminous fluxes from plural areas different in object areas, the image of the object is formed on plural photodetecting elements by a pair of separator lenses 15. That is, the light transmitted through a photographic lens 1 is partly reflected by a main mirror 2, then, goes toward a finder part 5, the remaining light is transmitted through the translucent part of the main mirror 2, then, goes toward an automatic focus detecting module 4 after being reflected by a submirror 3. Only the light in plural focus detecting sensitivity areas, for example, in four zones are transmitted through a visual field mask 12. After the light is deflected by a mirror 14 at an angle of 90 deg. after transmitted through a condenser lens 13, the pupil division of the light is executed by an image re-forming lens 15, then, two images of respective standard parts PAL1 to PAL4 and reference parts PAR1 to PAR4 are formed on a photoelectric conversion element 16.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a focus detecting device for an optical system in a camera or the like.

[0002]

2. Description of the Related Art Conventionally, a structure has been proposed in which image light of an object is separated and re-imaged by a separator lens and an image is formed on a light receiving element to detect a focus state of an optical system with respect to the object. .

[0003]

However, when the above-mentioned structure is applied to the apparatus for detecting the focus state for a plurality of different areas of the target area, the number of separator lenses and the number of received light are the same as the number of areas to be detected. The device will be installed. For example, when it is desired to detect the focus state in four areas, four sets of separator lenses and four light receiving elements are installed. However, if the number of members is increased according to the number of regions, the following inconvenience occurs.

First, increasing the number of members hinders miniaturization. If the members are installed close to each other for downsizing, mutual light beams overlap each other, which may cause harmful light. Further, when the members are installed in each area, the performance (manufacturing accuracy / installation accuracy) of all the members must be made equal, which is substantially difficult. It is an object of the present invention to obtain an apparatus for detecting a focus state for a plurality of different areas of a target area without causing the above inconvenience.

[0005]

To achieve the above object, the present invention provides a separator lens for separating an image of an object into two images and re-imaging the image, and an image of the object formed by the separator lens. In a focus detection device having a light-receiving means for receiving light, the light-receiving means has a plurality of light-receiving elements so as to receive light fluxes from a plurality of different target areas, and the separator lens has a plurality of light-receiving elements formed by a pair of lenses. The feature is that an image of an object is formed on the element. Further, it is desirable to further have a condenser lens so that the light flux formed by one condenser lens enters the plurality of light receiving elements.

[0006]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A conventional device will be described for comparison with the embodiments of the present invention. FIG. 1A is a field of view showing a conventional focus detection area. In the conventional device, as shown in the figure, one narrow area in the center of the field of view is indicated by the pointing frame as the focus detection area, and the sensitivity range of the actual focus detection device is also extremely narrow in the center of the area corresponding to this field of view. It was limited to one area. Therefore, when performing automatic focus adjustment (hereinafter referred to as AF), the operator first sets the focus detection sensitivity range to the main object by aligning the indication frame in the center of the field of view with the image of the main object, regardless of his or her framing intention. After performing AF at the same time, it is necessary to fix the focused state, so-called AF LOCK, and then perform framing according to the intention. Also, when following a moving object, it is extremely difficult to always keep the main object in this focus detection sensitivity range, and a high-speed moving object often deviates from the focus detection sensitivity range, so AF operation is difficult. Becomes unstable. In the embodiment of the present invention, a device which does not require complicated operations, that is, a device having a plurality of focus detection sensitivity regions as shown in FIG. 1B is used. Hereinafter, preferred embodiments of the present invention will be described. FIG. 2 shows an optical configuration of the embodiment of the present invention. In FIG. 2 (a), which schematically shows the entire optical system of an embodiment suitable for a single-lens reflex camera, the present invention shows that part of the light transmitted through the taking lens (1) is reflected by the main mirror (2). Proceeding to the section (5), the rest passes through the semi-transparent portion of the main mirror (2), is reflected by the sub-mirror (3) and proceeds to the automatic focus detection module (4). The light directed to the finder section (5) is focused on the matte surface (7) and output to the photographer's eye through the pentaprism (9). A part of the finder light is repeatedly reflected by the diffraction grating (8) in the mat and guided to the spot photometric element (10) arranged on the side of the mat.
It is used as the detection light of the photometric element.

FIG. 2C shows the arrangement of the diffraction grating body and the spot photometric elements (BV1) to (BV4) on the matte surface (7). The diffraction grating body is arranged in four planes as shown in the figure, and reflects the light respectively incident from below toward the side edge of the mat, and the photometric elements BV1 to BV1 to the respective light exit ports.
BV4 is arranged.

Light transmitted through the main mirror (2) to the lower part of the camera body by the sub mirror (3) is an infrared cut filter.
(11), on the photoelectric conversion element (16) via the field mask (12) arranged near the focal plane, the condenser lens (13), the mirror (14), and the re-imaging lens (separator lens) system (15). Is imaged. The details are shown in FIG. 2 (b).

In FIG. 2B, the infrared cut filter (1
Light that has passed through 1) is a field mask (1
Reach 2). The field mask allows only the light of the four zones shown in FIG. 1 (b) to pass through. This light is a condenser lens (1
After passing 3), it is deflected 90 ° by the mirror (14), and then the re-imaging lens (15) divides the pupil, and the first zone is (PAL1).
And (PAR1), the second zone is (PAR2) and (PAR
2), the third zone is (PAL3) and (PAR3), and the fourth zone is (PAL4) and (PAR4). Two images of the standard part and the reference part are formed on the photoelectric conversion element. .

The image of each of the standard part and the reference part (PA
Lz) and (PARz) (z = 1 to 4) are in focus when the image distance Xz is a predetermined distance Lz, and when Xz> Lz, the subject is close to the lens position and Xz <Lz.
At that time, the subject is far from the lens position. FIG. 2 (d) is an expanded view of the optical system of FIG. 2 (b).

Next, FIG. 3 shows an electrical configuration of this embodiment. In this embodiment, a microprocessor for controlling the entire camera (hereinafter referred to as control microcomputer) (COP), a microprocessor for AF control (hereinafter referred to as AF microcomputer)
(AFP). (S1) is photometry and A
A start switch for starting F, (S2) is a release switch for activating the photographing operation of the camera, and (S4) is a switch which is turned off by charging the shutter curtain of the main mirror and the focal plane shutter and turned on when the exposure is completed. The open / close signal is also input to the control microcomputer (COP).

The aforementioned spot photometric elements (BV1) to (BV1)
The output of 4) is controlled by a multiplexer (AEMP).
It is selectively output by a selection signal AEMPS from (COP) and input to the control microcomputer (COP) as a value digitized by the A / D converter (AEAD). Control microcomputer
(COP) is a lens data output circuit (LDM), which is used for AF, such as a conversion coefficient for converting the defocus amount detected by the automatic focus detection unit into the lens extension amount according to each lens, and the maximum aperture. Value, minimum aperture value, etc.
Input (LDS) and transfer only the data required for AF to the AF microcomputer (AFP).

The control microcomputer (COP) inputs data from the ISO data output means (SVM) which outputs the data of the apex value Sv of the film sensitivity. Control microcomputer (C
OP) calculates the exposure based on these input data, outputs the exposure value signal (AES) to the exposure display device (AED) for display, and after the release signal of the release switch (S2) is input, the exposure is performed. The control signal (BCS) is output to the exposure controller (BCR) to control the exposure.

On the other hand, the AF control microcomputer (AFP)
AF consisting of CCD via interface (AFIF)
The sensor (CCD) is driven, the output of the AF sensor (CCD) is subjected to analog processing and A / D conversion by the AF interface (AFIF), and digital image information is input.
AF calculation is performed according to the input information to calculate the defocus amount.

Further, the AF microcomputer (AFP) converts this defocus amount into a lens extension amount based on the lens data sent from the above-mentioned control microcomputer (COP), and drives the motor (MO) to the motor encoder (ENC). ) Output
While confirming the amount of rotation with (DCL), drive with a motor driver (MDR) with a motor drive signal. Further, the AF microcomputer (AFP) sends a focus state signal (FAS) to the focus display device (FA) to confirm the focus state.
It is output to D) and the in-focus state is displayed.

Next, the exchange of signals between the control microcomputer (COP) and the AF microcomputer (AFP) will be described. (AFS
T) is from the control microcomputer (COP) to the AF microcomputer (AF
AF start signal sent to P) to start the AF operation. When this signal (AFST) changes from “H” level to “L” level, the AF microcomputer (A
FP) starts the AF operation.

(AFE) is an AF end signal for transmitting from the AF microcomputer (AFP) to the control microcomputer (COP) that the AF operation is completed and the focus state is achieved.
When (AFE) becomes the “H” level, it is transmitted that the AF is in the end state. (AFSP) is a control microcomputer
The AF stop signal sent from (COP) to the AF microcomputer (AFP) to stop the AF operation.
When a pulse is input to (AFSP), the AF microcomputer (AFP) stops the AF operation.

Further, (AFZS) is an AF zone selection signal which becomes "H" level when any one of the above-mentioned four zones is selected, and (SZS) is a signal showing the selected zone. Is. (LDTS) is a control microcomputer
(COP) is an AF lens data bus for transferring to the AF microcomputer (AFP) only the data necessary for the AF operation among the lens data (LDS) input from the lens data output circuit (LDM).

The operation flow of these constituent elements of this embodiment will be described for each of the control microcomputer (COP) and the AF control microcomputer (AFP) with reference to FIGS. 4 and 5. Switch (S1) is turned on by pressing the release button to the first step.
And the interrupt terminal of the control microcomputer (COP)
An interrupt signal is applied to (INT0) (# 1 in FIG. 4). This signal causes the control microcomputer (COP) to exit from the stop mode, set the AF start signal (AFST) to "L" to operate the AF control microcomputer (AFP) (Fig. 4 # 2), and start the photometric operation (Fig. 4 # 3). Next, control microcomputer
(COP) inputs data required for exposure calculation. That is, Sv data is input from the Sv value output means (SVM), and various lens data is input from the lens data output means (LDM).
(# 4 in FIG. 4), only lens data necessary for AF is output to the AF control microcomputer (AFP) (# 5 in FIG. 4), and further photometric data is input.

Next, the control microcomputer (COP) inputs the AF zone selection signal (AFZS) from the AF microcomputer (AFP) and determines whether it is "H"(# 7 in FIG. 3). This signal
(AFZS) will be described later, but since "L" is output at the beginning of the operation, the case of "L" will be described here.

When the AF zone selection signal (AFZS) is "L" (when the AF zone is not selected), since the AF zone is not selected, the main subject cannot be limited and the photometric element cannot be selected. Photometric data (BV1)
The average of (BV4) is taken as a photometric value (# 8), and exposure calculation is performed from each data (# 11 in FIG. 4). When the exposure calculation is completed, the control microcomputer (COP) outputs the result to the exposure display device and displays it (# 12). Upon completion of the above-described one-loop operation, it is determined whether or not the switch (S1) is continuously pressed. If the switch is pressed, whether the shutter charge is completed (# 14) or the focus state ( # 15)
If both are satisfied, the release enable state is set, interrupts from the interrupt terminal (INT1) are enabled (# 16), and a loop is formed again after re-inputting each data (# 4). , If either one is not satisfied, re-enter each data without setting the release permission state (# 4)
To form a loop.

When the switch (S1) is not pressed, the metering and display are stopped, and the AF stop signal (AFSP) is output to stop the AF operation.
The F start signal (AFST) is set to "H" level. Next, enable the interrupt from the terminal (INT0) and switch (S2)
Disable the interrupt of the terminal (INT1) from the
The IF is reset to 0 and then the stop state is entered.

On the other hand, the operation of the AF microcomputer (AFP) is performed by the AF start signal (A) sent from the control microcomputer (COP).
FST is the interrupt terminal (INT) of AF microcomputer (AFP)
When the voltage is applied to A) (# 30 in FIG. 5), the operation is exited from the stop mode and started. The AF microcomputer (AFP) first drops the AF end signal (AFE) to "L" and sets the AF zone selection signal (AFZS) to "L" to output to the control microcomputer (COP) that the zone is not selected during AF operation. Flag L set to 1 when the lens is driven
DF is set to 0 (# 31 in FIG. 5).

Next, after initializing the CCD which is the AF sensor (# 32 in FIG. 5), the variable Z indicating the number of AF zones is set to 4 (# 33) and the control microcomputer (CO) is set.
Input the lens data required for AF operation from (P) (# 3
4). Next, the CCD is controlled. First, CCD integration is performed, and a shift pulse is applied when the integrated light amount reaches an appropriate level, or when the subject brightness is low, a preset maximum integration time is reached, and the CCD data, that is, the image. Enter information as a digital value
(Fig. 5 # 35). This operation will be described in detail later, but here, CCD data for all zones 1 to 4 are input.

Next, a low contrast (hereinafter abbreviated as low contrast) flag indicating whether or not the contrast of the object is low is set (# 36 in FIG. 5). This flag is cleared only when focus detection was possible during the previous CCD integration, and this flag is set because it was the first CCD integration here. This flag is used later for determining whether to perform low contrast scanning or to perform focus detection operation again with the lens position unchanged. Here, the low contrast scan is, when the contrast of an object is low at a certain lens position, seeks a lens position where the contrast is high, and drives the lens, for example, one round trip over the entire driving range.

Next, data preprocessing (# 37) is performed in order to determine the priority order for performing the focus detection calculation for the four zones.
~ # 57), pre-correlation (# 57- # 72), pre-correlation low contrast discrimination (# 73- # 81), zone prioritization (# 83
~ # 94) is performed. Although these operations will be described in detail later, only the zone including the closest subject among the subjects included in each zone, that is, the zone with the largest image interval calculated in each zone, is selected. This is because focus detection is performed, and if the main correlation calculation is performed for all zones, the calculation time becomes long, so that the calculation time can be shortened. In # 82, it is checked whether or not the variable Z is 0. If the variable Z is 0, it means that there is a low contrast in all AF zones.

Since the low contrast discrimination here is repeated once more after the main correlation, which is performed later, the discrimination is simple and the discrimination region of the low contrast is narrow. In this way, a more accurate focused state detection calculation is performed for the zone selected by the pre-correlation (# 96 to # 105). Based on this correlation calculation, a low contrast check is further performed, and it is determined that the selected zone is not a low contrast, the defocus amount is calculated (# 112), or all zones are determined to be low contrast. This correlation calculation and low contrast discrimination are performed for each zone according to the ranking. When all zones are determined to be low contrast and the low contrast flag is set, it is considered that the defocus amount cannot detect the focused state because the lens position is extremely far from the position where the subject is focused. By changing the lens position, CCD integration and calculation are repeated many times during one reciprocating movement from the closest shooting distance to infinity, and a low-con scan is performed to search for a lens position where the focus state can be detected (# 110).
# 33).

When the defocus amount is calculated as not being a low contrast and the defocus amount is calculated, the low contrast flag is first cleared to store this state (# 113), and even if a low contrast is obtained in the next integration, The lens is not driven, and reintegration and recalculation of all zones of the CCD are performed at the lens position as it is. This is because there is no change in the distance between the main subject and the camera, and when the main subject deviates from the zone containing the main subject the previous time, it becomes low-con in the previously selected zone, This is to prevent the lens position that has been in the position from largely changing.

Next, the AF control microcomputer (AFP) uses the AF control microcomputer (COP) to specify the photometric zone to the control microcomputer (COP).
The zone signal (SZS) selected by the control microcomputer (AFP) is output to the control microcomputer (COP), and the AF zone selection signal is output.
(AFZS) is set to High and output. After this, the flow on the control microcomputer (COP) side is # 7 for the AF zone selection signal AFZ.
When it comes to the branch by S (# 7), A
An F zone signal (SZS) is input (# 9), and spot photometry calculation is performed based on the photometric element output (# 10) of the photometry zone.

Here, the AF zone selection signal in # 7 of FIG.
When (AFZS) is determined to be "H", it is determined in # 7-1 whether the AF end signal (AFE) is "H". And
When the AF end signal (AFE) is "H" and the in-focus state is reached, it is determined whether or not the flag BIF is 1 in # 7-2. If this flag BIF is not 1, # 7-3
Then, the flag BIF is set to 1 and the process proceeds to # 9.
On the contrary, if the flag BIF is 1, the process proceeds to # 11 without updating Bvc in # 10. Therefore, the AF zone selection signal (AFZ
The photometric data corresponding to S) will be AE-locked. If the AF end signal (AFE) is not "H" in # 7-1 and the lens has not reached the in-focus state, the flag BIF is reset to 0 in # 7-4 and the process proceeds to # 9.

Next, the AF control microcomputer (AFP) determines whether the calculated defocus amount is within the preset focusing zone (# 115), and determines that the lens is present within the focusing zone. At this time, the AF end signal (AFE) is set to "H", the control microcomputer (COP) is instructed to complete the AF operation, the focus display is performed, and the release permission is prompted (#
121, # 123). On the contrary, when it is out of the focusing zone, the lens extension amount is calculated as the pulse count (LEP) of the encoder by using the conversion coefficient for converting the defocus amount into the lens extension amount input previously (# 116), and the counter ( PC) to drive the motor by the calculated pulse count number (# 117, # 118, # 119), move the lens position by the calculated lens extension amount, and stop the motor.
(# 120). After this, it is necessary to re-integrate the CCD and check again whether it is in-focus or not. At this time, in order to reduce the operation time, the CCD is re-targeted only for the CCD selected by the previous calculation. Integrate and output data
(# 127).

Prior to this, the variable Z is set to 1 and the data (LDTS) necessary for the AF operation is input to the AF microcomputer (AFP) in order to determine the low contrast of only the selected CCD (# 125, # 126). After that, the main correlation calculation is performed only for that block, and the lens is extended based on the in-focus / out-of-focus determination. In the unlikely event that a low contrast is determined at this stage, the operation from the integration of CCDs in all zones is repeated with the lens position left unchanged as described above. The above is the basic operation of the autofocus camera having means for performing appropriate automatic focus adjustment irrespective of the position of the main subject in the finder field of view and exposure control for the main subject by spot metering.

Next, a supplementary explanation will be added to the portions where the explanation is omitted. First, the data preprocessing, precorrelation, precorrelation low contrast discrimination, and zone prioritization portions provided for shortening the calculation time will be described with reference to FIGS. 6, 7, 8, and 9, respectively.

First, the data preprocessing routine shown in FIG. 6 will be described. The AF microcomputer (AFP) first sets a variable Z indicating the number of AF zones to 1, a variable C (Z) indicating a contrast value to 0, and a variable j indicating the number of times of contrast calculation to 0. Yes (# 37, # 3
8, # 39). Next, the difference of the A / D converted data between the adjacent pixels of the CCD which is the reference part is taken, it is judged whether this difference is positive or negative, and the data Ldj is signed for each judgment result. Memory (# 40, # 41, # 4
2). That is, the output data of each pixel of the reference portion is LDj
Then, in # 40, LDj (Z) -LDj + 1 (Z) ... (1) is calculated to determine whether the result is positive or negative. If this result is positive, the variable j
The value Ldj corresponding to is set to 1, and conversely, if the result is negative, the value Ldj is set to 0 in # 42.

Next, in # 43, the same calculation as in equation (1) is performed and the result is used as the contrast value C. In # 44, this absolute value | C | is added to the contrast value C (Z) up to the previous time. , The total contrast value C (Z) up to the obtained difference data is obtained. Then, in # 45, 1 is added to the variable j, and in # 46, the operations of # 38 to # 45 are repeated until the variable j becomes k−1 (where k is the number of pixels of the reference portion).

When j becomes equal to k-1 in # 46, whether or not the variable Z indicating the number of AF zones is 4 in # 47 in order to perform the operations of # 38 to # 45 for all four AF zones. To determine. If the variable Z is not 4, the variable Z is incremented by 1 in # 48 and the process returns to # 38.
The operations of # 38 to # 47 are repeated until 4 becomes 4.

When the variable Z becomes 4 in # 47, the process proceeds to # 49. In # 49 to # 57, # 37 to # 47 for the reference part
By the same method as (excluding # 38, # 42, and # 43), the contrast value is obtained for all four AF zones from the pixel data serving as the reference portion. Here, l is the number of pixels in the reference portion, and whether the difference data is positive or negative is stored as Rdj.

As described above, the contrast values C (1) to C (4) of the first to fourth reference portions and the reference portion difference code data Ldj (1) to Ldj.
dj (4), [j = 1 to k-1], reference part difference code data Rdj (1) to Rdj (4), [j = 1 to l-1] are prepared, and the preprocessing work is completed ( (Figure 6).

Next, a flowchart of the pre-correlation routine is shown in FIG. 7 and will be described. AF microcomputer (AF
P) sets a variable Z indicating the number of AF zones to 1 in # 58, and 1-bit difference data (Ldj) of the reference portion in # 59.
On the other hand, the variable n indicating the number of shifts when the 1-bit difference data (Rdj) in the reference part is shifted by one pixel to obtain the correlation value is set to 1. Furthermore, AF microcomputer (AFP)
Sets the variable hn (Z) indicating the degree of correlation to 0 in # 60, and sets the variable j indicating the number of operations performed when obtaining one correlation value in # 61 to 0.

Then, in # 62, the difference between the difference data (Ldj) of the 1-bit reference part and the difference data (Rdj) of the 1-bit reference part is calculated, and when both data are not the same (that is, the difference is not 0). Time) and the correlation is not good,
At 63, 1 is added to the variable hn (Z). When both data are the same, # 63 is skipped. This # 62, # 63
The above operation is performed for the number of contrasts (k-1) obtained in the reference section (# 64, # 65).

Further, the AF microcomputer (AFP) carries out an operation of calculating the shift number for which the maximum correlation is obtained. First, in # 66, it is determined whether or not the variable n is 1 (indicating that it has not been shifted). If n = 1, in # 68 the variable Mhn (Z) indicating the correlation degree is set to the variable hn (Z). Z) is set, and the image interval error Mn
(Z) is calculated by n-Lz. Here, Lz is an image distance in a focused state. On the other hand, when the variable n is not 1 in # 66, the correlation value Mhn (Z) stored in # 67 is compared with the hn (Z) obtained in the current calculation.

If the correlation value hn (Z) at this time is smaller than the stored correlation value Mhn (Z), it is determined that the correlation is high, and the process proceeds to # 68, and the correlation value at that time is determined. And an image interval error is calculated. On the contrary, this correlation value hn
If (Z) is not smaller than the stored correlation value Mhn (Z), # 68 is skipped. # 6 like this
The operations of 0 to # 68 are performed 1−k + 1 times to obtain the minimum correlation value (maximum correlation degree) and the image interval error at that time (# 69, #
70).

Further, the operations of # 59 to # 70 are performed by four AFs.
The pre-correlation routine ends after the minimum correlation values Mhn (1) to Mhn (4) and the image interval errors Mn (1) to Mhn (4) at that time are obtained for each AF zone. 71, # 72). Here, in # 71, the variable Z is 4
When the above calculation is completed for all the AF zones, the routine shifts to the pre-correlated low contrast discrimination routine shown in FIG.

In the pre-correlation low contrast discrimination routine of FIG. 8, low contrast discrimination is performed on the calculation result of the pre-correlation routine. First, the AF microcomputer (AFP)
In order to perform low contrast discrimination for the F zone,
In step # 73, the variable j is set to 1. Then, in # 74, the contrast value C (j) calculated for each AF zone is calculated.
Is greater than the predetermined value CS, and in # 75, it is determined whether the minimum correlation value Mhn (j) obtained in the previous correlation routine is less than the predetermined value SM. Then, for each AF zone, the contrast value C (j) is the predetermined value C
If the minimum correlation value Mhn (j) is greater than S and the minimum correlation value Mhn (j) is less than the predetermined value SM, it is determined that focus detection is possible for the AF zone, and at # 76, the focus corresponding to the zone is detected. The conzone flag LZF (j) is reset to 0.

On the other hand, the contrast value C (j) is the predetermined value CS
If it is less than or equal to or more than the minimum correlation value Mhn (j) is greater than or equal to the predetermined value SM, 1 is subtracted from the variable Z set to 4 as the initial value in # 78, and the AF zone is changed in # 79. Regarding, it is judged that the focus cannot be detected,
Low control zone flag LZF (j) corresponding to that zone
Set to 1. Then, in order to perform the operations of # 74 to # 79 for all the AF zones, it is determined in # 80 whether or not the variable j is 4, and if not 4, 1 is added to the variable j in # 81 and the process returns to # 74. To do.

If the variable j becomes 4 in # 80, it is determined whether or not the variable Z indicating the number of AF zones for which focus detection is determined in # 82 is 0. If this variable Z is 0, it is determined that focus detection is impossible for all AF zones, and the process proceeds to step # 109 in FIG. 5. If variable Z is not 0, there is an AF zone in which focus detection is possible. In order to determine the priority order of the AF zones for which the main correlation is made, the process proceeds to the zone priority order routine of FIG.

In the zone priority order routine shown in FIG. 9, the AF microcomputer (AFP) first sets the variable j to 1 in # 83 and sets the variable M1 for storing the image interval error.
Set ~ M4 to -LZ and the variable Q to 0.
Then, at # 84, it is determined whether or not the low contrast zone flag LZF (j) corresponding to each AF zone is set. Here, if the low-conzone flag LZF (j) is set for a certain AF zone, it is not necessary to determine the priority order for the main correlation for that AF zone, so proceed to step # 94 to set variable j. AF with 1 added
To determine the locon zone flag corresponding to a zone #
Return to 84.

When the low contrast zone flag LZF (j) corresponding to each AF zone is not set at # 84, the operations at # 85 to # 93 are performed, and only the AF zone where focus detection is possible is performed. The images are ranked in descending order of the image interval error, that is, in ascending order of the subject distance corresponding to the detected focus position, and the image interval error is also stored in correspondence with the ranking. That is, among the AF zones in which focus detection is possible, the image distance errors corresponding to the AF zone determined to have the shortest subject distance are stored as M1, M2, and M3 in order, and the AF zone numbers are also B1, B2, and in that order.
It is stored as B3 and B4.

The pre-processing, pre-correlation, pre-correlation low contrast discrimination, and zone prioritization have been described in the present embodiment. In addition to these, the CCD is determined by a predetermined value or the average output value of CCD data. Instead of the pre-processing that binarizes the data or the calculation (subtraction) of step # 97 in FIG. 10 for obtaining the correlation value of this correlation, the exclusive OR of the two data is taken, and the result is added to obtain the minimum value. The same function can be realized by means such as obtaining the shift position and performing the pre-correlation.

Next, a detailed description will be given of the procedure of this correlation. The correlation value is evaluated by the sum of the non-binarized output value differences between the reference pixel and the reference pixel in the zone designated by the previous correlation. This correlation value H (P) is obtained by shifting the reference portion pixel by 1 pixel to 1−k + 1 with respect to the standard portion pixel column (# 96 to # 99), and the minimum value H (PM) among them is obtained ( # 101).

Next, the AF microcomputer (AFP) performs interpolation calculation to obtain the true minimum correlation value. First, the AF microcomputer
In (AFP), the shift amount PM is 1 or 1-in # 102.
It is determined whether or not k + 1. When the shift amount PM is not 1 or 1−k + 1, # 103
In step # 104, the minimum correlation value YM is obtained by subtracting the image interval LBi at the time of focusing from the result of the interpolation calculation in
Ask for. On the other hand, when the shift amount PM is 1 or 1−k + 1 in # 102, the interpolation calculation cannot be performed.
In step # 105, the image distance XM is obtained from the shift amount PM, and the minimum value H (PM) calculated in step # 101 is used as it is as the minimum correlation value YM. In this calculation, the image interval having the minimum correlation value can be predicted in advance from the result Mn (Bi) of the pre-correlation. Therefore, calculation is performed only in the vicinity of the predicted image interval in the designated zone, and the calculation time It is also possible to shorten.

Based on the thus obtained minimum correlation value YM and the image interval at that time, low contrast discrimination is performed again. Here, the condition is that the value obtained by dividing the minimum correlation value YM by the contrast value obtained in the preprocessing is equal to or less than a predetermined value. If the value is equal to or larger than the predetermined value, the zone is considered to be a complete low contrast zone (# 106-1).

When the defocus amount is obtained in the first loop, the lens is already driven and the lens is moved to the vicinity of the in-focus state, and in the case of the second or more calculation work in the reintegration of the CCD (# 106- (2, LDF = 1) Considering that the target subject is a moving subject, when the defocus amount suddenly increases (when it becomes a certain value D or more), it is determined that the subject is out of that zone, and that zone is set to low contrast. It is determined that the state is the state, and reintegration of all zones is performed (# 107). On the contrary, if the absolute value of the image interval XM is less than the constant value D in # 107, the process proceeds to # 112 in FIG. 5 to calculate the defocus amount, and then the lens is driven.

Since the pre-correlation is simply a correlation, there is a possibility that a large difference may occur in the defocus amount between the special correlation and the special correlation in the first loop. At this time, there is a possibility that the main subject is closer to the camera than the subject in the other focus detection zones. Therefore, in this embodiment,
As an example, the image interval of the main correlation is subtracted from the image interval of the preliminary correlation (here, the variable q is set to 0 as an initial value (# 95 in FIG. 5), so that q = 1 in # 106-3. , # 106-4 to # 106-5.) The result is larger than 1, that is, 1 pit compared with the previous correlation result.
If the subject is far from the camera in this correlation over ch
The calculation result is stored in (# 106-5), and after the main correlation of the selected zone is finished (# 106-4),
Compare the main correlation results of the first and second selection zones
(# 106-7), a zone having a large image distance is selected, and the defocus amount is obtained according to the image distance calculation result, and the lens is driven.

On the contrary, when the subtraction result is smaller than 1, it is considered that the correct zone is selected by the pre-correlation from the beginning, the defocus amount is obtained according to the result of the main correlation image interval calculation, and the process proceeds to the lens driving. In FIG. 11, the same operation is performed, but the subtraction target in # 106-5 is performed by the second selected pre-correlation image interval amount instead of the pre-correlation image interval amount of the zone (# 106- 5 '), achieves exactly the same effect as described above.

With the above, the explanation of the flow of the whole operation in this embodiment is completed, and the electric circuit configuration, the AF sensor (CC
The detailed configurations of D) and the AF interface (AFIF) will be described.

FIG. 12 shows an AF sensor according to this embodiment.
Two configurations of the CCD used as (CCD) will be illustrated. FIG. 12 (a) shows a configuration in which output CCD registers are arranged in series, and FIG. 12 (b) shows an example in which output registers are arranged in parallel, both of which are one-chip CCDs.

First, the structure common to FIGS. 12A and 12B will be described. The images of the first block to the fourth block are pupil-divided, and the reference portion photodiode array (P
Images are formed on AL1) to (PAL4) and on the reference photodiode arrays (PAR1) to (PAR4) as a reference image. Here, each photodiode array includes a storage unit corresponding to the diode array. The standard photodiode array has k pixels, and the reference photodiode array has m pixels (k <m).

Reference part photodiode array (PAL1)
~ (PAL4) Photodiode for subject brightness monitor for the purpose of CCD integration time control near each
(MP1) to (MP4) are respectively arranged, and the photocurrents generated in the photodiodes (MP1) to (MP4) are charged to approximately the power supply level according to the integrated clear gate pulse (ICG). The electric charge of C4) is reduced with an inclination proportional to the amount of incident light. The voltage of this capacitor is output to the outside via a buffer with high input impedance and low output impedance (AGCOS1-4).
Is output as

The integrated clear gate pulse (ICG) is applied to the MOS gate provided between the storage unit (photodiode array) of each pixel and the power supply, and the integrated clear gate pulse (ICG) is "H". At the same time, the storage unit is charged up to almost the power supply voltage level and cleared. After this, when the integration clear gate pulse (ICG) is "L", the MOS gate (MOS)
Is in the open state, and the charge of the storage unit charged to the power supply voltage is discharged by the photocurrent proportional to the image brightness distribution generated in the photodiode array, and the brightness distribution information of each pixel is stored.

A MOS gate is provided between the charge storage section and the register for each pair of the standard section and the reference section of each block.
(MOS) is installed and each gate is closed when “H” of SH pulses (SH1) to (SH4) is applied, and the charges accumulated after the integration clear gate pulse (ICG) is applied to the storage unit are transferred to the register.

Monitor outputs (AGCOS1) to (AGCOS1)
A DOS circuit is installed as the compensation output of S4).
This circuit is formed with the same characteristics as the capacitor and buffer of the monitor output part, and the input end of which is in the OPEN state, and the potential charged up to almost the power supply voltage in response to the integral clear gate pulse (ICG) is applied. Output is continued even after the integration clear gate pulse (ICG) disappears.

Next, the differences between the configurations of FIGS. 12A and 12B will be described. 12 (a) and 12 (b) differ in the configuration of the CCD register and the configuration of the output stage of the CCD that follows. In FIG. 12 (a), CCD registers are arranged in series with respect to each zone, and an output buffer is provided at the end of the CCD register, the output of which is the first zone reference section, the second zone reference section, the first zone. The reference portion, the second zone reference portion, the fourth zone reference portion, the third zone reference portion, the fourth zone reference portion, and the third zone reference portion are sequentially output in synchronization with the falling edge of the transfer clock φ1.

On the other hand, the output configuration of the CCD image sensor shown in FIG. 12 (b) is a parallel configuration in which each zone has a different register, and a total of four outputs have buffers at the end of each CCD register. There is. First
From the buffer outputs of ~ 4, the outputs of the standard part and the reference part of the first to fourth zones are sequentially output in synchronization with the falling edge of the transfer clock φ1. Also, with this CCD image sensor, 4
Figure 1 to control the integration time in different zones
In 2 (a), pixels that are shaded by an aluminum mask are provided on the output end sides of both the standard part and the reference part of each zone (shaded areas), and pixels for dark output level correction that vary greatly with temperature and integration time are provided. Used as. In FIG. 12B, the dark output level correction pixel (hatched portion) is installed only on the output end side of each reference portion and is used for correction of both the reference portion and the reference portion.

Next, the circuit structure of the AF interface (AFIF) and the concrete driving method of the CCD image sensor will be described. First, the series CC shown in FIG.
A driving method of the CCD image sensor having the D register will be described with reference to FIG. In FIG. 13, the left side of the drawing is C
Connection part with CD image sensor, AF microcomputer on right side
(AFP) connection part. The first CCD integration after the start of the AF operation requires the output of all zones. At this time, the integration is started by applying an integration clear gate pulse (ICG) from the AF microcomputer (AFP). The application of this pulse initializes the all-pixel storage portion of the CCD and the monitor output, and after the disappearance of this pulse, both start storage of photoelectric conversion output at the same time.

On the other hand, the AF timing control circuit (AF) to which the original clock φ0 supplied from the AF microcomputer (AFP) and the clock φa obtained by dividing the clock by a plurality of stages are input.
In TC), when an all-zone output command is supplied from the AF microcomputer (AFP) with a zone selection symbol (AFZS), the transfer clock selects φa as a cycle at which the subsequent A / D conversion is possible. The integral clear gate pulse (ICG) is also input to the reset input of the R / S flip-flop,
By resetting the R / S flip-flop, the transfer clock φ1 to the CCD is fixed to the “H” state and φ2 is fixed to the “L” state. In this state, the accumulation of the pixel accumulating portion progresses, and at the same time, the accumulation of the monitor also starts to generate a monitor output which drops by a constant level V1 from the compensation output. At this time, the value of V1 is set in advance in that the charge accumulated in the pixel accumulating portion has an average output level suitable for A / D conversion in the subsequent stage and focus detection calculation.

The level V1 is exceeded in order from the zone with the highest subject brightness, and the comparators (COM11) to (CO
M14) is inverted, and its output is supplied as shift pulses (SH1) to (SH4) to the CCD image sensor through the OR gate and the one-shot pulse generator. The shift pulses (SH1) to (SH4) to the image sensor shift the charges in the pixel storage unit to the transfer registers, but since the transfer clock is not supplied to the registers, the charges are held in the potential of the register corresponding to the pixel. To be done. In this way, the comparator (COM11)
When the inversion of (COM14) is completed, that is, when the output (TINT) of the AND gate becomes "H", the output of each zone that has obtained the output of the appropriate average level is stored in the register. .

Here, the inversion of the output (TINT) of the AND gate is input to the AF microcomputer (AFP) as an integration completion signal of all zones of the CCD image sensor and to the R / S flip-flop via the OR gate and the delay element. As a result, it is used as a transfer clock application start signal. 14
Shows the time chart. Then from the OS terminal φ1
Each pixel output is output in synchronization with the falling edge of
The timing control circuit counts φ2 to generate a sampling signal at each timing when outputting the dark-time output correction pixel, and also supplies an AD start signal (ADS) to the AD converter.

Thus, the output of the CCD is the first zone reference section, the second zone reference section, the first zone reference section, the second zone reference section, the fourth zone reference section, the third zone reference section, the fourth zone reference section, After the dark time output correction is performed according to the respective integration times in the order of the third zone reference part, it is A / D converted, output in synchronization with the AD conversion completion signal, and input to the AF control microcomputer (AFP). Will be done.

Next, in this circuit, step # 127 in FIG.
The integral drive of the selected zone shown in 1) will be described. First, the zone signal (SZS) changes to the AF timing control circuit (A
When it is transmitted to the FTC), the number of transfer clocks required until the zone is output is set in the counter inside the circuit. After applying the integration clear gate pulse (ICG),
The AF microcomputer (AFP) outputs the monitor comparators (COM11) to (COM14) of the block to be output (I
NT1) to (INT4) are selected, a manual shift signal (SHM) is generated at the same time as the comparator is inverted, and the stop of the transfer clocks φ1 and φ2 is released.

The AF timing control circuit (AFTC) in which the counter is set counts the clock φa, and the original clock φ is counted until the counter becomes equal to the set value.
0 is supplied to the CCD, and an A / D conversion clock is supplied only when the output of the selected zone is output.
Data for only that zone is supplied to (AFP) in synchronization with (EOC), counter setting is performed, and high-speed transfer is performed when other zones are output, and when there are remaining pixels in the zone of interest. Performs the same operation. By doing so, the dead time of the data dump time and the integration time is reduced, and the speed of the AF operation is increased. A time chart of this operation is shown in FIG.

Finally, the parallel CCD shown in FIG. 12 (b).
FIG. 1 shows a driving method of a CCD image sensor having a register.
This will be described using 6. The left side of the drawing is the CCD image sensor, the right side of the multiplexer (MPX) is the AF interface (AFIF), and the terminal row at the right end is the AF microcomputer (A
FP).

With this CCD image sensor, the output of all zones of the CCD for the first time after the start of AF can be obtained by shortening the time by the following driving method. First, the AF microcomputer (AFP) uses an integration clear gate pulse (ICG) to eliminate charges accumulated in each pixel storage unit and monitor.
Generate. At this time, the output signal of the multiplexer (MPX) is changed by the zone signal (ZS) indicating the first zone.
The input signal (AGCOS1) is output from (AGCOS0), the input signal (SH0) is output from the output terminal (SH1), and the input signal (OS1) is output from the output terminal (OS0). .

Then, the charge accumulation of the CCD image sensor for the first zone is monitored by the multiplexer.
Comparing the signal (AGCOS 1) via (MPX)
This is done by monitoring (COM20). When the charge accumulation in the monitor section and each pixel section in the first zone progresses and the signal (AGCOS1) reaches a level V1 suitable for the subsequent analog processing circuit and the subsequent focus detection calculation, the output of the comparator (COM20) is inverted. Shift pulse
(SH0) is supplied as a shift pulse (SH1) to the CCD image sensor in the first zone via the multiplexer (MPX).

When the preset maximum integration time elapses without the signal (AGCOS1) reaching the level V1, the shift pulse (SH0) is applied by the application of the manual shift pulse (SHM) from the AF microcomputer (AFP). Are supplied to the CCD image sensor in the first zone as shift pulses (SH1) via the multiplexer (MPX).
The supply of this shift pulse (SH1) causes the CCD image sensor in the first zone to end the charge accumulation operation, and the charge accumulated in the pixel accumulation section is shifted to the CCD shift register (Rg1) in the first zone via the shift gate. To be done.

Here, the delay pulse generating the shift pulse (SH0) and the input signal of the one-shot circuit (DO) are φ1,
It is also supplied to the transfer clock generation circuit (TCG) that generates two transfer clocks of φ2, and the transfer clock φ1 is “H”.
The phase is adjusted so that the shift pulse (SH1) is supplied to the CCD image sensor in the first zone within the level section. Then, in synchronization with the fall of the transfer clock φ1, the photoelectric conversion output (OS1) of the image accumulated in the CCD image sensor in the first zone is transferred to the multiplexer (M
It is sequentially output via the output terminal (OS0) of PX).

Immediately after the generation of the shift pulse (SH0), the AF microcomputer (AFP) supplies the second integral clear gate pulse (ICG) to the CCD image sensor. This second integrated clear gate pulse (ICG) is the second
This is an integration start signal for the CCD image sensor of the zone, in order to continue the charge accumulation operation of the monitor section and the pixel section of the second zone and the discharge operation of the accumulated charge immediately after the end of the charge accumulation operation of the first zone. belongs to.

After that, the AF microcomputer (AFP) causes the sample and hold circuit (S / H) to store the output of the dark output correction pixel in the photoelectric conversion output (OS1) for the first zone, and then The difference between the output of each pixel output and the output of the stored dark output correction pixel is A / D converted and input as image information.

Here, when the charge accumulation of the CCD image sensor is forcibly terminated by the manual shift pulse (SHM) from the AF microcomputer (AFP), the output of the monitor section is output by the outputs of the comparators (COM20) to (COM22). The automatic gain adjustment circuit (AGC) automatically adjusts the gain in accordance with the average accumulation level of. That is, the automatic gain adjustment circuit (AGC) has a photoelectric conversion output.
(OS0) and the output of the sample and hold circuit (S / H) are input, and the difference between both outputs is appropriately amplified and output. The output of the automatic gain adjustment circuit (AGC) is A /
It is input to the D conversion circuit (ADC) and converted into a digital value, and this digital value is used as image information in the AF microcomputer (A
FP).

The image information of the first zone is thus detected by the AF
When input to the microcomputer (AFP), next, the charge accumulation state of the CCD image sensor in the second zone in which the charge accumulation has been started is detected. For this purpose, first, A
The F microcomputer (AFP) sets the signal (TINTC) to "L" and the manual shift pulse (SHM) shifts to the shift pulse (SH
0) output as a zone signal (ZS)
Is switched from the first zone to the second zone. As a result, the output terminal (AGCOS0) of the multiplexer (MPX)
The input signal (AGCOS2) is output from the
(SH0) is output from the output terminal (SH2)
It is set so that the input signal (OS2) is output from (OS0).

Then, the AF microcomputer (AFP) sends a signal (T
(INT0) is confirmed. When the signal (TINT0) is “H”, the charge accumulation of the CCD image sensor in the second zone is already excessive.
(ICG) is supplied to the CCD image sensor to restart the charge accumulation of the CCD image sensor in the second zone.
On the contrary, when the signal (TINT0) is “L”, the charge accumulation of the CCD image sensor in the second zone is C in the first zone.
It is not completed while the image information is being taken into the AF microcomputer (AFP) of the CD image sensor. Therefore, the AF microcomputer (AFP) sets the signal (TINTC) to "H" again.
Wait for (TINT0) to reverse.

Then, when this signal (TINT0) is inverted, or when the waiting time for inversion of the signal (TINT0) is added to the time required for taking in the image information from the CCD image sensor in the first zone is predetermined. When the maximum charge storage time reached has been reached, a shift pulse (SH0) is generated and the charge storage of the CCD image sensor in the second zone ends. Similarly, the charge accumulation of the CCD image sensor of the third zone is started, the image information of the second zone is taken in, and the charge accumulation state of the CCD image sensor of the third zone is detected in this order. Accumulation and capture of image information are performed.

Here, when the subject has a low luminance and a long charge accumulation time is required, the CCD is driven for a time of (image information acquisition time) × {(total number of zones) -1}. Although the time is shortened, the driving time of the CCD is not shortened when the subject does not have low brightness and a long charge accumulation time is not required.

However, in the circuit configuration shown in FIG. 12B, the shift pulse (SH1) gates (SHG1) to (SHG1).
By adding a buffer section and a shift gate section between the HG4) and the registers (Rg1) to (Rg4), even when the subject has high brightness, the charge is accumulated from the charge accumulation section to the buffer section when the charge accumulation operation is completed. Perform the first shift operation of the charge,
Integration complete signal (TINT0) when the charge accumulation state is detected
If has already occurred, register from the buffer
A second shift operation of charges to (Rg1) to (Rg4) may be configured to shorten the driving time of the CCD.

Also in the circuit configuration shown in FIG. 12A, by adding the buffer section and the shift gate section similar to those described above, the complicated circuit configuration of stopping the transfer clock φ1 during the charge accumulation operation is simplified. In addition, it is possible to reduce the inconvenience such as noise due to a complicated circuit configuration.

Further, although the above-described embodiment is a so-called AF priority type camera in which the shutter release is permitted when the lens reaches the in-focus state, the present invention is not limited to this. A so-called release priority type camera in which the shutter is released according to the shutter release operation regardless of whether it is in focus or not may be used.

Further, the focus detection sensitivity region corresponding to the AF zone and the photometry sensitivity region for exposure control do not necessarily have to be exactly the same. For example, one photometry sensitivity region includes one focus detection sensitivity region. It may cover a wide range,
In a range other than the photometric sensitivity range that looks at the center of the shooting range, one photometric sensitivity range may cover a plurality of focus detection sensitivity ranges.

In the latter case, whichever of the plurality of focus detection sensitivity areas covered by one photometric sensitivity area is selected, the photometric sensitivity area may be selected. Further, in order to monitor the charge accumulation state of the CCD image sensor, the output of the monitor unit provided for each CCD image sensor is used as it is as a photometric signal, and the CCD image sensor corresponding to the selected focus detection sensitivity region is used. The output of the monitor unit provided for monitoring the charge accumulation state may be used as information of the photometric sensitivity range selected corresponding to the focus detection sensitivity range.

Further, in the present embodiment, the focus adjustment of the lens is performed by giving the highest priority to the AF zone in which the shortest subject distance is detected, but the present invention is not limited to this. The focus may be adjusted so that the focus is in the middle between the shortest detected subject distance and the longest detected subject distance, or the focus may be adjusted to the longest detected subject distance. Focus adjustment may be performed. Further, it may be configured so that which zone is preferentially selected can be switched.
Here, whether or not to switch the zones may be determined at the time of camera design so that the most suitable zone for general photography is selected based on statistical data of photographs generally taken.

[0090]

According to the present invention, a focus having a separator lens for separating an image of an object into two images for re-imaging and a light receiving means for receiving the image light of the object imaged by the separator lens. In the detection device, the light receiving means has a plurality of light receiving elements so as to receive light fluxes from a plurality of different target areas, and the separator lens forms an image of the object on the plurality of light receiving elements by a pair of lenses. It is possible to perform focus detection in a plurality of areas without requiring as many separator lenses as there are a plurality of areas, and thus not hindering downsizing and degrading performance.

[Brief description of drawings]

FIG. 1 is an explanatory diagram comparing a focusing area and a photometric area in a conventional example and an embodiment of the present invention.

FIG. 2 is an explanatory diagram of an optical system according to the above embodiment.

FIG. 3 is a block diagram showing an entire circuit of the above embodiment.

4 is a flowchart showing the operation of the control microcomputer in FIG.

5 is a flowchart showing the operation of the AF microcomputer in FIG.

FIG. 6 is a flowchart showing a data preprocessing routine in the above embodiment.

FIG. 7 is a flowchart showing a pre-correlation routine in the above embodiment.

FIG. 8 is a flowchart showing a pre-correlation low contrast determination routine in the above embodiment.

FIG. 9 is a flowchart showing a zone prioritization routine in the above embodiment.

FIG. 10 is a flowchart showing the correlation routine of the above embodiment.

11 is a flowchart showing a modified example of FIG.

FIG. 12 is an explanatory diagram showing a configuration example of a CCD in the above embodiment.

13 is a circuit diagram showing a configuration example of a drive circuit of the CCD shown in FIG.

FIG. 14 is a timing chart showing the operation of FIG.

15 is a timing chart of the integral drive operation of the selected zone in FIG.

16 is a circuit diagram showing another configuration example of the drive circuit of the CCD shown in FIG.

FIG. 17 is a timing chart showing the operation of FIG.

[Explanation of symbols]

 15: Separator lens PAL1, PAL2, PAL3, PAL4: Light receiving element PAR1, PAR2, PAR3, PAR4: Light receiving element 13: Condenser lens

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Toshihiko Karazaki 2-3-3 Azuchi-cho, Chuo-ku, Osaka Inside the Osaka International Building Minolta Co., Ltd. No. 13 Osaka International Building Minolta Co., Ltd.

Claims (2)

[Claims] 1. A focus detection apparatus having a separator lens for separating an image of an object into two images and re-imaging the image, and a light receiving means for receiving image light of the object formed by the separator lens. The light receiving unit has a plurality of light receiving elements so as to receive light fluxes from a plurality of regions having different target regions, and the separator lens forms an image of an object on the plurality of light receiving devices by a pair of lenses. Characteristic focus detection device. 2. The focus detecting device according to claim 1, wherein the focus detecting device has a condenser lens, and a light flux formed by one condenser lens is incident on the plurality of light receiving elements.