JPS6061713A - Focus detecting device - Google Patents

Abstract

PURPOSE:To detect a focus of an optical image whose main body is a low frequency component, and to prevent generation of a pseudo focusing signal in a place where a shift is large, by setting a composite MTF of an MTF characteristic of a filter and an MTF characteristic by a shape of a photodetecting part, to a specified value. CONSTITUTION:A series of photoelectric outputs of image sensors 22A, 22B are sent to corresponding filters 26A, 26B, respectively, and sample holding circuits 27A, 27B sample-hold successively their added outputs I1, I2, I3..., respectively. An operator 28 detects a displacement of an optical image. An MTF characteristic of the first and the second filters 26A, 26B is synthesized with an MTF characteristic X0 determined by a shape of the photodetecting part, a composite MTF characteristic has one peak Ap between frequency zero, namely a DC component and a Nyquist frequency fN=1/2p as shown by a full line Y0, and also an extraction efficiency Ao of the DC component satisfies 0.1Ap<Ao<0.8Ap. According to such a constitution, an influence of various error factors is avoided, no false focusing is generated in a defocused place, a detecting accuracy at a true focused position is raised, and a focus can be detected with respect to an optical iamge whose main body is a low frequency component, as well.

Description

[Detailed description of the invention]

(Technical Field of the Invention) The present invention involves arranging a large number of light receiving sections (projecting a light image onto an array of two light receiving sections L7, processing a series of photoelectric outputs of the light receiving section array, and detecting the state of the light image). Example E relates to an optical image detection device such as a focus detection device for a camera, and in particular relates to a filter cylinder device that suppresses a specific spatial frequency component in an optical image. Photoelectrically detects the relative amount of distortion of a pair of subject images caused by the light flux,
Many camera focus detection devices have been proposed that detect the focus of a photographic lens or measure the distance based on the subject's weight based on the amount of grazing. FIGS. 1 and 2 show the optical system of a typical focus detection device. FIG. 1 shows the optical system described in Japanese Patent Application Laid-Open No. 104859/1985 (-1) The first and second exit pupils of the photographic lens 1
The light fluxes that have passed through the portions 1a and 1b form first and second object bodies, respectively, in the vicinity of the intended imaging plane 2 of the photographing lens l. These first and second object images are respectively re-imaged into first and second images through field lenses 3/4.5 onto first and second photoelectric three-element arrays 6.7. In both the photo-electronic element arrays 6 and 71d, the photo-electronic elements PT having a width p are arranged at a corner p, that is, with substantially no gaps, as shown in FIG. 1(b)K. The pattern of a series of photoelectric outputs a1.82, a3, . , b3...
The patterns / correspond to the illuminance distribution pattern of the second subject image, respectively. The first and second photoelectronic cover patterns
If the object image is relative, it will be detected.
The light beams that have passed through the first and second portions la and lb of the exit pupil of the camera can form first and second object images, respectively, in the vicinity of the intended image plane 2 of the photographic lens 1 via the field lens 3. A small lens array 8 is arranged near this planned image forming plane 2. This small lens array 8 is shown in FIG. 2(b).
As shown in the figure, a large number of small lenses 801, 802, .
It is composed of... Winter Shi/zu 8 () 1.802
Behind...is a pair of photoelectric element PT equipment, PT
2 is placed. The first and second subject images are formed in small parts by the small lenses of the small lens array 8, respectively.

[The divided first subject image is divided into] photoelectric element groups PT, PT
The second subject image is photoelectrically converted by the second photoelectric element group PT2, PT2, respectively. The photoelectric output a of the first photoelectric element group corresponding to the illuminance distribution pattern of the first subject image and the pattern of the second
An image shift is detected from the illuminance distribution pattern b3 of the subject image. The above-mentioned subject image is quantized and photoelectrically converted by one photoelectric element arranged at a pitch p in FIG. Photoelectric conversion is performed by the element. In this specification, the part that quantizes and photoelectrically converts the optical image is called the light receiving part, and the part in which the straws are arranged is called the light receiving part play. Therefore, in FIG. 1, the photoelectric element itself is the light receiving section, and in FIG. 2, the introduction of the small lens and the photoelectric element behind it corresponds to the light receiving section. However, in these focus detection devices, the first. If there is a slight error factor such as a difference in the light transmittance of each optical system related to the second photoelectric element array, a difference in the amplification factor of both plays, a difference in the amount of C offset, etc., it cannot be ignored in focus detection. There is a problem that errors occur. This will be explained in detail. In FIG. 3, the horizontal axis represents the actual relative displacement amount of A1 on the first and second photoelectric element arrays, and the vertical axis represents the detected displacement amount detected by the focus detection device. In an ideal case where the above error factors do not exist, the actual image shift of zero matches the detected amount of shift as shown by the solid line in Figure 3, but if the above error factors exist, the detected amount of image shift is shown. The straight line becomes like a broken line and does not pass through the origin of the coordinates, causing an error in focus detection. Therefore, the present applicant proposed in Japanese Patent Application No. 56-177827 a detection device that eliminates such detection errors. This detection device passes the photoelectric output of the photoelectric element array through a filter to completely remove the zero-frequency C component among the spatial frequency components in the optical image, and then detects image distortion. be. If the C component or its pole (low frequency component in the vicinity (hereinafter referred to as the C component, including the C component and this low frequency component) is completely removed, it is certain that However, for optical images that mainly consist of low frequency components, most of the information necessary for focus detection is removed, resulting in poor focus detection. At the very least, the accuracy would drop slightly (there was a problem with 7 errors. Also, completely removing the C component and switching 5 times would cause a false focusing signal to be generated in areas with large defocus. (Objective of the Invention) Therefore, the object of the present invention is to eliminate the above-mentioned focus detection device, and also to eliminate the focus detection device for optical images mainly consisting of low frequency components. An object of the present invention is to provide a focus detection device that can prevent the generation of false focusing signals in areas with large defocus.
'J' The composite MTF of the F% property and the NITF characteristics depending on the shape of the light receiving part is determined to suppress the C component to some extent but extract a predetermined amount. An embodiment of the present invention will be described with reference to the drawings. Fig. 4 shows an optical system of a focus detection device according to the embodiment,
This optical system is basically the same as the optical system shown in FIG. 1, but the overall configuration is made smaller. In the same figure, a field lens 15 is arranged near the planned focal plane (primary image plane) of the imaging optical system 1 as shown in FIG. The area other than the area 15a is a light-shielding area. The rectangular parallelepiped-shaped transparent block 16 is made of a high refractive index material such as glass or plastic, and this one end surface 16a
The above-mentioned field lens 15 is attached to. Opposed to this one end surface 16a is 1. *A pair of concave mirrors 17 and 18 slightly tilted in opposite directions are provided on the other end surface 16b. The stiff concave mirrors 17 and 18 respectively correspond to the C/Z 4.5 in FIG. Both end surfaces 16a, 16b
A pair of mirrors 19 and 20 are diagonally arranged at an angle of approximately 45 degrees in the block 16 between them, with a predetermined gap therebetween. Photoelectric conversion devices 21 are arranged below the transparent blocks 16, respectively. In this photoelectric conversion device 21, one-dimensional image sensors 22A and 22B are formed below the mirrors 19 and 20, respectively. The light flux that has passed through the imaging optical system I passes through the light transmission area i5a of the field lens 15, enters the second block 16, passes through the gap between the mirrors 19, 20, and passes through the pair of concave mirrors 17.
18. One concave mirror 17 directs the incident light to mirror 1.
The concave mirror 18ij incident on the concave mirror 18ij is reflected toward the concave mirror 9 and toward the mirror 5-20, respectively, and each reflected light reaches the image sensor 22A, 22B via the mirror 19, 20, respectively. In this way, a pair of subject images of substantially the same subject are formed on the seats 22A and 22B. A circuit system for processing the photoelectric output from the photoelectric device R21 will be explained with reference to FIG. 5th FIC, -dimensional image center 22A, 22
B includes light receiving section arrays 23A, 23B consisting of light receiving sections arranged at intervals p, and transfer darts 24A, 24B.
and charge transfer shift registers 25A and 25B. The charge signals of the light receiving sections of the light receiving section arrays 23A and 23B, that is, the photoelectric 1 output a The signals are sent in parallel to 25A, 25B, and are time-series in the order in which the light receiving sections are arranged. The series of photoelectric outputs of the image sensors 22A, 22B means a signal related to the output of the photoelectric element with the photoelectric output barb, which is sent to the corresponding filter means 26A, 26BIC, respectively, and thus linearly amplifies and/or amplifies the photoelectric element output. Naturally, it includes logarithmically amplified data. Both of the filter means 26A and 26BH are constituted by a traverse/sparsal filter shown in FIG. This ton/sparsal filter consists of shunning circuits D1 to Dq for one pixel connected in series and each delay circuit I)1 to
It consists of multipliers W] to W which are connected to the output terminals of D and D via ampules Am, and a multiplier T1 which adds the outputs of these multipliers. Multiplier W1 multiplies the output of delay circuit D1 by weight Wl, and the remaining multipliers W2 to W9 similarly multiply the outputs of delay circuits D2 to D by weights W2 to W9, respectively. Here, W1 to Wq are positive, zero, or the number of angles. Since the filter means 26A and 26B are the same structure as described above, for example, a series of photoelectric outputs a, a2, .
The first photoelectric output a1 is input to the delay circuit, and V
C is sent, filtered, followed by photoelectric, 14. As the force transfer progresses, the addition output ■2, i3,...
Output sequentially. The same applies to the filter means 26B. The output/delhold circuits 27A and 27B receive the addition outputs 1, I2, I3 and 126B of the filter means 26A and 126B, respectively.
Hold Sanzor in sequence. The arithmetic means 28Vt, which works as a position detection means, calculates the outputs of the sample and hold circuits 27A and 27B, calculates the phase difference between the output patterns of both circuits 27A and 27B, and calculates the optical image on the light receiving arrays 23A and 23B. In other words, the displacement of the optical image is detected. Based on the output of the calculation means 28, the focusing drive of the photographic lens or the display of the focus adjustment state is performed. MT of these first and second filter means 26A, 26B
The F characteristic is defined as the MTF characteristic of the entire focus detection device in FIG. 5, as shown by the solid line YO in FIG.
Extraction efficiency Ao of DC component for both C component and Nyquist frequency
is within the range of 10% to 80% of −1 extraction efficiency A, that is, 0, I A < A o < 0.8
It is determined to satisfy A p. More specifically, in the focus detection device shown in FIG.
Elements for filtering the spatial frequency components of the optical images on A and 22B include 4 and 11 second filter means 26A and 2.
In addition to 6B, there are light receiving arrays 23A and 23B. Therefore, the MTF characteristic of the entire focus detection device, the MTF percentage determined by the shape of the light receiving part, and the filter means 26A, 26
It is a composite of B's own MTF% property, that is, the product of both. The MTF characteristic determined from the shape of the light receiving part is as shown by the broken lines X and o in Fig. 7.
This is a characteristic that gradually decreases as the frequency becomes louder when extracted at 00%. Therefore, M of the first and second filter means
'l' The P% property is set so that the composite MTF% property satisfies the position conditions of the peak center mentioned above and the conditions of the C component Ao, as well as the MTF% property X and o determined by the shape of the light receiving part. There is. Of course, MT is determined by the shape of the light receiving part.
Since the F characteristic XO has a DC component extraction efficiency of approximately 100%, the composite MTF characteristic Yo regarding the DC component is the same as the MTF characteristic of the eleventh and second filter means. In this way, the composite MTF characteristic has a peak in the intermediate frequency component between the C component and the Nyquist frequency7, and the C component Ao is within the range of 0IAp to 0.8Ap. Therefore, the photoelectric element arrays 22A, 22
Among the spatial frequency components of the optical image included in the photoelectric output from B, the above-mentioned intermediate frequency components, which are important as focus detection information, are sufficiently extracted, and the D and C components are considerably suppressed and only a part of them is extracted. sample and hold circuit 27A1
27B to the calculation means 28. Therefore, for a normal optical image, as mentioned above, by suppressing the C component, the influence of the various error factors mentioned above can be extremely reduced, and the focus can be detected with high accuracy by the intermediate frequency component, and the optical image mainly consists of low frequency components. , the focus can be detected using the above-extracted part of the C component. Note that the extraction efficiency AO of the D and C components is 0 at the upper limit of the range.
The closer the value is to 8Ap, the smaller the effect of removing the influence of the various error factors described above becomes, and conversely, the closer it is to the lower limit of 01, the more difficult it becomes to detect the focus of an optical image mainly consisting of low frequency components. Therefore, the setting of the value of Ao is determined within the above range, taking into account the magnitude of the various error factors mentioned above and the decrease in detection accuracy of optical images mainly composed of low frequency components, but generally Ao is 0.8A. , the effect of removing the influence of error factors is quite weak, so at this point 0.
7A. It is desirable that it is below, and the t around the O I Ap
Well, the extraction of low frequency components is quite difficult/J't, so in this point it is 0
2A1 or above is desirable0 If you understand this,
The above-mentioned effect can be further exhibited by setting it in the range of 0.2 A≦Ao≦07A. Furthermore, by leaving a certain amount of the DC component, it is possible to improve the focusing accuracy and reduce the probability that a false focus determination will be made at a location where the defocus is large. That is, DC
If the components are completely removed and the output pattern is shuffled, the filtered output pattern will contain only high-order components, so the frequency of cases where the two river patterns match each other again due to a slight image shift will increase, resulting in a false combination. It had the disadvantage that it had a high possibility of making a judgment call. However, such drawbacks can be solved by leaving a certain amount of DC component. Next, the specific M′]”F of the first and second filter means
The characteristics and the value of the weight W that achieves them are illustrated in FIG. The broken line XO in Figure 8 ('A-1) to (E-1) is the MTF characteristic that is determined only from the shape of the light receiving part, and the shape of this light receiving part is as described above. ing. The broken line a and the solid line a2 in FIG. 8(A-1) respectively indicate the composite MTF characteristic of the MTFXo of the light receiving portion shape and the MTF characteristic of the first or second filter means. The dashed line a1, the solid line a2, and the C component extraction efficiency Ao are respectively 0.74A and 0.24A, assuming that the efficiency at the MTF characteristic XO slope position determined from the shape of the light receiving part is 09 in the figure.
p and pゝ, such MTF' characteristics are essentially two weights W]
, W3, M'
The II' F characteristic a1 is as shown in FIG. 8 (A-2).
, = 1, W3 = -0, 2, the ``MTE'' characteristic a2 becomes Wl-1, W3 = -0 shown in Fig. 8 (A-3).
, 64 weights. Figure 8 ('B-1), (B-2), and (B-3) are examples using three weights, and as in Figure 8 (B-2), Vv1 = -
('1.1. When W3=+l and W5==0.1 are selected, the efficiency at the A1 position of the M'l'F characteristic The broken line bl in Figure 8 (B-1) hits.8th
As shown in figure (B-'3')<'W]=-(1,32,"3
= 1, w5 = 0.32, then Ao = 0.2
4 Ap, and its M'TF characteristics are shown in Figure 8 (B-1).
The fruit is #bz. Figure 8 (C-1), (C-2), ('C-3) have 5 weights, Figure 8 ('D-1), (D-2), (D-3)
is an example in which the weight is 7 digits, the sugar 8 figure CE-1), and (E=2) the ld weight is 11 digits. As is clear from this, when the number of weights is increased, the position of the peak of the MTF moves to the lower sampling frequency side. In the above discussion, it is assumed that the sampling pitch is still p after passing through these filter means (-). For fN=-, it is impossible to extract 2p2 effective information for focus detection in the vicinity of fN/.Therefore, as shown in Fig. 8 (E-
In a case like 1), the sampling pitch after passing through one filter is set to 2p, and the Nyquist circumference reduces the number of the/pulls, making calculation processing easier. In addition, considering how to give the weight W, then, Fig. 8 ('B-2)
, (B-3), (D-2), (D73), cE-
As shown in 2), for the reason that the weight value sequence is symmetrical with respect to the central weight, suppose that the DC component is suppressed from the data given at the ring pitch p of S/7°. As a filter that can effectively extract as high-order spatial frequency components as possible, the filter shown in Figure 8 ('
It can be said that B-2), (B-3), (D-2), and (D-3) are preferable. Furthermore, when the weighted addition filter processing of the filter means is performed by a microcomputer or the like, the processing time is shorter when the number of weights is smaller. In particular, the three symmetrical weight values in B-3) are the most desirable. Therefore, this symmetrical 13 weight value strings W1, W3,
W5 motor mentioned above MTF characteristic 0.1 A p < A
o <0,8 A, -...■ and 0.2 A, ≦
The conditions for giving Ao ≦0.7 Ap "・■ are described below. Now, if W3-1 and W1 = W5 = X (number of dogs), this filter means 26 A, 26 Bt7) Mino DC Between the component extraction efficiency Ao and the maximum extraction efficiency A,', the maximum extraction efficiency of the composite MTP% of the entire focus detection device A, ke A, =
γ・A,′ (The value depends on the shape of the γ light receiving part.)
Next, the overall DC component extraction efficiency Ao and the maximum extraction efficiency A,
Substituting this into equation q) and rearranging it, we get 1+2X 1 01 (□-・-< O,S ] -2X γ Subtracting X from this, we get 1 10-γ1 1O-sr ---<X< --- 2i0+r 2.1o+8r The value of γ based on M'IF' determined from the shape of the light receiving part is generally about 1≧γ≧09, so when γ=1
When -0,41 (X<-0,06r = 0.9), it becomes -0,42 (x<-0,08).Similarly, by substituting the above A into equation 0 and subtracting X, we get 1 1O-
2r 1 1.0-7r -1 ≦X≦-1 2 10+21 2 10+7γ When γ=1 -0,33≦X≦-0,09γ=09
When , -0,35≦X≦-0,11. Note that the weight of such a filter means is 1. for every other data of W□, w3ζW5 and sample data. The purpose of this was to bring the MTF peak to 1/2 of the Nyquist frequency fH. Therefore, if spatial frequencies larger than the vicinity of fN are sufficiently removed by some other filter means, W1=W3=-0,
3. Even with continuous weighting such as W2=1, the overall system and 1. It is possible to make the composite MTF'' in a form similar to that shown in Figure 7.Furthermore, if the calculation of such a filter means is performed by a microcomputer, etc., as described below, it is possible to achieve the processing by 2'' operation. becomes easier. So, Koto,
x=-0,25,-0,125. The filter means 26A and 26B may be configured by a microcomputer or the like instead of being configured by hardware as shown in FIG. 6. In this case, the photoelectric output is sampled and held, and the A/D conversion All you have to do is store it in the upper memory and perform the calculation. In the example described in Section 1, each of the first and second filter means is composed of a single filter, but instead, they may be composed of a plurality of filters having mutually different MTF characteristics. . In this case, the MTF of Towara's multiple filters is
The '% property is selected such that the composite MTF characteristic of each filter satisfies the φ condition of the M'I'F characteristics of the first and second filter means. ! All of Takowara's multiple filters may be constructed from hardware, or at least a microcomputer.Also, some filters from Shichiwara may be constructed from hardware, while the remaining filters may be constructed from microcomputers. Good too. In addition to various error factors such as the difference in the transmittance of the optical system and the difference in the amplification factor of the photoelectric element array mentioned above, there is also curariness that occurs in the optical image on the photoelectric element array as error factors in focus detection. This vignetting generally occurs when the photographic lens is replaceable and has a large F value, that is, a dark lens. This influence of Kurare can be avoided by removing the C component to a greater extent than in the case of the various error factors mentioned above. A second embodiment in which both the influence of Kuraray and the various error factors mentioned above are removed will be described with reference to FIGS. 4 and 9. The focus detection device of this embodiment is incorporated into a camera with an exchangeable photographing lens, and in FIG.
1. A signal member (not shown) that is replaceable and corresponds to the F value of the camera is attached to the mounting surface of the camera. I'm here. Field lens]
5, block 16, photoelectric element array 22A, 22BIC
l, built into the camera body. Optical path dividing means for dividing the light transmitted through the photographing lens 1 and guiding it to focus detection optical systems 15 and 16, and to a film or an image pickup device (not shown) are not shown. In FIG. 9, the photoelectric outputs from the pair of photoelectric element arrays 22A122B are in this case FLISb'l, a2, b
2...... shall be output in chronological order alternately. This output is transmitted to the first filter means 42A or the second filter means 4 via the filter selection switch 4o.
2B, and the sample and hold section 43 holds the output at a predetermined pitch. The calculation unit 44 performs calculations for focus detection, and outputs a focus detection signal indicating the amount of defocus, for example. This first filter means 42
The second filter means 42B has different M'l'F characteristics, and the former extracts more DC components than the latter 1-1 For example, as in the first embodiment, the DC component Ao of the composite MTF characteristic of the entire focus detection device is The peak component A is selected to be 20% to 7% of the peak component A, and the latter is, for example, one that largely removes the DC component and makes it almost zero. The filter selection means 411"t connects the switch 4o to one of the first or second filter means 42A, 42B depending on the signal member of the attached photographic lens. FIG. This is an example in which the number of weights is 5 (Wl'''7W5). In this example, the outputs keal, bl, 12, and b2 are output alternately, so the weights W] to W5 are multiplied and added to the output of every other delay circuit D1 to D9 for one pixel. The weighted addition synthesis result In is calculated by the circuit T. That is, the output of D9 is al, the output of D8 is bi, the output of Dγ is ai+1, and so on...the output Kai+ of the river D□. When D appears, the weighted addition synthesis output for the a sequence +1 becomes D
When bi+4 appears in the output of l, weighted addition synthesis formula 11' for the b sequence:=Σ Wm·m=1 b, +m-1 appears. This effect will be explained. When a dark photographic lens with a large F value is attached to the camera, the selection means 41ij and the second filter means 42B, which connect the switch 40 to the second filter means 42B, select the C component and the low frequency component according to the signal system of the lens. Since this method largely removes the vignetting, it becomes somewhat difficult to detect the focus of objects that mainly consist of low frequency components, but it is possible to perform focus detection of ordinary objects with high precision without being affected by the vignetting or the various error factors mentioned above. can. On the other hand, when a bright photographic lens that does not cause flashing is attached, the signal member connects the switch 40 to the first filter means 42A. This first filter means 42A is a second filter means 42B.
Since a considerably large amount of C components are extracted compared to the above, it is possible to detect the focus of an optical image mainly composed of low frequency components with high precision, and also to eliminate the effects of the various error factors mentioned above. In addition to the above-mentioned Kuraray σF value, this phenomenon also occurs depending on the exit pupil position of the photographing lens. Therefore, in addition to the F-number, information on the exit pupil position may be added to the signal member. Also, the switch 40 may be switched manually instead of automatically depending on the signal member as in the above embodiment. It may be L7. FIG. 11 is an example in which the functions shown in FIG. 9 are achieved using a microcomputer. A pair of photoelectric element arrays 22A, 22B al, bl,
The 11th order photoelectric output such as a2, b2, . , are stored in the memory 49. The filter selection means 5 degrees color filter selection signal is read by the microcomputer 48, and the microcomputer 48 thereby selects the image information stored in the memory 49 from 7 to 7 in FIG. 8 (B-1).
1 or solid line b2, and the filtered image information thus obtained is also used to perform calculations for focus detection and output a focus detection signal. Therefore, all filter processing in the case of is performed by software. A third embodiment in which a desired MTF percentage is obtained by synthesizing a plurality of filters will now be described. In FIG. 12, the photoelectric outputs from the photoelectric element array 22A122B are al, b, , a2, b2, . . .
The signals are input to the input terminal 60aK of the filter means 60 in the order of -. As shown in FIG. 13, this filter means 60 is composed of two filters 60A and 60B and a combining means 60C. Each of the filters 60A and 160B has different MF characteristics and has the same configuration as the filter shown in FIG. 10. The synthesizing means 60C multiplies the output (1) of the filter 60A by a coefficient α and the output V2 of the filter 60B by a coefficient β, and then adds them, that is, calculates (1×α + V2×β), and sends the result to the output terminal 60b. Output. Sorting means 4 in Figure 12
First, the values of the coefficients α and β are set according to the information from the signal member of the photographing lens, and also according to manual operation. The defocus amount is calculated and outputted from the output terminal 62a.The function of the filter means 60 will be explained.The weight value of the filter 60A is set as shown in FIG. 14(a), for example. Wl = W2 :w1o=w1]= o, w
3: w9 = 0.16, W4 = We = 0.465
, Wδ; W, = 086, w, = 1, filter 6
For example, the weight value of 0B is as shown in FIG. 14(b).
=w, =018, W2:W, o= 0.32, W3=
W1 = 0.44, W, = W8 = 0.335, W
5 = W = 0.14, W6 = 0. Coefficient α
, β are α;1, β=-06, the combined weight values are as shown in Figure 15 (A-1), and this combined MTF
The characteristics, that is, the MT and F% characteristics of the filter means 60 are as shown in FIG. 15 (A-2').
It has a peak between the D and C components and fN/2. Also, if α; 1 and β = -1, the composite weight value is the first
As shown in FIG. 5 (R-1), the composite M TF characteristic has Ao # and 0.4 Ap as shown in FIG. 15 (R-2). Also, if α; 1 and β knee 1.4, the composite weight value and composite M T'F' are respectively shown in Figure 15 ('C-1) and (C
15 (D-1) and ('D-2). In this way, by appropriately selecting the coefficients α and βθ values, A. It is possible to obtain arbitrary M'T'F characteristics at =0 to Ap building. Therefore, the selection means 4Nd selects α and β according to the signal member of the photographing lens or according to the manual operation to select the M suitable for the photographing lens.
Set to obtain TF characteristics. Furthermore, instead of or in addition to setting the coefficients α and β according to the type of photographic lens as described above, they may be set according to the output of the calculation means 62. For example, when the photographic lens is far away from the in-focus position, the optical image on the photoelectric confectionery array is largely blurred, and the optical image has a relatively large amount of low-order spatial frequency components, causing the opposite effect. When the image is close to the in-focus position, the number of high-order spatial frequency components becomes relatively large. Therefore, as shown in FIG. 12, the determining means 63 determines the defocus signal from the calculating means 62, and when the focus amount is large, that is, when the photographing lens is far away from the in-focus position. , α, and β]) are set to extract a large amount of the C component, and when the amount of defocus is small, that is, when the photographing lens is close to the in-focus position, α and β are set to suppress the DC component. . By doing this, it is possible to avoid the occurrence of false focusing in areas with large defocus, and to increase the detection accuracy near the focus. Figure 16 shows the filters 60A and 60B in Figure 13 as CC
Consists of a D-tra/sparsal filter. Nine filters are shown, and such C0D)/sparsal filters are detailed in IEICE technical research report 88D7B-1. FIG. 17 shows a specific configuration example of the synthesizing means 60C shown in FIG.
, 60B outputs V1, V. The channels 82 to 85 are controlled to be turned on or off by the selection means 41 or the discrimination means 63. 86 to 88, the output terminal of this OP amplifier 88 is the output terminal 60b of this synthesizing means. With such a configuration, K for setting the values of α and β illustrated in FIG. 15 can be controlled by controlling the switches 82 to 85 as shown in the table below. Furthermore, when using the weighting shown in Figure 14, α=
By taking β=IVC, Fig. 15 ('E-”1), (E-
'2) It is possible to have the following characteristics. When using this filter, it is sufficient to sample at twice the sample depth when using filters Vi (A-1) to (D-1), so the area is twice as large with the same number of samples. Since this can be covered, it is particularly effective when detecting cases where the amount of defocus is large or when detecting an object consisting mainly of low frequency components that requires a wide area for determination and requires very little spatial consideration. Of course, it is also possible to configure all or part of the filters 60A, 60B and the synthesizing means 60C in FIG. 13 with a microcomputer. (Effects of the Invention) As is clear from the above description, according to the present invention, the MTF characteristic of the filter means is set to 0. IApguA. <0.8 Ap, and among the spatial frequency components in the optical image, D
Since the C component is suppressed but not completely removed, the influence of various error factors is avoided, and the detection accuracy at the true focus position is improved without causing false focusing at the defocused location. In addition, it is possible to perform focus detection even for optical images mainly composed of low frequency components.

[Brief explanation of drawings]

Figures 1 and 2 are optical diagrams showing the conventional σ focus detection optical system, Figure 3 is a graph diagram showing the relationship between the light image and its detected amount, and Figure 4 is the optical diagram of the conventional σ focus detection optical system. FIG. 5 is a block diagram showing the signal processing system of the above embodiment, FIG. 6 is a block diagram showing an example of the configuration of the filter means, and FIG. M of the filter means according to the invention
A graph diagram showing the T F characteristics, FIG. 8, shows the above M 'l'
9 is a block diagram showing a second embodiment of the present invention. FIG. 10 is a block diagram showing an example of the configuration of filter means. 11 is a block diagram showing a modification of the second embodiment. FIG. 12 is a block diagram showing a third embodiment of the present invention. FIG. 3 is a block diagram showing an example of the configuration of the filter means of the third embodiment. 14 is a diagram showing weighting values, FIG. 15 is a diagram showing weighting values and MTF % characteristics, FIG. 16 is a circuit diagram showing a specific example of a filter configuration, and FIG. 17 is a specific example of a synthesis means. FIG. 2 is a circuit diagram showing an example of the configuration. 1... Imaging optical system 22A, 22B... Vintage light section array 26A, 26B, 42A, 42B, 60... Filter means applicant Takashi Watanabe, representative of Nippon Kogaku Kogyo Co., Ltd. Figure '7' 10th floor "711 Danmine II

Claims (1)

[Scope of Claim] A focus detection device that detects the focus adjustment state of an imaging optical system that creates an optical image of an object, which includes a light receiving unit play in which a large number of light receiving units are arranged, and a photoelectric output of the light receiving unit array. It comprises a filter means for filtering, and a means for creating a signal representing the focus adjustment state based on the output of the filter means, and the MTFI determined by the MTF characteristic of the filter means and the shape of the light receiving section.
The capacity MTF with le property is (1,IAp(Ao<0.8A
p (Aoit frequency extraction efficiency, Ap is the extraction efficiency of the frequency component where the above-mentioned combined MTF is maximum.)
A focus detection device characterized by satisfying the following.