JPH0646257B2 - Focus detection device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a focus detection device for an optical device such as a camera that detects a focus adjustment state of an imaging optical system that forms an optical image of a target object.

(Background of the Invention) In a conventional focus detection device for a single-lens reflex camera, a large number of photoelectric conversion elements are arranged to form two images of almost the same part of a subject by a light flux passing through different regions on the pupil of an imaging lens. The relative displacement of the optical image on the photoelectric conversion element array is detected by calculating the photoelectric output of the photoelectric conversion element array, and focus detection is performed from the detection result. A displacement detection method is known.

As an optical configuration of a focus detection device using such a relative displacement detection method, a type using a microlens array described in Japanese Patent Laid-Open No. 54-159259 and a type using a pair of re-imaging optical systems are known. Has been.

In both cases, two images of almost the same part of the subject are guided onto a pair of photoelectric element arrays to detect the relative displacement of the optical image, and the arrangement of the photoelectric element arrays is different depending on the configuration of the focus detection optical system. In principle, the same output can be obtained as a pair of the photoelectric element array outputs, and they are equivalent as an optical system.

In this case, the pitch for sampling the two images corresponds to the pitch of the microlenses in the type using the microlens array, and in the reimaging lens type, the photoelectric element array is back projected onto the focus detection surface by the reimaging lens. The pitch of the pixel image when this is done corresponds to this.

In order to improve the focus detection accuracy and enable detection of a finer subject image, it is necessary to reduce the sample pitch. However, if the sample pitch is made fine, the number of pixels increases in inverse proportion to this, so that there is a problem that the time required for calculation processing increases and the responsiveness deteriorates.

Further, when comparing images for detecting the anti-phase displacement of the images, it is necessary to use a certain number of pixels in order to satisfy the accuracy, and the detection area determined by (sample pitch x number of pixels) becomes large. However, there is a problem in that focus detection is impossible or the accuracy is poor when a subject having a depth enters the detection area.

(Object of the Invention) The present invention has been made in view of the above problems, and determines the reliability of a focus detection signal obtained by focus detection, and has good focus detection accuracy for any subject. An object is to provide a focus detection device.

(Means for Solving Problems) A first invention of the present invention is a first processing means for performing a first filter process for suppressing a spatial frequency component, and creating the processed data at a first sampling pitch, It has a second processing means for performing a second filtering process for suppressing a spatial frequency component and creating the processed data at a second sampling pitch, and changing the proportional constant of the defocus amount calculation by the sampling pitch. It is a feature.

Further, the second invention of the present invention is that the first filter processing for suppressing the spatial frequency component is performed, and the data is obtained by the spread of the focus detection area (the spread of the data sampling area) on the photoelectric element array used for focus detection. Data is created by the first processing means to be created and the second filter processing that suppresses the spatial frequency component, and by the different spreads of the focus detection areas on the photoelectric element array used for focus detection (spreading of data sampling areas). And a second processing means for performing focus detection based on the data of the first or second processing means.

A third invention of the present invention is characterized in that the first defocus amount obtained by the first processing means and the second defocus amount obtained by the second processing means are combined. is there.

(Examples) Examples of the present invention will be described below.

In FIG. 1, a field lens 15 is arranged in the vicinity of a planned focal plane of an image forming optical system such as a taking lens, and the field lens 15 has a rectangular light transmitting area 15a at the center thereof and the area 15a. The other areas are light-shielded areas. The substantially rectangular parallelepiped transparent block 16 is made of a high refractive index material such as glass or plastic.
The field lens 15 is attached to the. A pair of concave mirrors 17, 18 slightly inclined in opposite directions are provided on the other end surface 16b facing the one end surface 16a. In the block 16 between the both end surfaces 16a and 16b, a pair of mirrors 19 and 20 are obliquely provided at an angle of about 45 ° with a predetermined gap. Below the transparent block 16, photoelectric conversion devices 21 are arranged, respectively. In the photoelectric conversion device 21, corresponding photoelectric conversion element arrays 22 and 23 are formed below the mirrors 19 and 20, respectively.

The light flux that has passed through the imaging optical system 1 passes through the light transmission region 15a of the field lens 15 and enters the block 16, passes through the gap between the mirrors 19 and 20, and has a pair of concave mirrors 17 and 1.
It is incident on 8. One concave mirror 17 reflects the incident light to the mirror 19
On the other hand, the other concave mirror 18 reflects the incident light toward the mirror 20, and the respective reflected lights reach the photoelectric conversion element arrays 22 and 23 via the mirrors 19 and 20, respectively. In this way, a pair of subject images of almost the same subject are arrayed 22,
23 is formed.

A circuit system for processing the photoelectric output from the photoelectric device 21 will be described with reference to FIG.

In FIG. 2, the photoelectric device 21 is a CCD image sensor, and includes at least a first photoelectric conversion element array 22, a second photoelectric conversion element array 23, a transfer gate 24, and transfer units 25 and 26. May include an amplifier for linearly or logarithmically amplifying the photoelectric output. Of course, the optoelectronic device may be a MOS image sensor or other structure. The photoelectric conversion element a ₁ ... a _n constituting the first photoelectric conversion element array 22 are arranged at a pitch Po in a row in a state where each other in close proximity, are identical also the configuration of the second photoelectric conversion element array 23 . This photoelectric device 21 has photoelectric outputs a ₁ ... A _n of the first array 22.
, And the photoelectric output b ₁ ... B _{n of} the second array 23 as a ₁ ,
b ₁ , a ₂ , b ₂ ... A _n , b _n alternately output each other, and a series of photoelectric outputs a ₁ , b ₁ ... A _n , b _n
Is repeatedly output at a predetermined time interval. Such photoelectric conversion element arrays 22 and 23 have MTF characteristics as shown in FIG. The input terminals of the first and second filter means 27 and 28 are connected to the output terminals of the photoelectric device 21. The first filter means 27 allows low-order spatial frequency components to pass as shown in FIG. 3 (b), but has an MTF characteristic of sufficiently suppressing high-order spatial frequency components at frequencies around 1/8 Po and above. The two-filter means 28 passes low-order spatial frequency components as shown in FIG. 3 (c), but has an MTF characteristic of sufficiently suppressing high-order spatial frequency components at frequencies around 1/4 Po or higher. As described above, the second filter means 28 is defined so as to pass the spatial frequency component on the higher order side as compared with the first filter means 27. The selection means 29 are the first and second
The outputs of the filter means 27 and 28 are selectively selected and sent to the sample hold means 30. The sample-holding means 30 is composed of a pre-stage sample-holding means 30A and a post-stage sample-holding means 30B which are connected in series with each other. The A / D converter 31 is a sample hold circuit 3
The output of 0B is A / D converted. The memory means 32 stores the result.

An example of the calculation contents of the calculation means 33 will be briefly described below.

The filtered output for one photoelectric conversion element array is output as a plurality of areas A (1), A (2), A (3) ... A (i).
Separated by (L), the filtered output for the other photoelectric conversion element array also corresponds to B (1), B (2), B
(3) ... B (j) ... B (L) are divided, the matching is calculated for a plurality of pairs of A (i) and B (j) in these areas, and the combination having the best matching (i, The defocus amount Zi is calculated by calculating the relative displacement amount of the optical image on both arrays using the value of j) and the combination in the vicinity thereof. For example, when the number of constituent elements in each area is equal (M + 1), A (i) = {A _i , A _{i + 1} , A _{i + 2} , ... A _{i + M} } B (j) = { B _j , B _{j + 1} , B _{j + 2} , ... B _{j + M} }. As the correlation amount representing the degree of consistency, l = i-j is used when the image is displaced by l data positions,
[X] represents the maximum integer not exceeding x, Given by.

This correlation amount C (l) is represented by l =-(L-1), ...- 1,0,
1 ... (L-1) for each shift amount, and the shift amount l = l at which the position of maximum correlation, that is, C (l) has the smallest value
₀ is obtained. When l ₀ is not equal to the value at both ends (L-1 or -L + 1), the fraction Δl ₀ of a finer shift amount is calculated by the following formula, for example. Can be extrapolated by.

The amount of deviation l _{0 +} △ l ₀ containing the fraction obtained in this manner
From this, a focus detection signal Zi representing the defocus amount is obtained.

In the embodiment, as will be described later, the sampling pitch is different when the first filter means is used and when the second filter means is used, and the proportional constant when calculating the defocus amount from the shift amount l ₀ + Δl ₀ is different. . Further, the number L of the divided plural areas is not necessarily the same in the case of selecting the first filter means and the case of selecting the second filter means, and therefore the calculation content of the calculation means 33 will be somewhat different depending on the selected filter means. . This is identified by inputting the filter selection signal 34a from the discriminating means 34 to the arithmetic means 33 in FIG. 2 as well, and a partially different arithmetic operation is performed.

It also represents the amount of information being output by the first and second filter means.
As the information amount signal Di, for example, the difference between the maximum output data and the minimum output data of the filter means can be used, but the maximum correlation data C (l) C (l) _MAX.
And the smallest one, C (l) _MIN , may be taken. Further, the image shift amount is the image shift determination region,-(L-1) ≤l≤ (L-1)
Considering the case where the value does not fall within the range of (1), the value of C (l) is also confusing in this case because it is the minimum at a certain value of l within the above range. However, in such a case, C
Since (l) _MIN / (C (l) _MAX- C (l) _MIN ) does not become as small as when the image shift amount is within the range of the judgment area, it can be excluded by providing an appropriate threshold value C _TH. . Immediately C (l) _MIN / (C (l) _MAX- C (l) _MIN )> C
_At the time of _TH, the information amount signal Di which normally takes a positive value is excluded from the correlation by giving zero or a negative value. The discriminating means 34 receives the focus detection signal Zi and the information amount signal Di as described above, and the information amount signals Di (1) and Di (2) at the time of selecting the first and second filter means are respectively predetermined values Do.
If smaller than (1) and Do (2), focus detection signal Zi
, The selection means 29 is now the first filter means 27.
Is selected, the second filter selection signal for selecting the second filter means 28, specifically, the H level output, and conversely, if the second filter means 28 is currently selected, the first filter means is selected. 27 outputs a first filter selection signal, specifically an L level output, to each output terminal 34a.
On the other hand, on the other hand, the information amount signal Di (1) or Di (2)
Is greater than or equal to the corresponding predetermined amounts Do (1), Do (2), the absolute value of the focus detection signal Zi (1) or Zi (2) is greater than the corresponding predetermined value Zo (1), Zo (2). At this time, the first filter selection signal is set to a predetermined value Zo (1), Zo
(2) In the following cases, the second filter selection signal is generated at the output terminal 34a. Further, the discriminating means 34 determines that the information amount signals Di (1) and Di (2) are the predetermined amounts Do (1) and D (D).
When it is o (2) or more, the memory update signal is output to the output terminal 34b.
Occurs in. In response to this storage update signal, the memory circuit 35
Stores the focus detection signal Zi at that time. In accordance with the focus detection signal Zi stored in the memory circuit 35, the display device 36 displays the focus adjustment state, and the drive device 37 drives the imaging optical system 1 toward the in-focus position. The sample pulse generating circuit 38 is connected to the output terminal 34a of the discriminating means 34 and supplies the sample and hold circuits 30A and 30B with a sample pulse for starting the sample and hold. The period of this sample pulse changes according to the output of the discriminating means 34, and the period when it is the first filter means selecting signal is larger than that when it is the second filter means selecting signal. It has been selected twice. The sample pulse generating circuit 38 has an output terminal 39 of the first counter 39.
When a start signal, specifically an H level signal, is received from a, the generation of the sample pulse is started, and when an end signal from the output terminal 40a of the second counter 40, specifically an H level signal, is received, Stop sample pulse generation. The first counter 39 is a presettable counter, which presets a preset value sent from the setting section 41 through the gate means 42, and the AND gate 43.
The pulse output from is down-counted, and when the content becomes zero, the H-level start signal is output. The second counter 40 is also a presettable counter and is the gate means 4
4 is preset to a preset value from the setting section 41, and the sample pulse to the subsequent sample hold circuit 30B is down-counted, and when the content becomes zero, an H-level end signal is generated. The setting unit 41 is
First preset value for first counter and second preset value for second counter used when selecting first filter means 27 and second preset value for first counter and second counter used for selecting second filter means The second preset value is stored in advance, and the output terminals 41a and 41c respectively output the first preset value for the first counter and the first preset value for the second counter when the first filter means is selected. A second preset value for the first counter and a second preset value for the second counter when the second filter means is selected are output to the respective. In this example, the first preset value of the output terminal 41a is smaller than the second preset value of the output terminal 41b, and the first preset value of the output terminal 41c is set equal to the second preset value of the output terminal 41d.

An H level signal is input to the input terminal 45 from a sequence control unit (not shown) in synchronization with the start of transfer of a series of photoelectric outputs a ₁ , b ₁ ... _An , b _n from the photoelectric device 21. This signal indicates the next preset value after the sample and hold of all data is finished.
It is reset to L level at an appropriate time until it is set to 9, 40. A clock synchronized with the transfer clock for transferring the series of photoelectric outputs is input to the input terminal 46.

This action will be described below.

It is assumed that the discrimination means 34 generates the L level output which is the first filter means selection signal at the output terminal 34a. Based on this selection signal, the selection means 29 selects the first filter means 27, and the gate means 42 and 44 select the first counter preset value and the second counter preset value from the output terminals 41a and 41c of the setting section 41. It inputs to the 1st counter 39 and the 2nd counter 40, respectively, and presets each counter to the preset value. The photoelectric output _a 1 from the photoelectric device 21 a series of a signal from the sequence controller _{Thereafter, b 1, a 2, b} 2 ... a n, b n are read. This series of photoelectric outputs a ₁ , b ₁ ... A _n , b _n
Shows a photoelectric output _{a 1, a} 2 ... _{a n} from the first photoelectric conversion element array 22 in FIG. 4 (a) of. The series of photoelectric outputs a ₁ , b ₁ ... _An , b _n are the first filter means 27.
Is subjected to filtering processing by and is converted into the filtered outputs A ₁ , B ₁ ... A _n , B _{n shown} in FIG. Of these filtered outputs A ₁ , B ₁ ... A _n , B _n , those associated with the first photoelectric conversion element array A ₁ , ... A _n are shown in FIG. 4 (b). Comparing FIG. 4 (b) with FIG. 4 (a), the effect of suppressing the higher spatial frequency components by the first filter means 27 is clear. On the other hand, since the H level signal is input to the input terminal 45 in synchronization with the reading from the photoelectric device 21, the AND gate 43 outputs the transfer clock from the input terminal 46. The first counter 39 subtracts the number of transfer clocks from the preset value, and when the number of input clocks becomes equal to the preset value, generates a start signal H level output. The start signal is input to the sample pulse generation circuit 38, inverted and input to the AND gate 43 to close the gate.
The sample pulse generation circuit 38 supplies the front-stage and rear-stage sample pulses SP1 and SP2 shown in FIGS. 5B and 5C to the front-stage and rear-stage sample hold circuits 30A and 30B, respectively, in response to the start signal. The pre-stage sample hold circuit 30A samples the filtered outputs A ₁ , B ₁ ... A _n , B _n to A ₄ , B ₄ , A ₈ , B ₈ , A ₁₂ , B ₁₂ ... In response to the pre-stage sample pulse SP1. To do. The pre-stage sample and hold circuit 30A outputs outputs A ₄ , A ₈ ... Related to the first photoelectric conversion element array for a short time and is related to the second photoelectric conversion element array as shown by the range of arrow in FIG. 5 (b). The outputs B ₄ , B ₈ ... Are held for a relatively long time. In order to equalize the retention time of both,
The rear sample-hold circuit 30B samples and holds the output of the front sample-hold circuit 30A in response to the rear sample pulse SP2. The second counter 40 counts the rear-stage sample pulse SP2, generates an end signal when it becomes equal to the first preset value, and stops generation of the front-stage and rear-stage sample pulses SP1 and SP2.

In FIG. 4B, the sampled filtered outputs A ₄ , A ₈ , A ₁₂ ... Of the filtered outputs A ₁ ... A _n relating to the first photoelectric conversion element array have a mark M under the output. _s is attached. As you can see from this figure,
This sampled filtered output distribution range (hereinafter referred to as a sampling region) ₁₂ is the filtered output A.
It can be seen that it occupies most of the range ₁ ... A _n .

The A / D converter 31 A / D converts the output of the post-stage sample and hold circuit 30B and sends it to the memory means 32. The reason for providing the latter-stage sample hold circuit 30B is as follows. If you if front sample and hold circuit outputs a direct A / D conversion of 30A, the pre-stage sample-and-hold circuit 30A, the output B _4, B 8 _... in comparison with the retention time of the output A
₄ , A ₈ ... Has a short holding time, so an expensive high-speed A / D converter is used as the A / D converter 31 so that the A / D conversion operation is completed within the shorter holding time. There must be. Even if a high-speed A / D converter is used, the high-speed characteristic cannot be utilized in the A / D conversion of the outputs B ₄ , B ₈ ... Having a long holding time. However, the use of the post-stage sample and hold circuit 30B solves the above problem. As shown in FIGS. 5 (d) and 5 (e), the sampling period when the second filter means is selected is smaller than that when the first filter means is selected. In other words, the holding time of the latter-stage sample hold circuit 30B is Since it is shorter when the second filter means is selected, the conversion required time of the A / D converter 31 is determined by the holding time when the second filter means is selected. Then, of course, when the first filter means is selected, the holding time of the post-stage sample and hold circuit becomes unnecessarily longer than the conversion required time. In order to avoid this waste, the frequency of the transfer clock of the photoelectric output when the first filter means is selected is made higher than that when the second filter means is selected, and the latter sample-hold circuit 30B when both are selected.
The holding times of the can be made equal.

The computing means 33 computes the filtered output stored in the memory means 32 to obtain the focus detection signal Zi (1) and the information amount output D.
i (1) is output. The determination means 34 uses the signal Zi
(1), Di (1) corresponding predetermined values Zo (1), Do
Compare with (1).

(A) When Di (1) is Do (1) or more In this case, the determination unit 34 outputs the memory update signal to the output terminal 34.
The focus detection signal Zi (1) at this time is stored in the memory circuit 35. Display device 36 and drive device 37
On the basis of the stored signal Zi (1), displays the focus adjustment state and drives the imaging optical system 1 to the in-focus position. When the signal Zi (1) is larger than the predetermined value Zo (1), the discrimination means 34 continues to output the first filter means selection signal to the output terminal 34a. Therefore, at this time, the photoelectric device 21 further series of photoelectric outputs _{a 1, b 1 ... a n} , b n
When this is output, all the circuits perform the same operation as described above.

When the signal Zi (1) is less than or equal to the predetermined value Zo (1), the discrimination means 34 outputs the H level output which is the second filter means selection signal to the terminal 34a. In response to the second filter means selection signal, the selecting means 29 causes the second filter means 2
8 is selected, and the gate means 42 and 44 send the second preset values for the first and second counters from the output terminals 41b and 41d of the setting section 41 to the first and second counters 39 and 40, respectively. After that, the series of photoelectric outputs a ₁ , b ₁ ... A _n , b _n read from the photoelectric device 21 are filtered by the second filter means 28 and converted into A ₁ , B ₁ ... A _n , B _n . A first filter already output _A 1 relates to a photoelectric conversion element _{array, A} 5 ... _{A n} at this time is shown in FIG. 4 (c). Comparing this FIG. 4 (c) with FIG. 4 (b),
It can be seen that the figure in FIG. 7C is not smooth, and that the second filter means 28 allows higher-order spatial frequency components to pass than the first filter means 27. On the other hand, the first counter 39 counts the output transfer pulses of the AND gate 43 in synchronization with the reading of the photoelectric output, and generates a start signal when the counted value matches the second preset value. Since the second preset value for the first counter when the second filter means is selected is set to be larger than the first preset value for the first counter when the first filter means is selected, the start signal generation time at this time is , Is later than the start signal generation time when the first filter means is selected. With this start signal, the sample pulse generating circuit 38 is shown in FIGS.
The sample pulses SP3 and SP4 for the front stage and the rear stage shown in are generated. The cycle of the sample pulses SP3 and SP4 is in accordance with the second filter means selection signal sent from the discriminating means 34, and the sample pulse S when the first filter means is selected.
It is shorter than P1 and SP2, and is set to 1/2 in this embodiment. Therefore, as shown in FIGS. 5 (d) and 5 (e), the above-mentioned and latter-stage sample hold circuits 30A and 30B are
The filtered outputs A ₁ , B ₁ ... A _n , B _n are sampled at a cycle ½ that of the time when the first filter means is selected, and A ₈ ,
_{B 8, A 10, B 10} , A 12, B 12 ... to output. The second counter 40 counts the post-stage sample pulse SP4, and when the count value matches the second preset value, generates an end signal to stop the generation of the sample pulses SP3 and SP4. Since the second preset value for the second counter is set equal to the first preset value for the second counter when the first filter means is selected, the filtered outputs A ₈ and B sampled when the second filter means is selected. ₈ ,
The number of A ₁₀ , B ₁₀ ... Is equal to that when the first filter means is selected.

The first of the filtered outputs sampled in this way
The photoelectric conversion element array is shown by a mark M _s in FIG. 4 (c). In the present embodiment, the sampling period and the number of samples when the first filter means are selected are half and equal to those when the second filter means is selected, so that the sampling region l ₁ shown in FIG. It is twice as large as l ₂ in FIG. Of course, the sample numbers of both do not necessarily have to be equal. In addition, FIG. 5 (d), (e)
The graph in () is drawn earlier than the sampling start time for drawing purposes.

The sampled output is sent to the calculating means 33 via the A / D converter 31 and the memory means 32 and is calculated.
Since the filtered output at this time contains more high-order spatial frequency components than when the first filter means is selected, this second output
The focus detection signal Zi (2) at the time of selecting the filter means has higher accuracy in the vicinity of the in-focus position. Di (2) ≧
When Do (2) and | Zi (2) | ≦ Zo (2), the determination means 34 outputs the second filter means selection signal to the output terminal 34.
The memory update signal is output to the output terminal 34b.
The focus detection signal Zi (2) at this time is stored in the memory circuit 35. Display and imaging optical system drive are performed according to the stored contents. Di (2) ≧ Do (2),
When | Zi (2) |> | Zo (2) |, the discrimination circuit 34 outputs the first filter means selection signal.

(B) Di (1) or Di (2) is Do (1) or D
If smaller than o (2).

In this case, irrespective of the focus detection signal Zi, the discrimination means 34 selects the second filter means selection signal if the first filter means is selected, and conversely selects the first filter means if the second filter means is selected. The signals are output to the output terminals 34a, respectively. This switches the filter means for selecting the selection means 29. Further, since the focus detection signal Zi when the signal Di is smaller than the predetermined value Do is extremely low in accuracy, the discriminating means 34 does not generate the storage update signal. Therefore, the signal Zi at this time is not used for driving the display and imaging optical system. The switching of the filter means irrelevant to the signal Zi is performed for the following reason. That is, for example, when the subject contains almost no low-order spatial frequency components but a large amount of high-order spatial frequency components, necessary information can be obtained by selecting the second filter means 28.

In this embodiment, as shown in FIG. 4B, when the first filter means is selected, that is, when the defocus amount is large and the relative displacement amount of the subject images on the first and second photoelectric element arrays is large. When the sampling area l ₁ is wide and the second filter means is selected, that is, when the shift amount is small, FIG.
The sampling area l ₂ is narrowed as shown in FIG. This is very effective for focus detection. That is, when the sampling area is widened, even if the subject image relatively shifts, the shift can be detected. Therefore, the defocus amount can be detected even if the taking lens is far away from the in-focus position. On the other hand, widening the sampling area increases the possibility that an object having a different distance or an object having a depth will come in. Focus detection when the defocus amount is large is sufficient if it is possible to determine whether it is the front focus or the rear focus or to roughly determine the defocus amount, and accurate measurement of the absolute value of the defocus amount is not always necessary. Even if a subject or the like exists in the sampling area, the influence is small. However, when the defocus amount is small and the absolute value of the defocus amount must be accurately measured, the presence of the deep subject tends to cause a large error in the measurement. Therefore, when the second filter means that requires highly accurate focus detection is selected, high-order spatial frequency component information is used to improve detection accuracy and to narrow the sampling area to reduce the possibility that a deep subject will enter there. There is.

Generally, it is not necessary to increase the sampling period just because the sampling area is widened. For example, the area of l ₁ related to the output of FIG. 4B is set to be smaller than the sampling pitch 4Po shown in the drawing, Po or 2Po.
Of course, sampling may be performed at the pitch. However, it is not preferable to reduce the sampling pitch to Po or 2Po because the number of samplings becomes 4 times and 2 times respectively and the storage capacity of the memory means 32 and the operation scale of the operation means 33 increase remarkably. Therefore, even when the sampling area is changed as in this embodiment, it is extremely effective to set the number of samples to the same level.

As described above, when selecting the first filter means that passes only the low-order spatial frequency component, the sampling period is increased to 4Po, and when selecting the second filter means that passes the high-order spatial frequency, the sampling period is reduced to 2Po. It is extremely advantageous in terms of utilization. Here, the relationship between the sampling pitch and the MTF characteristic of the filter will be described in detail.
Since the sampling period was 4 Po when the first filter was selected, the Nyquist frequency at this time is 1/8 Po. From the sampling theorem, this spatial frequency component of 1/8 Po or higher frequency may cause a malfunction, so it is desirable to remove it. As shown in FIG. 3 (b), the first
The MTF of the filter means is the Nyquist frequency 1 / 8P
The components above about 0 are sufficiently suppressed and the components below are allowed to pass, so that the passed components can be effectively used. However, if the sampling period is Po when the first filter means is selected, the Nyquist frequency at this time is 1 / 2P.
Therefore, spatial frequency components below this frequency can be used for focus detection. However, as shown in FIG. 3 (b), the components having a frequency of ⅛ Po or higher are removed by the first filter means, so that the sampling period Po is increased four times as much as the sampling period 4Po. However, the amount of spatial frequency components that can be used is the same, and the increase in the number of samplings is completely useless. As is clear from the above, from the viewpoint of effective use of information, it is desirable to determine the sampling period so that the Nyquist frequency determined by the period is near the end of the MTF frequency band of the filter means.
Further, if the preset value of the setting means 41 can be changed from the outside, the spread of the sampling area, that is, the subject area used for focus detection can be arbitrarily changed, and a deep subject can be included in the area. It can be prevented.

Next, a specific example of the configuration of the block in FIG. 2 will be described.

FIG. 6 shows an example of the filter means 27, 28. Delay circuits D ₁ , D ₂ ... D _m for one pixel are connected in series, and the delay circuits D ₁ , D ₂ , D ₃ ... D _m are connected. Amplifier A respectively
_m is connected to the multipliers W ₁ ... W _s . These multipliers W ₁ ... W _s multiply their inputs by weights W ₁ ... W _s , respectively. This weight is a positive or negative number. The adder circuit T ₂ adds the outputs of the multipliers. The optoelectronic device 21 is connected to the delay circuit D _1.
When a series of photoelectric outputs from the
₂ outputs the filtered output. Predetermined MT
There are various possible ways of taking the weights W ₁ ... W _s that give the F characteristic, but they are not uniquely determined, but some specific examples are shown below.

In order to obtain the filter means having the MTF characteristic as shown in FIG. 3 (c), D _m = D ₉ and W _s = W ₅ are set, and W ₁ ... W ₅ tend to have mutual magnitudes as shown in FIG. 7 (a). It is decided to be as shown in. As a specific example, W ₁ = 0.28 W ₂ = 0.76
_{W 3 = 1 W 4 = 0.76} , a _{W 5} = 0.28. _{D m} is the likewise obtain filter means of the MTF characteristic of FIG. 3 (b)
= D ₁₇ , W _s = W ₉ and W ₁ ... W ₉ are determined as shown in FIG. 7 (b). As a specific example, W ₁ = 0.28 W ₂ = 0.52
_{W 3 = 0.76 W 4 = 0.94} W 5 = 1 W 6 = 0.94 W
_{7 = 0.76 W 8 = 0.52 W} 9 = 0.28. The characteristics of FIG. 3 (d) have the weights of FIG. 7 (c) or (d), and the characteristics of the dotted line (e ₁ ) and the solid line (e ₂ ) of FIG. The weights of (e) and (f) may be used, and the weights of FIG. 7 (g) may be used for the characteristics of FIG. 3 (f).

A combination of the first filter means and the second filter means shown in FIG. 8 can be obtained by appropriately combining the MTF characteristics of FIGS. 7 (a) to 7 (g).

If a CCD transversal filter is used as the filter means, the filter means can be easily constructed.

FIG. 9 shows a specific configuration example of the discriminating means 34 shown in FIG.

In FIG. 9A, the first memory 340 and the second memory 3
41 are predetermined values Do (1), Do (2), and Zo, respectively.
(1) and Zo (2) are sent to the comparators 344 and 345 via the gate means 342 and 343. The comparator 344 compares one of the outputs Do (1) and Do (2) of the memory 340 selected by the gate means 342 with the information amount signal Di from the arithmetic means 33. Similarly, the comparator 345 outputs the outputs Zo (1) and Zo of the memory 341.
(2) One is compared with the focus detection signal Zi. The gate means 346 inputs the output α of the comparator 344, the output β of the comparator 345 and the output γ of the discrimination means 34. The concrete structure of the gate means 346 is shown in FIG. The D-type flip-flop 347 is the above α,
A clock pulse generated at a timing after the output of β is determined is received by 348 and the output δ of the gate means 346 is input and stored. The updated output of the flip-flop 347 is used as the output of the discriminating means 34. The following table shows an example of the operation of the discriminating means 34.

However, when Di (1) <Do (1) or Di (2) <Do (2), α = L Di (1) ≧ Do (1) or Di (2) ≧ Do (2), α = H | Zi ( 1) | <Zo (1) or | Zi (2) | <Zo (2) and β = L. In other cases, β = H.

In the description of the embodiment, the main theme is to switch a plurality of filter means depending on the information amount Di and the defocus amount Zi, and depending on the information amount Di and the defocus amount Zi. It has been described that the spread of the sampling area is switched and the sampling pitch is switched correspondingly. In practice, it is most preferable to have both of them as in the embodiment, but the filters are switched depending on the information amount Di and the defocus amount Zi.
Switching the sampling region and the sampling pitch depending on the information amount Di and the defocus amount Zi is a matter of some problems related to the Nyquist frequency, and even if only the latter is used, effective focus detection is achieved. It is possible to provide a device. For example, as the filter means, only one having the MTF characteristic of FIG. 3 (c) is used,
In the vicinity of the focus, the sampling cycle 2 as shown in FIG.
The data is sampled over the region of l ₂ at Po, and the filter means is left as it is at the place where the defocus is large, but the data is sampled over the region corresponding to l _{1 of} FIG. 4B at the sampling period 4Po. Performs front-back pin determination with. In this case, compared with the case where the MTF characteristic of the filter means is switched to the one shown in FIG. 3 (b) in a place where the defocus is large, a component higher than the Nyquist frequency is slightly extracted, so that some malfunction may occur or a higher order may occur. The possibility of false focusing due to the presence of the spatial frequency component of is increased, but except for some of these possibilities, it is expected that there will be a large blur of the optical image in the place of large defocus, and a certain effect is expected. Can be done. That is, the effect of changing the sampling area and the sampling pitch depending on the defocus amount is expected as it is. Of course, as a single filter, the M
The TF characteristic is not limited to that shown in FIG. 3 (c), and may be one having a dotted line (e ₂ ) in FIG. 3 (e) or other characteristics.

In the above embodiment, the focus detection signal based on the filtered output of the first filter means and the focus detection signal based on the filtered output of the second filtering means are selectively selected according to the information amount Di and the defocus amount. It was something to do.

Next, another embodiment of the present invention will be described in which, instead of the alternative selection, respective focus detection signals are simultaneously used in a predetermined relationship.

In FIG. 10, a series of photoelectric outputs a ₁ , b ₁ ... A _n , b _n from the photoelectric device 50 are sent to the first filter unit 52 and the second filter unit 53 directly via the delay unit 51. To be The photoelectric device 50 has the same configuration as that of the first or second embodiment, and the first and second filter means 5 are provided.
2 and 53 are the same as those in the above-described embodiment, and the center of the MTF frequency band of the first filter means 52 is the second filter means 5.
It is shifted to the lower spatial frequency side than that of 3. Regarding the delay time of the delay means 51, the filtered output of the first filter means 52 is sent to the sample hold means 54 after all the filtered outputs of the second filter means 53 have been sent to the sample hold means 54 for the series of photoelectric outputs. 54
Is set to be sent to. Of course, this delay means 51 can also be provided on the second filter means side. The sample / hold means 54, the A / D converter 55, the memory means 56, and the arithmetic means 57 following the sample / hold means 54 have the same configurations as those of the above-described embodiment. This computing means computes the filtered output of the second filter means 53 which was sent first, calculates the signals Di (2), Zi (2) and then the signal Di (of the filtered output of the first filter means 52. 1) and Zi (1) are calculated. The memory circuit 58 receives the signals Di (1), Di (2), Zi from the calculation means 57.
All of (1) and Zi (2) are stored. Synthesizing means 59
Receives the above-mentioned signal from the memory circuit 58 and calculates an output Z by combining the focus detection signals Zi (1) and Zi (2) in the following predetermined relationship. That is, Z = (1-α) Zi
(1) + αZi (2) where the weight α is a number between 0 and 1 inclusive, and the signals Zi (1), Zi (2), Di (1), Di
It is determined according to the size of (2). Specifically, α is determined when the signal Zi (1) or Zi (2) is small, that is, when the imaging optical system is located near the in-focus position, the band is the first on the higher spatial frequency side. The signal Zi (2) based on the output of the two-filter means 53 has α set to 1 or a value close to it so as to be emphasized. Conversely, when the signals Zi (1) and Zi (2) are sufficiently large, the band has a low-order space. Α is set to zero or a value close to it so that the signal Zi (1) based on the output of the first filter means 52 on the frequency side is emphasized. Further, when the signal Di (2) is very small near the focus, the signal Zi
Since (2) is reduced in accuracy, the signal Di
If (1) is large, the weight of the signal Zi (1) is increased, and conversely, even if the in-focus position is far, the signal Di
If (1) is very small and the signal Di (2) is large, the weight of the signal Zi (2) is increased.

The memory means 60 stores the combined output Z when at least one of the signals Di (1) and Di (2) exceeds the corresponding predetermined values Do (1) and Do (2). In accordance with the output of the memory circuit 60, display and image forming optical system driving are performed as in the above-described embodiment. Sample pulse generation circuit 61
Is similar to that of FIG.

(Effect of the Invention) As described above, according to the first invention of the present application, the sampling pitch of the photoelectric output used for the focus detection calculation is made variable according to the spatial frequency component of the subject, and the first processing means or the second processing means is used. Since any one of the focus detection calculation results of the processing means is used, the processing can be performed in the optimum focus detection calculation time according to the spatial frequency component of the subject, and the increase of the calculation time can be suppressed. Further, since the focus detection calculation is performed using proportional constants that differ depending on the sampling pitch, highly accurate focus detection can be performed.

Further, according to the second aspect of the invention, since the width of the focus detection area is variable, it is possible to perform focus detection with high accuracy even for a subject having a depth or a subject having few high frequency components. For example, focus detection accuracy can be improved by using a narrow focus detection area for a subject with depth, and focus detection accuracy can be improved by using a wide focus detection area for a subject with few high frequency components. .

According to the third invention, the first processing means obtains the first information amount and the first defocus amount, and the second processing means obtains the second information amount and the second defocus amount. Since the defocus amount that combines the first and second defocus amounts is calculated depending on the two information amounts or the first and second defocus amounts, the focus detection accuracy is improved regardless of the subject. Can be improved.

[Brief description of drawings]

FIG. 1 is an optical diagram showing the optical system of the first embodiment of the present invention, FIG. 2 is a block diagram showing the circuit system of the first embodiment, and FIG. 3 (a) shows the MTF characteristics of the photoelectric conversion element array. Graphs (b) to (f) of FIG. 3 are graphs of MTF characteristics of the filter means, and (a), (b) and (c) of FIG. 4 are photoelectric output, output of the first filter means and output of the second filter means, respectively. FIG. 5 (a) to FIG. 5 (e) are timing charts showing the output of the filter means and sample pulses, FIG. 6 is a block diagram showing a concrete configuration example of the filter means, and FIG. 7 (a). ) To (g) are diagrams showing the weight of the filter means, and FIGS. 8 (a), (b) and (c) are MTs of the first and second filter means.
FIG. 9 (a) and FIG. 9 (b) are block diagrams showing a concrete configuration example of the discriminating means, and FIG. 10 is a block diagram showing a circuit system of another embodiment of the present invention. (Explanation of symbols of main parts) 4; 21; 50 ... Photoelectric device, 5; 27; 52 ... First
Filter means, 6; 28; 53 ... second filter means,
7; 29 ... selecting means, 10; 33; 57 ... computing means

Claims (8)

[Claims] 1. A pair of photoelectric conversion element arrays in which a large number of photoelectric conversion elements are arranged, a focus detection optical system for projecting two images of substantially the same portion of an object on the pair of photoelectric conversion element arrays, respectively. 2 based on the photoelectric output of the pair of photoelectric conversion element arrays
In a focus detection device that calculates a relative displacement of an image and detects the focus adjustment state of the imaging optical system based on the calculation result, a predetermined spatial frequency for the photoelectric output of the pair of photoelectric conversion element arrays. A first processing means for applying a first filter process for suppressing a component and creating the processed data at a first sample pitch, and for the photoelectric output of the pair of photoelectric conversion element arrays, the first A second processing unit that performs a second filtering process different from the filtering process, and creates data of the processed data at a second sampling pitch different from the first sampling pitch, and the first processing unit or the Based on the deviation amount detecting means for calculating the relative displacement of the two images based on the data by the second processing means and the deviation amount calculated by the deviation amount calculating means, A focus detection device comprising: a focus detection unit that calculates a defocus amount by using a proportional constant that varies depending on a sampling pitch of data and detects the focus adjustment state. 2. A patent characterized in that the first filter processing mainly extracts low-order spatial frequency components, and the second filter processing mainly extracts high-order spatial frequency components. The focus detection device according to claim (1). 3. The focus detection device according to claim 1, wherein the first sampling pitch is set to be coarser than the second sampling pitch. 4. A pair of photoelectric conversion element arrays in which a large number of photoelectric conversion elements are arranged, a focus detection optical system for projecting two images of substantially the same portion of an object on the pair of photoelectric conversion element arrays, respectively. 2 based on the photoelectric output of the pair of photoelectric conversion element arrays
In a focus detection device that calculates a relative displacement of an image and detects the focus adjustment state of the imaging optical system based on the calculation result, a predetermined spatial frequency for the photoelectric output of the pair of photoelectric conversion element arrays. A first filter process for suppressing a component is performed, a spread of a focus detection region on the photoelectric conversion element array used for detecting the relative displacement is set, and the processed data is created. Processing means, and to the photoelectric output of the pair of photoelectric conversion element arrays, while performing a second filter processing different from the first filter processing, the focus detection area spread different from the first processing means. And a second processing means for creating the processed data, and a focus for detecting the focus adjustment state based on the data by the first processing means or the second processing means. Focus detecting apparatus characterized by comprising a means out. 5. The first filter processing mainly extracts low-order spatial frequency components, and the second filter processing mainly extracts high-order spatial frequency components. The focus detection device according to claim (4). 6. The spread of the focus detection area on the photoelectric conversion element array of the first processing means is set wider than the spread of the focus detection area on the photoelectric conversion element array of the second processing means. Claims characterized by
The focus detection device according to item (4). 7. A pair of photoelectric conversion element arrays in which a large number of photoelectric conversion elements are arrayed, a focus detection optical system for projecting two images of substantially the same portion of an object on the pair of photoelectric conversion element arrays, respectively. 2 based on the photoelectric output of the pair of photoelectric conversion element arrays
In a focus detection device that calculates a relative displacement of an image and detects the focus adjustment state of the imaging optical system based on the calculation result, a predetermined spatial frequency for the photoelectric output of the pair of photoelectric conversion element arrays. A first processing unit that performs a first filtering process that suppresses a component and creates the processed data, and a photoelectric output of the pair of photoelectric conversion element arrays that is different from the first filtering process. Second processing means for performing second filter processing to create the processed data, and a first information amount indicating the reliability of the photoelectric output based on the data by the first processing means, and the imaging optical system. The first defocus amount indicating the size of the out-of-focus is calculated, and similarly, based on the data by the second processing means, the second information amount indicating the reliability of the photoelectric output and the imaging optical system Big out of focus And calculating the second defocus amount indicating the degree, and making the first defocus amount and the second defocus amount dependent on the first and second information amounts or the first and second defocus amounts. A focus detection device comprising: a focus detection unit that calculates a combined defocus amount. 8. A patent characterized in that the first filter processing mainly extracts low-order spatial frequency components, and the second filter processing mainly extracts high-order spatial frequency components. The focus detection device according to claim (7).