JP3211280B2 - Focus detection device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

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

[0002]

2. Description of the Related Art There is a phase difference detection method as a focus detection method in a single-lens reflex camera, and this method will be described with reference to FIG. The luminous flux incident through the region 101 of the objective lens 100 is reflected by the field mask 200, the field lens 300, the aperture opening 401, and the re-imaging lens 50.
1 to form an image on the image sensor array A. Similarly, the light beam incident through the region 102 of the objective lens 100 is reflected by the field mask 200, the field lens 300,
The light passes through the aperture 402 and the re-imaging lens 502 to form an image on the image sensor array B. A pair of subject images formed on the image sensor arrays A and B are:
In the so-called front focus state in which the objective lens 100 forms a sharp image of the subject before the planned focal plane, they move away from each other. At the time of so-called focusing in which a sharp image of the subject is formed on the focal plane, the subject images on the image sensor arrays A and B relatively match. Therefore, this pair of subject images is converted to image sensor arrays A,
The signals are converted into electrical signals by photoelectric conversion in B, and these signals are arithmetically processed to obtain the relative positions of a pair of subject images, thereby obtaining the focus adjustment state of the objective lens 100, here, the amount of the object lens 100 away from the in-focus state and its amount. The direction (hereinafter, referred to as a defocus amount) is known.

Next, an arithmetic processing method for obtaining the defocus amount will be described. Image sensor arrays A and B are
Each of the photoelectric conversion elements includes a plurality of photoelectric conversion elements.
As shown in (a) and (b), a plurality of photoelectric conversion output signal strings a1. . . an, b1. . . bn is output. And
The correlation operation is performed while relatively shifting each data string by L for a predetermined amount of data. Specifically, the correlation amount C (L)
Is calculated by the following equation. C (L) = Σ | ai−bj | (1) where Σ represents the sum of i = k to r, and ji = L,
L = -lmax,. . , -2, -1, 0, 1,
2,. . , Lmax. L is an integer corresponding to the shift amount of the data string as described above, and the first term k and the last term r may be changed depending on the shift amount L. The shift amount giving the minimum correlation value among the correlation amounts C (L) thus obtained is multiplied by a constant determined by the pitch width of the optical system and the pitch of the photoelectric conversion elements of the image sensor arrays A and B shown in FIG. Is the defocus amount. However, the correlation amount C (L) is a discrete value as shown in FIG. 11, and the minimum unit of the detectable defocus amount is limited by the pitch width of the photoelectric conversion elements of the image sensor arrays A and B. I will.

To solve this problem, Japanese Patent Laid-Open No. Sho 60-37 discloses a method in which an interpolation operation is performed on the basis of a discrete correlation amount C (L) to newly calculate a minimum value Cex and to perform in-depth focus detection.
No. 513, disclosed by the applicant. In this method, as shown in FIG. 12, a shift amount Fm that gives the minimum value Cex is first calculated by a method of calculating the correlation amount C0 which is the minimum value and the correlation amounts C1 and C-1 of the shift amounts on both sides thereof. Fm = L + DL / E (2) where, DL = (C-1 -C1) / 2, E = MAX @ C1-
C0, C-1-C0}, and MAX {Ca, Cb} is Ca
And Cb. Then, the defocus amount is calculated as follows from the shift amount Fm that gives the minimum value Cex. DF = Kf × Fm (3) Here, Kf is a constant determined by the pitch width of the photoelectric conversion elements of the optical system and the image sensor arrays A and B shown in FIG.

It is necessary to determine whether the defocus amount obtained in this way truly indicates the defocus amount or the fluctuation amount of the correlation amount due to noise or the like. When the following condition is satisfied, the defocus amount is determined. The quantity is reliable. E> E1 and Cex / E <G1 Condition (1) Here, Cex = C0−DL, and E1 and G1 are certain predetermined values. E is a value that depends on the contrast of the subject, and the larger the value is, the higher the contrast is and the higher the reliability is. Cex / E mainly depends on the noise component and is 0.
The closer to, the higher the reliability. Then, when it is determined that there is reliability, the objective lens 100 is driven based on the defocus amount DF. The above correlation calculation, interpolation calculation, and condition determination are collectively referred to as focus detection calculation.

In the above description, the focus detection calculation is performed using the outputs of the image sensor arrays A and B. However, the harmful components having a frequency equal to or higher than the Nyquist frequency and the balance between the outputs of the pair of image sensors A and B are used. In order to eliminate the influence of the DC component, the output of the image sensor arrays A and B is filtered to generate new data, and the focus detection calculation is performed using the newly generated data. The method is described in, for example,
-245123. That is, assuming that the output signal sequence of the image sensor is D (i), first, a filter process for removing high frequency components is performed to create a new data sequence H (i). H (i) = D (i) + 3D (i + 1) + 4D (i + 2) + 3D (i + 3) + D (i + 4) (4) Note that the right side of the above equation is You may divide by 12. The data string H (i) is further subjected to a filtering process for removing a DC component to create a new data string F (i). F (i) =-H (i) + 2H (i + N) -H (i + 2N) (5) where N is a natural number of about 1 to 10, and the value of N is large. The lower the frequency component is extracted, the lower the frequency component becomes. Therefore, two types of N, for example, 3 and 6, are prepared and switched depending on the subject, or at first, a data sequence F (i) is created with N = 3, and this is used to perform the correlation operation of the equation (1), (2) If the formula (1) is not satisfied, the data sequence F (i) is created by changing N = 6, and the calculation is performed again. By the way,
When N = 3, the first data F (1) of the data sequence F (i) is converted to the data sequence D which is the output signal of the image sensor itself.
(i) is as follows. F (1) =-D (1) -3D (2) -4D (3) -D (4) + 5D (5) + 8D (6) + 5D (7) -D (8) -4D (9) -3D ( 10) -D (11) ... (6)

In such a focus detection device, when a plurality of subjects having different distances form an image on the image sensor array, if focus detection calculation is performed based on all data of the image sensor array, the forward and backward subjects are calculated. However, there is a problem that it is determined that the focus is achieved at the intermediate distance, or that the focus detection becomes impossible. Therefore, the subject image is subdivided by dividing each of the pair of image sensor arrays into a plurality of blocks, a focus detection calculation is performed for each block, and a shift amount Fm (i) and a defocus amount Df that give a minimum in each block. (i) is calculated, for example, a block having the closest defocus amount or the block having the largest contrast is selected from the plurality of blocks, and the defocus amount of the block is set to the focus detection state of the objective lens. A method of driving the objective lens based on the defocus amount and focusing is disclosed in, for example, JP-A-62-163007, and is known. Here, dividing into blocks means, for example,
R> (1) is used to form a plurality of pairs of the first term k and the last term r at the shift amount L = 0. As shown in FIG. 13A, in order to divide the image sensor array into five blocks each consisting of nine data, block 1 is set to k = 3 and r = 11 at a shift amount L = 0. The correlation amount C (L) is calculated by the equation (1), the shift amount Fm giving the minimum is calculated by the equation (2) based on these, the defocus amount DF is calculated, and the blocks 2, 3, and 4 are similarly calculated. , 5 and k at the shift amount L = 0
= 12, r = 20, k = 21, r = 29, k = 30, r
= 38, k = 39, r = 47 and the calculation is performed.

[0008] Even when the above-described filtering process is performed, newly created data may be similarly divided into a plurality of blocks. In the filtering process for extracting high-frequency components, the number of data forming blocks is reduced to form many narrow blocks, and in the filtering process for extracting low-frequency components, the number of data forming blocks is increased to increase wide blocks. May be formed as a decimal number. Alternatively, there is a method in which the focus detection calculation is performed in a narrow block at first, and if the focus cannot be detected in all the blocks, the block is widened and the focus detection calculation is performed again to make the width of the block variable.

In the above description, it is assumed that the subject images on the image sensor arrays A and B are relatively coincident with each other at the time of so-called focusing in which a sharp image of the subject is formed on the expected focal plane. However, in practice, when the optical system shown in FIG. 9 (excluding the objective lens 100) is mounted on a camera or the like, the relative position with respect to the planned focal plane is shifted. Shifts without matching. Therefore, it is necessary to store the deviation amount and to perform the correction after the calculation of the equation (2) is completed. For example, when a Z whose unit is the data pitch is used as a correction value for correcting the shift amount, the correction is performed as follows. Fm ′ = Fm−Z (7) Then, the defocus amount is calculated as follows instead of the equation (3). DF = Kf × Fm ′ (8) Since this correction value Z differs for each camera body, it is necessary to measure the correction value Z for each camera body and store it in, for example, an EEPROM in the camera body. .

Ideally, it is desired that the shift amount be constant irrespective of the position of the image sensor array. However, the shift amount is not constant due to the aberration of the focus detection optical system.
As shown in FIG. 14, the shift amount differs depending on the position of the image sensor array. Therefore, when dividing the focus detection area into a plurality of blocks, the correction value Z needs to be set for each block. Therefore, the correction value Z (i) is measured for each block, these plurality of correction values Z (i) are stored in an EEPROM or the like, and a corresponding correction value Z (i) is used for each calculated result. To make corrections. This is disclosed and known by the present applicant in Japanese Patent Application Laid-Open No. 60-262004. Hereinafter, the shift amount due to the position of the image sensor array is referred to as a Z value, and the distribution of the Z values is referred to as a Z distribution.

[0011]

However, in a focus detection device that divides a focus detection area into a plurality of blocks as described above, the output signal data of the image sensor or the output signal data obtained by filtering the output signal data is obtained. When performing focus detection calculation using the data
In addition to providing a correction value Z for correcting the aberration of the focus detection optical system for each of a plurality of blocks, it is necessary to change the correction value Z depending on the presence / absence of filter processing and the type of filter processing. When the width of each block is variable, it is necessary to change the correction value Z for each block. These problems will be described.

FIG. 13 shows a pair of image sensors.
It is assumed that the pixel consists of 49 pixels as shown in FIG. First, the 49 data are converted into 45 data strings H (1) to H (45) as shown in FIG. Further, 39 data strings F (1) are obtained by the filter processing (N = 3) according to the equation (5).
To F (39). If this data is to be divided into five blocks, for example, as shown in FIG. 13C, a block in which each block is composed of seven pixels can be formed. When represented by k and r in the equation (1), blocks 1 to 5 have k = 3, r = 9, k = 10, r = 16, and k = 3, respectively.
17, r = 23, k = 24, r = 30, k = 31, r =
37. Consider what these areas correspond to on the image sensor. As shown in the equation (6), the filter processing data F (1) is represented by D (1) to D (11).
However, since D (6) is the center and the coefficient is the largest, the filtered data F (1) is considered to be a signal corresponding to the light image at the position of the sixth pixel on the image sensor. . Therefore, the 1st to 5th blocks are the 8th to 14th pixels and 15th to 15th pixels on the image sensor, respectively.
21 pixels, 22-28 pixels, 29-35 pixels, 3rd
This corresponds to a range of 6 to 42 pixels.

However, when the data not subjected to the filtering process, that is, the output itself of the image sensor, is used as described above, the 1st to 5th blocks include the 3rd to 11th pixels, the 12th to 20th pixels, and the 21st to 29th pixels, respectively. Pixel, the 30th to 38th pixels, and the 39th to 47th pixels. Therefore, since the range of the block differs depending on the presence or absence of the filtering process, if only the data sequence when using the data sequence without the filtering process is stored as the correction value Z (i), the data sequence after the filtering process is performed. Correction is not performed when the calculation according to Therefore, it is necessary to store the correction value Z not only for each block but also for each of the presence and absence of the filter processing. Further, since the range of the block may be different depending on the type of the filter processing, it is necessary to store the data for each type of the filter processing. When the width of the block is variable, the range of the block is naturally different. Therefore, it is necessary to store the correction value Z for each width of the block.

However, this causes an increase in the memory area for storing the correction value Z, so that an EEPROM having a large memory capacity needs to be mounted, which leads to an increase in cost. In order to improve the aberration of the focus detection optical system so that the correction value Z does not change depending on the position of the image sensor, the number of lenses constituting the focus detection optical system may be increased, or an aspherical lens which is difficult to manufacture may be used. It is necessary to use an expensive lens material and to increase the cost, and the size of the optical system becomes large, so that it cannot be mounted on a small device such as a camera.

An object of the present invention is to store correction data corresponding to at least one focus detection mode, and to correct a focus detection error of a phase difference detection method caused by lens aberration in various focus detection modes based on the data. It is to provide a focus detection device for correction.

[0016]

(1) The present invention will be described with reference to FIG. 1 which is a diagram corresponding to a claim. Photoelectric conversion means 100 for guiding the light on a pair of light receiving element arrays, and outputting a signal corresponding to the light intensity distribution of the subject image formed on each of the light receiving element arrays; and each light receiving element array of the photoelectric conversion means 100 Dividing means into a plurality of blocks each having a predetermined number of light receiving elements as a unit.
1, calculating means 102 for calculating, for each of the blocks divided by the blocking means 101, a defocus amount indicating a focus adjustment state of the taking lens based on an output signal of a light receiving element included in each block; A storage unit 103 for storing a correction value of each block for correcting a focus detection error of each block, and a correction unit 105 for correcting a defocus amount of each block based on the correction value. In a focus detection device for obtaining a final defocus amount based on a focus amount, the blocking unit 101 has a plurality of blocking methods for performing division in units of different numbers of light receiving elements, and the storage unit 103 includes the plurality of storage units. Storing a correction value of each block corresponding to a specific blocking method among the blocking methods, When a blocking method other than the specific blocking method is selected by the means 101, the correction value of each block corresponding to the selected blocking method is calculated based on the correction value of the specific blocking method. And a correction value calculating unit 104 for performing the correction. (2) According to the second aspect of the present invention, a pair of subject light beams that have passed through a photographing lens are guided to a pair of light receiving element rows composed of a plurality of light receiving elements via a focus detection optical system, and formed on each of the light receiving element rows. A photoelectric conversion unit 100 for outputting a signal corresponding to the light intensity distribution of the subject image, and a filter processing unit 20 for performing a filtering process on an output signal of the photoelectric conversion unit 100
0, a blocking unit 201 that divides each light receiving element row of the photoelectric conversion unit 100 into a plurality of blocks each having a predetermined number of light receiving elements as a unit, and for each block divided by the blocking unit 201. Calculating means 202 for calculating a defocus amount indicating a focus adjustment state of the photographing lens based on an output signal of the light receiving element included in each block after filtering, and each block for correcting a focus detection error of each block And a correction means 205 for correcting the defocus amount of each block based on the correction value, and obtains a final defocus amount based on the plurality of corrected defocus amounts. In the focus detection device, the filter processing means 200 performs a plurality of types of filter processing,
The blocking unit 201 has a plurality of blocking methods for performing division in units of different numbers of light receiving elements, and the storage unit 203 stores a specific type of filter processing among the plurality of types of filter processing, A correction value of each block corresponding to a specific blocking method among the blocking methods is stored, and when a type of filter processing other than the specific type is selected by the filter processing unit 200, or When a blocking method other than the specific blocking method is selected by 201, the correction value of each block corresponding to the type of the selected filtering process or the blocking method is determined by the filtering process of the specific type and the filtering process. Correction value calculating means 204 for calculating based on the correction value corresponding to the specific blocking method
Is provided. (3) The focus detecting device according to claim 3 is a blocking unit.
According to 01 and 201, when the focus detection state cannot be detected, the blocking method is changed.

[0017]

According to the first aspect, a correction value of each block corresponding to a specific blocking method among a plurality of blocking methods for performing division in units of different numbers of light receiving elements is stored. When a blocking method other than the above blocking method is selected, the correction value of each block corresponding to the selected blocking method is calculated based on the correction value of the specific blocking method. (2) In claim 2, a correction value of each block corresponding to a specific type of filtering among a plurality of types of filtering and a specific type of blocking among a plurality of types of blocking is stored. , If a type of filtering other than a specific type is selected, or if a blocking method other than a specific blocking method is selected, each block corresponding to the selected blocking method or type of filtering is selected. Is calculated based on the correction values corresponding to the specific type of filter processing and the specific blocking method.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment will be described. This is to perform focus detection calculation using data that is not subjected to the filter processing.The focus detection area is divided into small blocks with narrow widths, and focus detection calculation is performed. The focus detection calculation is performed again by setting a large block having an increased width. FIG. 2 shows the first
1 shows the configuration of the embodiment. Reference numeral 1 denotes an objective lens, reference numeral 2 denotes an optical system shown in FIG. 9, and reference numeral 3 denotes a pair of storage type image sensor arrays, such as CCDs, composed of rows A and B shown in FIG. Reference numeral 4 denotes a calculation unit, and the image sensor array 3
Is divided into a plurality of blocks, a focus detection calculation is performed in each block, and a shift amount Fm (i), a defocus amount Df (i), and the like that give a minimum for each block are calculated. Reference numeral 7 denotes a selection unit, and a plurality of defocus amounts Df calculated in the plurality of blocks by the calculation unit 4
(i) The one satisfying the predetermined condition is selected. A drive control unit 8 drives the motor 9 based on the defocus amount selected by the selection unit 7. Reference numeral 10 denotes a storage unit that stores a plurality of correction values Z (i). Numeral 11 denotes a correction value processing unit, which corresponds to a block width or the like set in the focus detection calculation by the calculation unit 4 and obtains a new correction value Z '() from the correction value Z (i) stored in the storage unit 10. Create i). Note that the correction value Z (i) may be directly used as the correction value Z ′ (i). A correction unit 12 uses the correction value Z ′ newly created by the correction value processing unit 11 to calculate the shift amount Fm (i) that gives the minimum value in each block calculated by the arithmetic unit 4 by (7 )) To calculate Fm ′ (i). In this configuration, the operations of the calculation unit 4, the selection unit 7, the drive control unit 8, the correction processing unit 11, and the correction unit 12 are realized by a microcomputer or the like. Storage unit 10
May be a memory in a microcomputer or an external EEPROM.

Now, it is assumed that the image sensor array 3 is composed of a pair of sensor arrays (rows A and B), each of which is composed of 49 pixels. In addition, the combination of the filter processing and the block width used for the focus detection
Block mode) are the following two types. (1) Small block mode The data itself (49 data pairs) of the image sensor array 3 is divided into five blocks of nine data, and focus detection calculation is performed in each block (FIG. 3).
(A)). (2) Large block mode The data itself (49 data pairs) of the image sensor array 3 is divided into three blocks of 19 data, and focus detection calculation is performed in each block (FIG. 3).
(A)).

The storage unit 10 stores correction values Z (1) to Z (5) for blocks 1 to 5 in the small block mode. As a method of measuring the correction values Z (1) to Z (5), for example, the objective lens 1 is focused on a subject using a planar chart, and the calculations shown in (1) and (2) are performed in the small block mode. , The shift amount Fm that gives the minimum value in each block is determined by the correction amount Z.
(1) to Z (5). The correction amount processing unit 11 inputs which of the modes (1) and (2) is selected from the calculation unit 4 and the correction amounts Z (1) to Z (Z) input from the storage unit.
The processing corresponding to the mode selected in (5) is performed to calculate the correction amount Z ′ (i). As a processing method, for example, the following method can be considered. The correction values Z (1) to Z (5) measured as described above are used for the block 1 in the small block mode.
5 is the average of the Z distribution in each range, and as shown in FIG. 4, it is considered to be the Z value at the center data position (hereinafter abbreviated as M1 to M5). Therefore, in modes other than the small block mode, the Z value of the block center position is calculated by the interpolation calculation described below, and the correction value Z ′ is calculated.
And it is sufficient.

Now, as shown in FIG. 5, it is assumed that the center position of a block S in a certain mode other than the small block mode corresponds to the position of the data Ms'. And Ms'
It is assumed that it is between the center positions Mn and Mn + 1 of the block n and the block n + 1 in the small block mode. The correction value Z ′ (S) of the block S is calculated based on the correction values Z (n) and Z (n + 1) of the block n and the block n + 1 in the small block mode by linear interpolation using the following equation. Z ′ (S) = Z (n) + (Ms′−
Mn) × (Z (n + 1) −Z (n)) ÷ (Mn + 1−Mn) = Z (n) +
(Ms'−Mn) × (Z (n + 1) −Z (n)) ÷ 9
(9) When the calculation is performed according to the above equation, the correction values Z '(i) of each block in each mode are calculated as the correction values Z (1) to Z (5) and M1 to M5 of the small block mode (sensor array, respectively). 3, element numbers 7, 16, 25, 34, 43) are represented as follows. (1a) Small block mode
Z (1) to Z (5) are used as they are. Z '(1) = Z (1) Z' (2) = Z (2) Z '(3) = Z (3) Z' (4) = Z (4) Z '(5) = Z (5) (10) (2a) Large block mode Since the center positions of the blocks 1 to 3 are the element numbers 12, 25, and 38 of the sensor array 3, Z '(1) = Z (1) + (12- M1) × (Z (2) −Z (1)) ÷ 9 = Z (1) + 5 × (Z (2) −Z (1)) ÷ 9 Z ′ (2) = Z (3) + (25− M3) × (Z (4) −Z (3)) {9 = Z (3) Z ′ (3) = Z (4) + (38−M4) × (Z (5) −Z (4))} 9 = Z (4) + 4 × (Z (5) −Z (4)) ÷ 9 (11)

The correction unit 12 uses the correction value Z ′ (i) calculated by the correction value processing unit 11 to calculate the shift amount Fm (i) that gives the minimum value in each block calculated by the calculation unit 4 to ( Correction is performed by equation 7) to obtain Fm '(i).
The calculation unit 4 calculates the defocus amount Df (i) based on Fm ′ (i) calculated by the correction unit 12 according to equation (8). The selection unit 7 selects, for example, one indicating the shortest distance from the plurality of defocus amounts Df (i). The drive control unit 8 controls the motor 9 so that the objective lens 1 is moved by the defocus amount selected by the selection unit 7.

Next, the operation of the first embodiment will be described with reference to the flowchart shown in FIG. In step S1,
The charge accumulation in the image sensor array 3 is performed, and data is input from the image sensor array 3 in the subsequent step S2. In step S3, the block is divided into five blocks in the small block mode described above, and in step S4, the correlation of the equations (1) and (2), the interpolation calculation, and the condition (1) are determined in each block. Next, in step S5,
It is determined whether or not focus detection is not possible without satisfying the condition (1) in all blocks. If focus detection is not possible in all blocks, the process proceeds to step S11, and if there is a focus detectable block, the process proceeds to step S6. . In step S6, a correction value Z '(i) (i = 1 to 5) is calculated by equation (10). In step S7, the shift amount Fm (i) that gives the minimum by the equation (7) is corrected by the correction value Z ′ (i), and Fm ′ (i)
Is calculated. In step S8, the defocus amount Def (i) is calculated by equation (8), and in step S9, one defocus amount Def (i) that satisfies a predetermined condition such as, for example, the closest distance is selected. And step S10
Then, the objective lens 1 is driven by controlling the motor 9 based on the selected defocus amount.

On the other hand, in step S11, the image data is re-divided into three blocks in the large block mode. Next, in step S12, (1),
The correlation of equation (2), interpolation calculation, and determination of condition (1) are performed. In step S13, it is determined whether or not focus detection is impossible without satisfying the condition (1) in all blocks. If focus detection is not possible in all blocks, the process is terminated as it is. If so, the process proceeds to step S14. In step S14, a correction value Z '(i) (i = 1 to 3) is calculated by equation (11), and thereafter, the process proceeds to step S7. In step S13, when the focus cannot be detected in all the blocks in the large block mode, the process is terminated without driving the lens, but the focus can be detected while forcibly driving the lens in a predetermined direction. A so-called scan operation for searching for a position may be performed.

Next, a second embodiment will be described. This second
In the embodiment, first, the data that has not been subjected to the filter processing is divided into small blocks having a small width, and focus detection calculation is performed. When focus detection cannot be performed in any of the blocks, the filter processing is performed, and the data subjected to the filter processing is performed. Is divided into narrow blocks, and the focus detection calculation is performed again.
If the focus cannot be detected in any of the blocks, the data subjected to the filter processing is divided into large blocks having a wide width, and the focus detection calculation is performed again. FIG.
Shows the configuration of the second embodiment. In addition, the first shown in FIG.
The same reference numerals are used to denote the same elements as those constituting the embodiment, and the description will focus on the differences. Reference numeral 6 denotes a filter processing unit that performs filter processing on data of the image sensor array 3. The calculation unit 4 divides the data of the image sensor array 3 or the data subjected to the filtering process by the filtering unit 6 into a plurality of blocks,
In each block, the above-described focus detection calculation is performed to calculate a shift amount Fm (i), a defocus amount Df (i), and the like that provide a minimum for each block. Correction value processing unit 11
Is a new correction value Z ′ (i) from the correction value Z (i) stored in the storage unit 10 based on the presence / absence of filter processing of the data used in the focus detection calculation by the calculation unit 4 and the width of the block. Create processing. Note that the correction value Z (i) may be directly used as the correction value Z ′ (i).

Now, it is assumed that the image sensor array 3 comprises a pair of sensor arrays A and B, each of which comprises 49 pixels, as in the first embodiment.
The block mode in the focus detection calculation is as follows.
Suppose that it is. (3) Small block mode The data itself (49 data pairs) of the image sensor array 3 is divided into five blocks of 9 data, and focus detection calculation is performed in each block (FIG. 3).
(A)). (4) Small block filter mode The image processing unit 6 controls the image sensor array 3
The data (39 data pairs) obtained by applying the filter processing of the expressions (4) and (5) to the data of (4) is divided into five blocks each consisting of seven data, and focus detection calculation is performed in each block (FIG. 3 (b)). )). (5) Large block filter mode The image processing unit 6 controls the image sensor array 3
The data (39 data pairs) obtained by applying the filtering processes of the expressions (4) and (5) to the data of
It is divided into two blocks, and focus detection calculation is performed in each block (FIG. 3B).

The storage unit 10 stores correction values Z (1) to Z (5) for the blocks 1 to 5 in the small block mode. The method of measuring the correction values Z (1) to Z (5) is the same as in the first embodiment. The correction amount processing unit 11 sends the above-mentioned (3) to
Which mode is selected from among (5) is input, and processing corresponding to the selected mode is performed on the correction amounts Z (1) to Z (5) input from the storage unit. '(i) is calculated. The processing method is the same as in the first embodiment.
Then, the correction value Z 'of each block in each of the modes is set.
(i) shows correction values Z (1) to Z (5) and M1 to M5 in the small block mode (element numbers 7, 7 of the sensor array 3, respectively).
16, 25, 34, 43) as follows. (3a) Small block mode Naturally, Z (1) to Z (5) are used as they are, and are the same as equation (10). (4a) Small block filter mode The center positions of the blocks 1 to 5 are the element numbers 6, 13, 20, 27, and 34 of the sensor array 3, and these are the element numbers 11, 1 in the unfiltered data, respectively.
8, 25, 32, and 39 correspond to the following. Z ′ (1) = Z (1) + (11−M1) × (Z (2) −Z (1)) ÷ 9 = Z (1) + 4 × (Z (2) −Z (1)) ÷ 9 Z ′ (2) = Z (2) + (18−M2) × (Z (3) −Z (2)) ÷ 9 = Z (2) + 2 × (Z (3) −Z (2)) ÷ 9 Z ′ (3) = Z (3) + (25−M3) × (Z (4) −Z (3)) ÷ 9 = Z (3) Z ′ (4) = Z (3) + (32−M3 ) × (Z (4) −Z (3)) ÷ 9 = Z (3) + 7 × (Z (4) −Z (3)) ÷ 9 Z ′ (5) = Z (4) + (39−M4 ) × (Z (5) −Z (4)) ÷ 9 = Z (4) + 5 × (Z (5) −Z (4)) ÷ 9 (12) (5a) Large block filter mode Block 1 The center positions of ３3 are element numbers 10, 20, 3
0, which correspond to the element numbers 15, 25, and 35, respectively, in the data without filter processing, and are as follows. Z ′ (1) = Z (1) + (15−M1) × (Z (2) −Z (1)) ÷ 9 = Z (1) + 8 × (Z (2) −Z (1)) ÷ 9 Z ′ (2) = Z (3) + (25−M3) × (Z (4) −Z (3)) ÷ 9 = Z (3) Z ′ (3) = Z (4) + (35−M4 ) × (Z (5) −Z (4)) ÷ 9 = Z (4) + 1 × (Z (5) −Z (4)) ÷ 9 (13)

The correction unit 12 uses the correction value Z ′ (i) calculated by the correction value processing unit 11 to calculate the shift amount Fm (i) that gives the minimum value in each block calculated by the calculation unit 4 to ( Correction is performed according to equation (7) to obtain Fm '(i). The calculation unit 4 calculates the defocus amount Df (i) based on Fm ′ (i) calculated by the correction unit 12 according to equation (8).
The selection unit 7 selects, for example, one indicating the shortest distance from the plurality of defocus amounts Df (i). Drive control unit 8
Controls the motor 9 so that the objective lens 1 is moved by the defocus amount selected by the selection unit 7.

Next, according to the flowchart of FIG.
The operation of this embodiment will be described. In step S21, charge accumulation in the image sensor array 3 is performed.
At 22, data of the image sensor array 3 is input. In step S23, the block is divided into the five blocks shown in the small block mode described above, and in step S24, the correlation of the equations (1) and (2), the interpolation calculation, and the determination of the condition (1) are performed in each block. In step S25, it is determined whether or not focus detection is not possible without satisfying the condition (1) in all blocks. If focus detection is not possible in all blocks, the process proceeds to step S31. Proceed to step S26. Step S26
Then, the correction value Z '(i) (i = 1 to 5) is calculated by the equation (10). In step S27, the shift amount Fm (i) that gives the minimum by the equation (7) is corrected by the correction value Z ′ (i), and Fm
(i) is calculated. In step S28, the defocus amount Def (i) is calculated by equation (8), and in step S29,
One of the defocus amounts Def (i) that satisfies a predetermined condition such as the closest distance is selected. Step S3
At 0, the motor 9 is controlled based on the selected defocus amount to drive the objective lens 1.

On the other hand, in step S31, the data of the image sensor array 3 is subjected to a filtering process according to the equations (4) and (5). In step S32, the data subjected to the filtering process is divided into five blocks in the small block filter mode described above. In step S33, the correlation of the equations (1) and (2), the interpolation calculation, and the determination of the condition (1) are performed in each block. In step S34, it is determined whether or not focus detection is impossible without satisfying the condition (1) in all blocks. If focus detection is not possible in all blocks, the process proceeds to step S36. Proceed to step S35. In step S35, the correction value Z '(i) (i = 1
After calculating (5), the process proceeds to step S27.

In step S36, the filtered data is divided into three blocks in the large block filter mode described above, and in step S37, the correlation and interpolation calculations of the equations (1) and (2) are performed in each block. And the condition (1) is determined. Step S38
Then, it is determined whether or not focus detection cannot be performed without satisfying the condition (1) in all blocks. If focus detection cannot be performed in all blocks, the process is terminated as it is. Proceed to S39. In step S39, the correction value Z ′ (i) (i = 1 to
After calculating 3), the process proceeds to step S27. In step S38, when focus detection is not possible in all blocks in the large filter mode, the process is terminated without driving the lens. However, the lens can be detected while forcibly driving the lens in a predetermined direction. A so-called scan operation for searching for a position may be performed.

In the first and second embodiments, if the edge of the contrast of the object pattern is located at the boundary of the block, the calculation result may be unstable. In other words, based on the contrast state of the output of the image sensor array near the block boundary,
This problem can be prevented by using a method of moving the boundaries of blocks. In this case, the correction value Z ′ (i) is set based on the center position of the block range set after moving the block boundary. Also, when there are a plurality of filtering processes, the same processing as in the present invention may be performed.

As described above, the correction values of any one of the large block mode, the small block mode, the large block filter mode, and the small block filter mode are stored in the memory, and based on these correction values, The correction value of each block in the selected block mode is calculated by interpolation, and the focus detection error of each block is corrected according to these correction values. There is no need to store a dedicated correction value every time, and the storage capacity of the memory can be saved.

In the configuration of the above embodiment, the image sensor array 3 is a photoelectric conversion unit, the calculation unit 4 is a blocking unit and a calculation unit, the storage unit 10 is a storage unit, and the correction value processing unit 11 is a correction value calculation unit. The corrector 12 constitutes the corrector, and the filter processor 6 constitutes the filter processor.

[0035]

As described above, according to the first aspect of the present invention, a plurality of divisions are performed in units of different numbers of light receiving elements.
Corresponds to a specific blocking method among the blocking methods
The correction value of each block to be stored
If a blocking method other than the blocking method is selected,
Of each block corresponding to the selected blocking method
Calculate the positive value based on the correction value of a specific blocking method
To correct aberrations in the focus detection optical system.
The correction value for each block is calculated for each block method.
It is not necessary to prepare for
Only the correction value of each block corresponding to the
Memory, especially expensive EEPROM
The storage capacity of any nonvolatile memory can be reduced. According to the second aspect of the present invention, of the plurality of types of filter processing,
Certain types of filtering and multiple blocking methods
Of each block corresponding to the specific blocking method in
Memorize the positive value and filter other than the specific type
When processing is selected or when a specific blocking method
If an external blocking method is selected,
The type of blocking or filtering used.
The correction values for each block are
Calculated based on the correction value corresponding to the specific blocking method
To correct aberrations in the focus detection optical system.
The correction value for each block to be used
And the presence or absence of filtering and the type of filtering
It is necessary to prepare for each focus detection mode and store it in advance.
No specific type of filtering and specific blocking method
If only the correction value of each block corresponding to
Good, non-volatile memory, especially expensive EEPROM
The storage capacity of the volatile memory can be reduced.

[Brief description of the drawings]

FIG. 1 is a diagram corresponding to claims.

FIG. 2 is a block diagram showing the configuration of the first embodiment.

FIG. 3 is a diagram illustrating a division mode of the image sensor array.

FIG. 4 is a view showing a center data position of each block in a small block mode and a correction value corresponding thereto;

FIG. 5 is a view for explaining a method of calculating a correction value by an interpolation method in a mode other than the small block mode.

FIG. 6 is a flowchart showing the operation of the first embodiment.

FIG. 7 is a block diagram showing a configuration of a second embodiment.

FIG. 8 is a flowchart showing the operation of the second embodiment.

FIG. 9 is a diagram showing a configuration of a focus detection device using a phase difference detection method.

FIG. 10 is a diagram showing output data of an image sensor array.

FIG. 11 is a diagram illustrating a relationship between a shift amount of an output data string of an image sensor array and a correlation amount.

FIG. 12 is a diagram illustrating a method of calculating a shift amount that gives a minimum value of a correlation amount.

FIG. 13 is a diagram showing a method of dividing an image sensor array.

FIG. 14 is a diagram illustrating a shift amount of a subject image caused by a lens aberration or the like on a pair of image sensor arrays.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Objective lens 2 Focus detection optical system 3 Image sensor array 4 Operation part 6 Filter processing part 7 Selection part 8 Drive control part 9 Motor 10 Storage part 11 Correction value processing part 12 Correction part

Claims (3)

(57) [Claims] 1. A method for focusing a pair of subject light beams passing through a photographing lens.
Pair of light receiving elements via a point detection optical system
The object formed on each of the light receiving element rows
Photoelectric conversion means for outputting a signal corresponding to the light intensity distribution of the body image
And a predetermined number of light receiving elements,
And blocking means for dividing into a plurality of blocks in units of each blocked by said blocking means blocks
Output signal of the light-receiving element included in each block
Calculating means for calculating the defocus amount indicating the focus adjustment state of the taking lens based on each block ; and each block for correcting a focus detection error of each block
Storage means for storing a correction value, and a correction means for correcting a defocus amount of each block by the correction value, the final differential based on the plurality of defocus amounts of the corrected
In the focus detection device for obtaining the amount of focus, the blocking means separates the number of different light receiving elements into units.
Having a plurality of blocking methods for performing the division,
The storage means stores a specific block among the plurality of blocking methods.
The correction value of each block corresponding to the block forming method is stored, and the block forming unit stores the correction value after the specific block forming method.
If an external blocking method is selected,
The correction value of each block corresponding to the
Correction calculated based on the correction value of a specific blocking method
A focus detection device comprising value calculation means . 2. A method for focusing a pair of subject light beams passing through a photographing lens.
Pair of light receiving elements via a point detection optical system
The object formed on each of the light receiving element rows
Photoelectric conversion means for outputting a signal corresponding to the light intensity distribution of the body image
A filter processing unit for performing a filtering process on an output signal of the photoelectric conversion unit; and a light receiving element array of a predetermined number of light receiving element rows of the photoelectric conversion unit.
And blocking means for dividing into a plurality of blocks in units of each blocked by said blocking means blocks
Filters for light-receiving elements included in each block
Calculating means for calculating a defocus amount indicating a focus adjustment state of the photographing lens based on the output signal after processing ; and each block for correcting a focus detection error of each block
Storage means for storing a correction value, and a correction means for correcting a defocus amount of each block by the correction value, the final differential based on the plurality of defocus amounts of the corrected
In the focus detection device for determining the amount of focus, the filter processing unit includes a plurality of types of filter processing.
And the blocking means is a different light receiving element
There are multiple blocking methods that perform division in units of numbers.
And the storage means is configured to store the plurality of types of filter processing.
Specific types of filtering and the multiple blocking methods
Each block corresponding to a specific blocking method in the method
And the filter processing means stores the correction value of the type other than the specific type.
Filter processing is selected, or
Blocking means other than the specific blocking method
This filter is selected when the encryption method is selected.
-Each block corresponding to the processing type or blocking method
The correction value of the filter to the specific type of filter processing and the characteristic
Calculated based on the correction value corresponding to the fixed blocking method
A focus detection device comprising: a correction value calculating unit . 3. The focus detection device according to claim 1, wherein said blocking means changes said blocking method when a focus detection state cannot be detected.