JPS604914A - Focus detector - Google Patents

Abstract

PURPOSE:To reduce an error in detection due to the noncongruity between two images and to improve the precision of detection of shifted quantity by using signals corresponding to the difference between the outputs of line sensors which show two image patterns and patterns obtained by shifting said outputs by the specified number of bits. CONSTITUTION:Output signals of respective sensors are shifted in phase as specified to generate signals corresponding to the phase differences between the original signals and shifted signals. Then, those two difference signals corresponding to the output signals of the respective sensors are compared with each other to detect the interval between the two images. Namely, a CCD22 is used as the two sensors, and a photoelectric conversion signal outputted in time series from the CCD22 is delayed through a delay circuit 30 by a signal component corresponding to specific picture elements, and the signal passed through this delay circuit 30 and the signal which is not passed are supplied to a subtracting circuit 24 to obtain a difference signal. Consequently, a DC component in the original signal which may cause an error is suppressed and an effective AC component is emphasized by the comparison. Thus, focus detection precision is improved.

Description

TECHNICAL FIELD The present invention relates to a focus detection device for a camera that detects the focus state of a photographic lens by receiving object light that has passed through a photographic lens.

Prior Art The first lens of the photographic lens is symmetrical to the optical axis.
The object light beams that have passed through the first and second regions are respectively re-imaged to create two images, and the mutual positional relationship of these two images is determined to determine the amount of deviation of the image forming position from the expected focal position and its deviation. A focus detection device has already been proposed that can obtain the direction (whether the imaging position is in front of or behind the expected focus position, that is, whether the focus is in front or behind). The optical system of such a focus detection device has a configuration as shown in FIG.
4) Alternatively, there is a condenser lens (6) located further back from this surface, and a re-imaging lens (6) further behind that.
8) and (10), and on the imaging surface of each re-imaging lens there is a line sensor (for example, a CCD1 light receiving element).
121, (14J is arranged.Each line sensor (12
1.

(As shown in Figure 2, the images on 141 are close to the optical axis (18) and approach each other in the case of so-called front focus, in which the image of the object to be focused is formed in front of the planned focal plane. , on the other hand, in the case of rear focus, they are far from the optical axis (18).When the two images are in focus, the distance between the two corresponding points of the two images is determined by the configuration of the optical system of the focus detection device. be a certain distance.

Therefore, in principle, the focus state can be determined by detecting the distance between the two images. The following method is known as one of the methods for detecting this image interval.

That is, in FIG. 3, sensors (12), (141
For example, each of the photodiode cells a1-a1o, bt-bta includes 10 and 16 photodiodes. For convenience, it is assumed that the symbols attached to customer cells also represent the output of each cell. Here, if we consider the set of 10 consecutive cells in the sensor (14J), seven sets Bl, B2...B7 are created as shown in the figure.Which set of these seven sets has an image associated with the sensor (12 ) The focus state is known by detecting whether the image of the sensor (12) most closely matches the image of the sensor (14). .That is, cell al.

B2.1- Each output of alo and safety tz bt, bl
, ・-・b4o O) Between each output at = bt
, a2=bz, -1alo-=bx.

It is assumed that the relationship holds true. In this case, 5t=l
at-bll-4-1a2-B21-1-...1at
o-btol -... (1) 20, but S+ is group B! It is smaller than the results of similar calculations for other sets of images, and is the smallest among the results of calculations for all sets of images. In order to find a set that has such a minimum value, the above calculation is first performed for each set of images. Next, an operation is performed to find the minimum value from among the obtained calculation results. In this way, the focus state was detected. However, the minimum value among the calculated values obtained above corresponds to the case where the two images most closely match, which means that the patterns of the two images are congruent and the sensitivities of the two sensors are relative to each other. This is the case when they are equal.

If the congruence breaks down or there is a sensitivity difference between the sensors,
The above minimum value does not necessarily correspond to the state of coincidence between the two images, and this results in a focus detection error. The above will be explained using FIG. 4. Figure 4 +al shows a tip detection target with step-like bright and dark areas, and the part in the frame ■ is assumed to be imaged on the two sensors (121, (Iω). Figure 4 (bl is a solid line) (Outl), dotted line (out2)
is the sensor (12) for the detection target in Fig. 4 Fa+,
(This is a graph showing the output of 141.As shown in the graph, consider a case where the two outputs are not congruent and there is a difference between the outputs in the area corresponding to the bright part of the detection target.The graph shows a case where the image on the sensor (12) matches the image of the third group B3 portion of the sensor (1ω), and the two outputs are superimposed corresponding to the matching of the images. Now, there are two outputs (outl) and (ouL
2) is congruent, then 53-1Σ"" -bi-J-zl +z (2)1=
The calculated value S3, which is indicated by 1, becomes 0 and is the smallest among all the other total values. However, as shown in the graph, if the two are not congruent, the calculated value S3 is not the minimum, but rather the calculated value S2 when the graph of the output (outl) is shifted to the left by one pinch of the cell. is smaller. In other words, a detection error occurs by l pitches.

Now, if l pitch corresponds to, for example, 30μ, the detection error for l pitch is equivalent to the detection error in the optical axis direction of the photographic lens, for example, about l yrvr. Such an error is of an amount that poses a practical problem for, for example, a single-lens reflex camera.

On the other hand, the optical system in the focus detection device according to the present invention is
As shown in FIG. 2, the configuration is such that the two images formed on the upper and lower sensors are asymmetrical with respect to the optical axis, and this is one of the factors that disrupts the congruency of the two images.

Furthermore, field curvature occurs due to the aberration characteristics of the condenser lens and the reimaging lens, which causes congruence to deteriorate. However, it is possible to improve the field curvature by changing the condenser lens to an aspherical lens or by combining multiple lenses, but even then it is difficult to improve the field curvature to the point where complete satisfaction is obtained. be.

Furthermore, it is difficult in terms of production to configure and arrange the two re-imaging optical systems sufficiently symmetrically with respect to the optical axis of the photographing lens, which causes non-congruency between the two images. Due to the various factors mentioned above, it is inevitable that the two images or the output patterns corresponding to each image will not be congruent, and therefore, in the conventional image comparison method as described above, the occurrence of a focus detection error is unavoidable.

Purpose The present invention does not generate focus detection errors as described above.
The purpose is to provide an improved focus detection device,
A further object of the present invention is to provide a focus detection device with high detection accuracy.

Summary To achieve this objective, in the present invention, instead of the conventional method of directly comparing the output patterns of two sensors, the phase of each output signal of each sensor is calculated. Create a signal that is shifted by a predetermined amount in a predetermined direction, obtain a signal corresponding to the difference between this shifted signal and the original signal before shifting, and generate two signals corresponding to the output signals of each sensor obtained in this way. The gist is that the interval between two images is detected by mutually comparing the difference signals. In the embodiment of the present invention, a series of f CODs are used as two sensors, and a photoelectric conversion signal output from the ccn in time series is delayed by a signal corresponding to a predetermined pixel via a delay circuit. The signal that has passed through this delay circuit and the signal that has not passed are fed to a subtraction circuit to obtain a difference signal.

Example First, Fig. 4 (output of bl (outl), (out2
) Let us consider what kind of results the focus detection method in the apparatus of the present invention produces in the case of (2). The output 0uLl' shown in FIG. 5 is the output 0uLl' shown in FIG.
uL1 (; The shifted output is subtracted from the output 0uLl before the shift by shifting two pitches to the right.

The output ouL2' is similarly created from the output Ou[2. In FIG. 5 the two outputs 0utl'.

out2' is shown corresponding to the state in which two images are superimposed on each other as in Figure 4 (bl). When calculating the calculated values, the calculated value (53') is the smallest for the combination of outputs as shown in Figure 5.By the way, in Figure 5, the output ouLl' is set to 1.
Assuming that the pitch is shifted to the right side, and considering the calculated value S2' in this case, it can be easily found that 52'> S a'. In the case of FIG. 4 fbl, S3 should be the minimum, but the relationship S2<53 was established, resulting in a focus detection error.

As seen above, focus detection accuracy can be improved by using signals converted as shown in FIG. The reason for this improvement effect is thought to be that the DC component included in the original signal, which is the basis of the error, is suppressed, and the AC component, which is effective for comparison, is emphasized.

Next, FIG. 6 is a block diagram showing a first embodiment of the present invention, in which a COD (charge coupled device) is used as a sensor. Although the sensors (12+ and (14) are shown as separate sensors in Figure 1, one CCD formed on the same chip is actually used, and the output of each cell on the CCD is chronologically recorded. output from the CCD (22
Among the + cells, the code Ll is used for those that use the output.
, R2,...◆L28. R1, R2,...R30
It is shown with . Here, the portion of cell L1ζI・28 and the portion of cells R1 to R30 do not correspond to, for example, sensors l12) and 114) in Figure 1, respectively, and are oriented at 6 degrees, and the numbers assigned to each cell also indicate the output of each cell. shall be indicated. Further, the portion of cells L1 to L28 will be referred to as a reference portion, and the portion of cells -R1 to Rao will be referred to as a reference portion.

Now, from the CCD (22), for example, Ll, Lz, .
... R28° The integral high force is sent out in the order of R1, R2, ... R3o. Actually, the left side of output 1, output L
There is also an unused cell portion on the right side of the output Rso, between R1 and R1. Returning to FIG. 6, the output signal of the CCD sent out as described above is given directly to the input of the subtraction circuit example, and is also sent to the other side of the subtraction circuit via the delay circuit (30). input (28).

The delay circuit (30) can use, for example, a CCD and delays the output by four pixels. An example of a subtraction circuit is input (2
6) Subtract the signal Sd of the input example from the signal Sn.

Figure 17 shows examples of signals Sn and Sd for the reference section, and Figure 17 (bl shows subtraction and output (Sn-3d) for signals Sn and Sd of Figure 17 Fal). The first subtraction output for the reference part signal is shown as 11=Ls-Ll, and the last subtraction output is shown as 124=L2g-L24.Thus, the delayed output for 4 pixels and If you want to combine to get the subtracted output,
For the 28 cell outputs, 24 subtracted outputs are obtained, which is 4 fewer. Similarly, cell 3 from the reference part
26 subtraction outputs are obtained for 0 integral outputs. These subtracted outputs are sequentially converted into digital values by the analog-to-digital conversion circuit □□□, and then stored in the memo circle. When all the subtraction outputs are obtained, the correlation circuit (1) performs a calculation to determine the amount of deviation of the photographic lens using the data in the memo IJ (34). Hereinafter, the above subtracted output will be referred to as a secondary output, and its waveform will be referred to as a secondary image.

Note that the calculation method will be explained in detail later.

FIG. 7 is a block diagram showing another embodiment of the present invention;
The integral output from the CD (22) is first converted into a digital value by the analog-to-digital conversion circuit (38), and then the converted digital value is directly applied to the input (44) of the subtraction circuit (42), and is also delayed. It is applied via the circuit (40) to the other input (46) of the subtraction circuit (42).

The output of the subtraction circuit (42) is stored in a memo circle. Note that the delay circuit (4) and the subtraction circuit (42) are constructed of digital circuits, but each has the same intended function as the delay circuit (301) and subtraction circuit (28) in FIG. (34) The subsequent signal processing is the sixth
Same as in the figure.

Next, we will explain the data processing for calculating the amount of deviation using the data of the memo force, but before that, we will refer to FIG. (1
The relationship with the image shift amount e in 2) will be considered. still,
FIG. 10 is the same optical system as shown in FIGS. 1 and 2. In FIG. 10, the light ray (48) passes near the outer periphery of the exit pupil of the photographic lens, passes through the intersection ((3)) of the focal plane and the optical axis, and reaches a point (52) on the sensor. This ray (48) is, for example, - in the group of rays that form an image at the intersection (■).The ray (circle) is the ray (48)
This is from the same light source as , but only when the photographic lens is extended by a distance of d. This ray of light passes through the intersection point (the point at a distance g from the point) and reaches the point on the sensor. The ray image is obtained when the photographing lens is moved in by d, passing through the intersection with the focal plane (621) and reaching the point (64) on the sensor. , or considering the line segment between the intersection (circle) and (62) as one image on the focusing plane, the point (5) on the sensor
The line segment between 2) and (58) or the point (52) and (circle) is an image that is refocused for each image on the focal plane.The length of each line segment is therefore Using and e, β= e
/g, and this will be called the image magnification. (This image magnification β is determined by the configuration of the optical system such as lenses (61, (8), etc.) On the other hand, if the angle between the light ray (48) and the optical axis (18) is θ, then the distances g and d are Between Lanθ=
g/d holds true. Here, θ is the lens (6), (
81, the aperture aperture [9+, etc., is a constant that depends on the configuration of the optical system. From equation 2, the relationship 4 between d and e is as follows.

d-□ (3) βtanθ As shown in equation (3), if the amount of shift e of the image on the sensor is known, the amount of deviation d of the image from the in-focus position can be calculated. Strictly speaking, the image magnification β is not a constant;
It changes according to the shift amount e, and tends to be large when the front focus is on, and small when the back focus is on the image magnification at the time of focusing. Furthermore, it also varies depending on the position of the image on the sensor surface due to aberrations such as field curvature of the optical system. Therefore, as will be described later, when calculating a more accurate shift amount, an image magnification corresponding to the shift amount e is prepared in advance and used. Detection of the shift amount e and the deviation amount d will be described below.

Returning to Figure 6, the January memo example contains 24 secondary outputs 1i (i
=l, 2. ...24) and 26 subtraction outputs r1 (i=1.2...26) created from the cell outputs of the reference section
can be stored. Note that the memory address for storing each secondary output is determined in advance. Here, for convenience of subsequent signal processing, each secondary output li, fi is associated with cells of the reference section and the reference section, as shown in FIG. For example, secondary output 1+
is obtained by subtracting the output of the cell L5 from the output of the cell L1 in the reference section, and is associated with the signal representing the cell L3 located at the center of the cells L1 and LS. Below, other secondary outputs 12, 13---124. rt, r2.
- Thinking in the same way as γ26, each cell is L4
．． LS, . . . L26 are associated with signals representing R3R4, . . . R2g.

Next, consider dividing the secondary output sequence (the center) corresponding to the reference section into three signal blocks, the first block, the second block, and the third block, as shown in FIG.

The first block is a portion of secondary outputs 41 to 112, the second block is a portion of secondary outputs 15 to 12o, and the third block is a portion of secondary outputs 41 to 112.
This is the part of the next output 11s to lz4. Each of these three signal blocks corresponds to the cell L on the line sensor.
+ +-L16. L5-L24. L13~L2[1
corresponds to the image of each part of. In detecting the amount of deviation, a pattern comparison operation as shown in FIG. 3 is performed for each of the three blocks. In Figure 3, 10
A reference pattern A1 consisting of 16 cell outputs was compared in 7 ways with a reference section consisting of 16 cell outputs. More comparisons are possible, but in that case the number of cell outputs compared to each other would be less than 10. In the focus detection device of the present invention, the comparison is performed with the number of signals forming the reference pattern constant.

Now, returning to Figure 8, the first and third blocks contain 12
Since there are 26 secondary outputs in the reference circle, the first and third blocks can each be compared in 15 ways. Furthermore, since the second block has 16 secondary outputs, 11 comparisons are possible. By the way, referring to FIG. 2, the reimaging lens (81, (
The two images according to 10) are closer to the optical axis (18) in the case of front focus, and are further away from the optical axis (18) in the case of rear focus.
When in focus or near focus, it is located at an intermediate degree between the two. Therefore, the first block is used to detect the amount of deviation when the back focus is large, the second block is used to detect the amount of deviation when in focus or near focus, and the third block is used to detect the amount of deviation when the back focus is large. Used to detect the amount of deviation in the case of front focus.

Now, as a design standard, the second block and reference part (16
The configuration of the optical system is determined so that the in-focus state is achieved when the secondary outputs γ6 to γ21 (referred to as focusing blocks) of

Figure 9 (al) shows the state in which the second block is compared to the in-focus block. Figure 9 (bl) shows the state in which the second block is shifted to the right by five cells from this state. Five comparison scenes can be obtained by right-shifting.If the second block is shifted further, a portion of the second block will protrude from the reference part.Therefore, the comparison when the second block is shifted to the right is as follows. In the case of further shifting to the right, the first block is used for comparison. In the state of FIG. 9, the first block is the secondary output r7 to γ18 of the reference section. This state is called the "1 right 5 shift state". This right 5 shift state corresponds to the final right shift state of the second block and also corresponds to the right shift start state of the first block. The first block performs nine comparisons, starting from this 5-position shift state to the right 13-position shift state, as shown in FCI in Figure 9.The first block starts from the right 5-position shift state and performs nine comparisons up to the right 13-position shift state as shown in FCI in Figure 9. It is possible to shift to the left from the state, and six ways of comparison are possible except for the state shown in the figure, but these six ways of comparison are not adopted.In other words, detection in this case is performed by the second block. In the same way as above, the second block is changed from the state of cell 5 in FIG.
It can be shifted to the left by an amount, and five comparison scenes can be obtained. In this way, a total of 11 comparison scenes including the state shown in FIG. 9 are obtained by the second block.

The third block will be responsible for detecting a further shift to the left. FIG. 9 idl shows the case where the third block is shifted 13 places to the left. Similarly to the case of the first block, when using the third block, the 9th shift from the left 5 shift state to the left 5 shift state
A comparative scene of the street is obtained. FIG. 11 shows the area of the comparison scene that each block is responsible for. In FIG. 11, the position of the number 0 on the number line (6G) represents the contrasting scene of ta+ in FIG. Below, the position of number 5 is shown in Figure 9 (
It shows the contrasting scene of bl, and the position of number -13 is in Figure 9 +
It shows a contrasting scene of dl. As described above, in the embodiment of the present invention, the area is shifted by 13 cells as the total amount of movement in opposite directions from the case where the sensor (12) and the image on the circle are in focus. This is the amount detection area. In this way, if we know at which position on the number line (mark) the degree of agreement between the comparison results is highest, that is, the amount of movement of the two images (denoted as ΔE), we can calculate The amount of deviation can be calculated.

Note that half (ΔE/2) of the movement amount ΔE of the two images based on the in-focus state corresponds to the shift amount e in Equation 31.

The method for detecting the amount of image shift on the sensor will be described in more detail below. For convenience of explanation, the number line (66) in Fig. 11 is replaced with the number line t68), and for example, the number line (66) corresponding to the number -13 on the number line (66) is used.
The above numerical value 0 is defined to represent that the shift position of the third block is 0 when comparing the third block with the reference section from the viewpoint of signal processing in the circuit. Therefore, for example, the number 0 (
The number 5 on the number line (681) corresponding to the amount of movement (0) indicates that the shift position of the second block is 5.

Next, FIG. 12 will be explained. C in the $12 diagram
The configuration from Cl21 to the memory (circle) is the same as that in Fig. 7.
1 to fz6 are stored. The circuit block (72) performs a comparison operation between each signal block corresponding to the reference part and the reference part, and detects a shift position with the highest degree of coincidence for each block and a comparison output corresponding to the shift position. The contents of this comparison output will be described below. First, a comparison operation using the first block will be explained. The first block is composed of the signals of the secondary outputs 11 to lxz, as described above, and is compared with the secondary outputs r7 to r26 of the reference section in nine different cases. The first of the nine cases is the block of secondary outputs 11 to lxz and r7
This is a comparison of the contents expressed by the following equation with the blocks of ~r18.

Hlfol-X Ilk -7' to 10e I ”14
)k=x This corresponds to the shift position 0 of $1 pro tsk on the number line ((g) in Figure 11.Then, the second comparison content is Hl(1)-kAlllk-rk+71 ...(5), which corresponds to shift position 1 of the first block.Hereinafter, shift positions 3, 4, .
. . . Comparisons corresponding to each of 8 are performed. Shift position j of the first block (j=0.1, 2...8
) is shown in a general formula as town (Jl =
kJ: 1111=γ to O. +jI...(6). The calculated value town (
j) is called the comparison output. The smallest one among these nine comparison outputs is regarded as the one with the highest degree of coincidence in the comparison scene regarding the first block, and is denoted as m1nH1(j), and the corresponding shift position is expressed as 51(j). Indicated by To detect this m1nH1(j), for example, first Hl
When Hl(1) is calculated next to (o), the two are compared and the smaller one is selected. Next, when Hl(2) is calculated, the previously selected comparison output and Hl(2) are compared, and the smaller one is selected. We will proceed in the same manner below. In this way, when Hl(8) is calculated and compared with the previously selected comparison output, m1nH1(jl is determined.On the other hand, when detecting 5t(j), for example, 0 is initially set and the comparison output is If you compare a counter that is incremented every time an output is calculated and two comparison outputs, and the newly calculated comparison output is smaller, you can prepare a register that captures the contents of the counter. , 51(j) is obtained in the register at the time of comparison processing between the final comparison output CI(8) of the first block and the previously selected comparison output.However, the register is initially set to 0.

Next, for the second block, H2ijl = , 4th I/1k-H-γ +j l
...The comparison shown in (7) is performed, and m1nH2(
J) and 52fjl are detected. Tatashi, 14
In the case of a +2 block, j is an integer from 0 to 10. Here, H2+01 is secondary output 15~120 and rI~
This is the comparison output from the comparison with r+a, and corresponds to the case of shift position O of the second block. Similarly, for the third block, H3(j) -kgx'lk+t2-rk-1-j
The comparison shown in (8) is performed, and m1nH3
(J) and 53(j) are detected. However, in this case, j is from 0 to 8. Here, [3(0) is the third
Figure 10 shows the comparison output in the case of dl, and the detected value m1nH1 (jl, m1nH2
fjl s m1nH301, S1 mountain, 52 (jl, 5
3(j) is memory (Ml), (N2), (N3),
It is stored in (N4), (N5), and (N6).

Next, the circuit block (7) calculates a contrast signal defined as follows using the second-order output data of the memory (en).In general, the contrast of the image is calculated in two specific areas with different contrasts. For example, the difference between the outputs of adjacent cells is used as a signal indicating the contrast.In contrast, in the embodiment of the present invention, the difference between the outputs of two cells is used. Instead, the difference between one secondary output is called a contrast signal, and this is detected and used as a signal indicating the contrast of an image.Of course, the difference between two cell outputs can also be used as a contrast signal. However, in this embodiment, for the purpose of saving memory, a memory for decompressing and storing the output of each cell is not prepared, and the contrast signal is mainly used as a substitute for the secondary output.Here, The reason why a substitute contrast signal can be used to indicate the contrast of an image can be seen from Figures 4 and 5.The contrast signal is used when the contrast of the subject is too low to detect the focus. Thus, in the circuit block group, the contrast signals C1 and C expressed by the following equation for each signal block are
2, C3 is observed.

C3= Σ l lk+lz −lk+ul...
(11) Here, CI is the secondary signal 1l of the first block.
-1x2, for one secondary signal lk, the difference from two adjacent secondary signals (t signal lk+2) is calculated over 10 ways, and the sum of each absolute value is calculated. Although it is possible to reduce the difference between the outputs lk and lk+1, it is better to leave a gap equal to one secondary output because the difference can be emphasized and extracted, which is advantageous in subsequent processing. Then, C2 and C3 are the contrast signals for the second and third blocks.C1, C2,
C3 is memory (N7), (N8).

(N9) respectively.

The circuit block (76) calculates the minimum comparison output m1nH1(jl, m
1nH2 (jl.

m1nH2(J) N2(j)=□ (13) 2 is calculated as the ratio between each of the circuit blocks (Jl) and each of the contrast signals C11C2 and C3.Here, the first The minimum comparison outputs of block 7 and the third block are 12 comparison outputs 11 to l*z, respectively.
The second block is calculated using ls3 to lz4, or 16 comparison outputs of 15 to 12 degrees. Therefore, for the same image, the minimum comparison output of the first or third block and the minimum output of the second block are not the same, and the minimum comparison output of the second block, which has a larger amount of signal compared with each other, is becomes larger. For this reason, it is not appropriate to treat each as equal and to compare them for the purpose of detecting which of the three minimum comparison outputs indicates the shift position with the highest degree of agreement. On the other hand, the contrast signal CI.

In Cz and C3, C2 becomes larger than 01 or C3 for the same image. Although this is the case, by taking the ratio as in the above formula, N1(j), N2(j)
.

It is assumed that each of N5(j) has been converted into information that is equal to the other as a tendency. Note that these determined ratios are referred to as normalized minimum values. Further, they are stored in memories (MIO), (Mll), and (MN2), respectively.

Next, the circuit block (78) includes each contrast signal C
It is determined whether or not I, C2, and C3 are higher than 1 at a predetermined contrast determination level. For example, if the determination result is higher than 1, 11 is output; if not, 10' is output, and CC1, C3 are respectively output. (Stored in memory (Mlo), (!vl+t), (MN2) as (:C2, CC3).Furthermore, if all three contrast signals have not reached the + judgment level, one signal indicating this This signal is treated as indicating that the contrast of the subject is insufficient and focus detection is impossible.

Based on the determination result of the circuit block (78), the circuit block (80) selects one of the three normalized minimum values determined by the circuit block (76) with the highest matching degree as follows. Selected as a signal indicating a high shift position. first,
The result of contrast determination by the circuit block (78),
The normalized minimum values corresponding to blocks determined to have insufficient contrast are excluded, and the normalized minimum values of the remaining blocks are selected.

If there is no block with insufficient contrast, the normalized minimum value N1(j) of all blocks.

N2(j) and N5(j) are selected. The smallest normalized minimum value is selected from among the normalized minimum values selected in this way, and this is determined as indicating the shift position with the highest degree of coincidence. In other words, it is determined which position in the number line system () or (inscription) in FIG. 11 is the corresponding position of image coincidence. For example, in the case of N1(j)-N2(jl, N2(j) with the larger number after N is selected.The normalized minimum value selected here is MN(j). ), and the corresponding shift position is MS (j).These are stored in memories (MN4) and (Mls), and data indicating which block the corresponding signal block is is stored in memory (N16). .Through the above,
This means that the shift position is detected in units of cell arrangement pitch. By the way, for example, the cell arrangement pitch is 30μ;
When θ in equation (3) is 2.8° and image magnification β is 0.3, the amount of movement for one pitch is equivalent to the amount of deviation of the photographic lens in the optical axis direction, for example, about one pitch. Become. In other words, according to the above detection results, the detection accuracy is about lam. However, for example, with a single-eye camera, 50
Since detection accuracy on the order of μ is required, it is desirable to detect the shift position with a resolution of less than 1 pitch. FIG. 13 is a graph showing an example of a comparison output for each shift position when the photographing lens is at a certain specific position. In this case, the point at shift position 2 of the second block appears to be the matching position.

Next, a description will be given of a circuit blower function that uses the above detection results to calculate the amount of deviation with even higher accuracy.

The normalized minimum value MN (j) determined by the circuit block (821 is the circuit block mark) and the predetermined limit value 2
Make a comparison with If MN(j) > N2, it is assumed that focus cannot be detected. Then, based on the shift position MS(j) that is currently in place, if it indicates, for example, rear focus, control is performed to extend the photographic lens, and the focus detection operation is performed again from the collection of CCD data. conduct. Note that the value of N2 is determined experimentally.

A circuit block (841) is a shift position Ms corresponding to the normalized minimum value MN(j) determined in the circuit block (80).
For each of the three positions 5(j-1) and S(j+1), which are 1 bit before and after the cell tJ+ and that position, the comparison output H(j-] is obtained using the secondary output of the memo 1 month circle.
).

Add H(j) and H(j+1). Note that these three comparison outputs are 0, and the circuit pro ts (7) including the other comparison outputs.
2), but in order to save the number of memories in the system, all but the minimum comparison output of each signal block are discarded. If you have enough memory, you can create a circuit block (
It is also possible to store all the comparison outputs calculated in 7 steps in the memory, and to read out the comparison outputs for this soft position and the shift positions before and after it from the memory at the stage when the shift position MS (j) is determined. . Now, the requested shift position is, for example, shift position 6 of the first block.
If the circuit block (84) corresponds to
Comparison output for three shift positions 5, 6, and 7 is possible (5).

Hl (61, Hl [71) is set. Note that if the shift positions at both ends of the signal block correspond to MS(j), for example, if the shift position of the first block is 0, that position is set to the coincidence of the two images. It is regarded as the position, and the calculation in the circuit program is canceled.

The circuit block (80) calculates the true matching position of the two images using the three comparison outputs determined by the circuit block (80).
S(j) does not necessarily indicate the true matching position.

This means that the two images can only be compared intermittently for each unit of cell arrangement pitch, whereas the distance between the two images changes continuously. attributed to.

Here, in order to understand the method of calculating the matching point, the relationship between the three comparison outputs and the matching point will be considered with reference to the graph of FIG. 14. Figure 14 Fal shows two comparison outputs n(j-1
) and H(j+1) are equal to each other, and the shift position j in this case is considered to be a true match. Figure 14 Fbl.

+CI, [dl is the graph in Figure 14 (al) shifted to the left by 1/4 pitch, and corresponds to the case where the true matching position is shifted by 1/4 pitch. Figure 14 (bl, [dl) d
l indicates that the true matching position exists between the two shift positions corresponding to the smallest comparison output and the next smallest comparison output, and is closer to the shift position SM(j). . FIG. 14 fcl shows that when the minimum comparison output is two, the center point of their shift positions is a true match.

A calculation formula for finding such matching positions will be explained below. Now, consider the case of FIG. 15 (al) as an example. This corresponds to FIG. 12 (bl +c).
(j)-H(j-1) and the difference HC between the other adjacent comparison output H(j+1) of the minimum comparison output and the minimum comparison output
j+o−H(j), and shift each to the shift position j−
1 and j and the center point Q2 between shift positions j and j+1. In FIG. 15fbl, the two slopes considered as described above are shown at points P1. It is plotted as Pa.

Next, consider the line segment PtPa, and consider the intersection P2 with the horizontal axis X to be the true matching point. In this way, the distance χ from point Q2 to point P2 is calculated. Note that the cell arrangement pitch is taken as the unit of -distance for the time being, and this is defined as one unit.

Therefore, for the position XM of 22 points, this equation (16) can be applied to cases other than fal in FIG. Therefore, the circuit block (ffi) is (16
) to calculate the true matching position XM. The matching position XM found here belongs to any of the three signal blocks, and corresponds to one point on the number line (681J) in FIG. 11, for example, the shift position 2. 20. Also, if you look back on the number line (66), this shift position 2.20 becomes the numerical value -10,80 (=-13-1-) shown by the arrow (71).
0,20), which means that the two images of the reference part and the reference part have moved toward each other on the sensor by 10.80 pitches compared to when they are in focus, that is, the amount of movement. This shows that ΔE is 1.0.80 on the front pin side. Thus, when a point on the number line (68) is determined by the circuit block (86), the corresponding point on the number line (66), that is, the amount of movement ΔE, is determined. The determination of this ΔE is performed in circuit block t88.
). Referring to FIG. 11, the amount of movement ΔE corresponding to the shift position XM for the first block is ΔE−XM−
1-5...(17) The movement amount ΔE corresponding to the shift position
XM -13 (19). Furthermore, this amount of movement Δ
As mentioned above, half of E corresponds to the shift amount e. The circuit block (88) performs (17), (18) based on the information on the shift position

The ΔE required for calculating either of the equations (19) is based on the cell arrangement pitch.

The next circuit block ((a)) is (1) depending on the value of ΔE.
) Determine the image magnification β included in the equation. As mentioned above, β
Since is a quantity that changes depending on the shift ie, the relationship between β and e is determined experimentally at the design stage, and based on this, e
In other words, prepare β corresponding to ΔE in the memo January song.

In this case, ΔE is divided into multiple regions, multiple βs are prepared according to the divided regions, and a circuit block (in labor, ΔE
It is determined which divided area belongs to, and then β is read from the memory corresponding to the determined divided area and outputted to the next stage circuit block (92).

The circuit block (92) uses the above information to determine the amount of deviation d
Calculate. Here, the length of one pitch of the cell array is P
, the shift amount e becomes ΔE, P/2.

Therefore, (Equation 11 is expressed as the following equation.

ΔE β −−(20) However, K is a constant and corresponds to P/2tanθ, and Memo IJ (
M2 is prepared in advance. Thus, the circuit block (92
) calculates the deviation amount d by calculating the (addition) formula.

FIG. 16 is a block diagram showing a configuration in which the circuit of the focus detection device according to the present invention shown in FIG. 12 is implemented using a microcomputer. In FIG. 16, the microcomputer system is a general 8-bit one-chip microcomputer (for example, Type 6502 manufactured by Motorola) with an increased memory capacity. Using a microcomputer, the first
Of the circuit blocks shown in FIG. 2, a part of the configuration of the A-D converter circuit (2) and the subsequent stages are all microcomputer-based components. External to the microcomputer a) are an A-D conversion circuit (39), a switch for commanding focusing operation ((g), a COD drive circuit a [phase], a motor drive circuit (1, 02), and a display circuit (108). ) etc. are connected.

The A-D converter circuit (39) includes a voltage comparator circuit and a D-A converter circuit, and is connected to a microcomputer t during A-D conversion.
A time-varying digital code is given from %l, the corresponding analog voltage and the cell output from the CCD (22) are compared, and when the two reach a predetermined relationship, the digital code is determined. A-D conversion value.

The CCD drive circuit QOI generates control signals necessary for drive control of the CCD (22) from clock pulses supplied from the microcomputer ((g)). Based on this, power supply to the lens drive motor (104) is controlled to drive the photographing lens (106) to the 7il-focus position.The display circuit (108) controls the front focus, focus, and rear focus based on the signal of the amount of wear. Indications such as out-of-focus, out-of-focus, etc. are displayed in the viewfinder.The above focus detection device repeatedly performs the above-mentioned shift amount detection operation while the switch (98) is closed. is moved toward the in-focus position.Incidentally, between the CCD ■ and the A-D conversion circuit (39), for example, a circuit for removing the dark output component peculiar to the CCD,
A circuit for amplifying the output signal of each cell according to the subject brightness is provided, but its explanation will be omitted since it is not directly related to the present invention. In addition, the actual device is
Needless to say, it is constructed by applying various known techniques, such as changing the integration time of D depending on the intensity of light incident on the cell.

Effects As described in detail above, the present invention provides a focus detection device that detects the relative positional relationship between two images and detects the amount of deviation of a photographing lens. Instead of comparing the two images with each other, the present invention uses a signal corresponding to the difference between the output of the line sensor and the output shifted by a predetermined number of bits. Figure 2 is a schematic configuration diagram of the optical system of the focus detection device related to the 4, a partially enlarged view of the optical system in Figure 1, Figure 3 is an explanatory diagram for explaining the conventional detection method, @Figure 4 is a graph showing the sensor output in FIG. 3; FIG. 5 is a graph for explaining the present invention in detail; FIG.
The figure is a block diagram showing the second embodiment of the present invention, and FIG. 8 is an example of the arrangement and configuration of each cell of the line sensor used in the focus detection device of the present invention, and the relationship between each cell and the secondary output. FIG. 9 is an explanatory diagram showing a comparison of the secondary outputs of the reference part and the reference part in FIG.
Fig. 0 is an explanatory diagram showing the relationship between the amount of grazing d of the photographing lens and the amount of shift e of the image of the line sensor 2, and Fig. 11 is an explanatory diagram for explaining the signal processing procedure of the focus detection device of the present invention. An explanatory diagram, FIG. 12 is a block diagram showing the configuration of the correlation calculation circuit section in FIG. 7, FIG. 13 is a graph showing an example of a comparison output of two secondary outputs, and FIG. 14 and @1
Figure 5 is a graph for explaining the calculation method for determining the relative position of two images on the line sensor, and Figure 16 is a graph when the circuit block in Figure 12 is constructed using a microcomputer. The block diagram, FIG. 17, is a graph showing the relationship between the line sensor output and the secondary output.

2: Shoot! Lens, 12, 14: Photodiode cell row, 0uLl', 0uL2': Secondary signal Applicant: Minolta Camera Co., Ltd. Figure 1 Figure 2 Figure 3 Figure 13 Figure 74 /-/ ♂ ♂? / /77

Claims (1)

[Claims] 1. Determine the relative positional relationship between the first and second images created from the subject light beams that have passed through the first and second parts of the photographic lens, which are sandwiched by the optical axis of the photographic lens. A focus detection device that detects the amount of deviation from the focus position of the photographing lens receives the first and second images and outputs an image signal according to the light intensity distribution pattern of each image. a line sensor in which a predetermined number of photodiode cells are arranged in a line, and a circuit for obtaining a secondary signal corresponding to the difference between the image signal and the image signal shifted by a predetermined number of photodiode cells; Focus detection characterized in that the relative positional relationship between the first and second images is detected by comparing the secondary signals corresponding to each of the first and second images. Device.