Focus position detection device, imaging apparatus, and focus position detection method

Abstract

To improve focus position detection accuracy for a subject having a large number of high-frequency components.SOLUTION: An image plane phase difference pixel processing unit 115 as a focus position detection device performs estimation processing of estimating a plurality of estimated pixels ZA', ZB' arranged between respective phase difference pixels ZA, ZB on the basis of data of two pixels Gr for image which are included in a row adjacent to a row in which the plurality of phase difference pixels ZA, ZB are arranged in a pixel array of an imaging element 102, and which are arranged in the same columns as the plurality of phase difference pixels ZA, ZB. Furthermore, a ratio between an average (graph C) of characteristics of pixel data along two rows including the two pixels Gr for image and a sum (graph A+B) of characteristics of pixel data of the phase difference pixels ZA, ZB is grasped in advance adjustment, and this ratio is used as a correction factor F in the estimation processing.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present invention relates to a focal position detection device, an imaging device, and a focal position detection method. [Background technology]

[0002] As a method of autofocus detection for imaging devices, on-chip phase-difference AF, which performs phase-difference AF using signals obtained from an image sensor, is already known. An image sensor for achieving on-chip phase-difference AF has, in addition to imaging pixels for obtaining image data, a phase-difference pixel A that is partially shielded so as to receive light beams that have passed through part of the pupil region of the photographing lens, and a phase-difference pixel B that is paired with the phase-difference pixel A and partially shielded so as to be paired with the phase-difference pixel B, which are arranged in predetermined positions. Then, a phase difference is calculated between the pixel data string of phase-difference pixel A and the pixel data string of phase-difference pixel B, to perform focus detection.

[0003] Patent Document 1 describes a highly accurate refocusing focus detection method using signals from discretely arranged phase difference pixels. The refocusing method involves adding pixel data strings of phase difference pixels A and B pixel by pixel to generate an added data string. The relative positions of the pixel data strings of phase difference pixels A and B are then shifted little by little in the column direction to generate similar added data strings in stages. Finally, a contrast evaluation value is calculated for each added data string to estimate the contrast peak position and perform focus detection. Patent Document 1 describes how the density of the phase difference pixel data string is apparently increased by estimating phase difference pixel data between discretely arranged phase difference pixels, thereby improving focus detection accuracy. Summary of the Invention [Problem to be solved by the invention]

[0004] However, there is a demand for improving focus detection accuracy for subjects with many fine patterns (i.e., many high-frequency components). Patent Document 1 does not address this point of view, and there is room for improvement.

[0005] An object of the present invention is to improve the accuracy of focus detection for a subject with a large amount of high frequency components. [Means for solving the problem]

[0006] In order to solve the above-described problems, a focus position detection device according to one aspect of the present invention is a focus position detection device that detects a focus position based on pixel signal values ​​from an image sensor in which image pixels that receive an object image via an imaging lens group and photoelectrically convert the image, and pairs of first and second phase difference pixels that receive and photoelectrically convert light beams that pass through paired pupil regions of the imaging lens group, are arranged, the focus position detection device estimating a plurality of first estimated phase difference pixels that are arranged between each of the plurality of first phase difference pixels based on data of two image pixels that are included in a row adjacent to a row in which a plurality of pairs of the first phase difference pixel and the second phase difference pixel are arranged and that are arranged in the same column as each of the plurality of second phase difference pixels, and an estimation unit that performs estimation processing to estimate a plurality of second estimated phase difference pixels that are arranged between each of the plurality of second phase difference pixels based on data of two image pixels that are rare and arranged in the same column as each of the plurality of first phase difference pixels, and a focus detection unit that detects a phase difference and calculates a defocus amount to a just-in-focus position by using a data string of the first phase difference pixels to which the plurality of first estimated phase difference pixels estimated by the estimation unit have been added and a data string of the second phase difference pixels to which the plurality of second estimated phase difference pixels have been added, wherein the estimation unit grasps, through advance adjustment, a ratio between an average of characteristics of pixel data along two rows including the two image pixels and a sum of characteristics of pixel data of the plurality of first phase difference pixels and characteristics of pixel data of the plurality of second phase difference pixels, and uses the ratio in the estimation processing. [Effects of the Invention]

[0007] This makes it possible to improve the accuracy of focus detection for a subject with a large number of high frequency components. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a perspective view showing an example of the appearance of a digital camera as an example of an imaging apparatus according to an embodiment; [Figure 2] FIG. 1 is a diagram showing an example of functional blocks of a digital camera. [Figure 3] FIG. 1 is a diagram showing an example of an arrangement of image pixels and phase difference pixels in an image sensor; [Figure 4] A diagram showing an example of a subject with many high-frequency components [Figure 5] FIG. 5 is a diagram showing an example of a focus position detection result in a state where the subject shown in FIG. 4 has a large amount of high-frequency components and is in focus using a conventional automatic focusing method. [Figure 6] FIG. 1 is a diagram illustrating a concept of a method for estimating phase difference pixel data between phase difference pixels. [Figure 7] FIG. 10 is a diagram showing an example of phase difference pixel data and image pixel data for a certain row when the image sensor is exposed to uniform light through a photographing lens. [Figure 8] FIG. 10 is a diagram showing an example of a focus position detection result using a pixel data string of phase difference pixels after estimation by the phase difference pixel estimation method of the present embodiment. [Figure 9] 1 is a flowchart illustrating an example of a pre-adjustment process. [Figure 10] 10 is a flowchart illustrating an example of an imaging surface phase difference detection process. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, embodiments will be described with reference to the accompanying drawings. To facilitate understanding of the description, the same components in the drawings will be denoted by the same reference numerals as much as possible, and duplicated descriptions will be omitted.

[0010] <Digital camera configuration example> FIG. 1 is a perspective view showing an example of the appearance of a digital camera 1 as an example of an imaging device according to an embodiment.

[0011] As shown in FIG. 1, digital camera 1 has a body 4 that is roughly rectangular parallelepiped in shape. Attached to the front surface 4A of body 4 of digital camera 1 is a group of taking lenses 2. This group of taking lenses 2 is broadly composed of two components. The first is a group of zoom lenses 6 (see FIG. 2) that change the magnification of the subject, and the second is a group of focus lenses 7 (see FIG. 2) that adjusts the focus on the subject. Arranged behind the group of taking lenses 2 are an aperture / shutter 103 (see FIG. 2) and an image sensor 102 (see FIG. 2).

[0012] A release button 3 is provided on the top surface 4B of the body 4 of the digital camera 1, and it is possible to distinguish between a half-pressed state and a fully-pressed state. When half-pressed, it serves as an automatic focusing (hereafter referred to as AF) start trigger, and when fully pressed, it serves as a shooting start trigger. A group of operation buttons 116 (see FIG. 2) and an image display unit 114 (see FIG. 2) that displays a monitoring screen, etc., are arranged on the back surface 4C of the body 4 of the digital camera 1. The back surface 4C is the surface on the opposite side of the normal to the front surface 4A, and is the surface at the back in FIG. 1.

[0013] Fig. 2 is a diagram showing an example of functional blocks of the digital camera 1. Fig. 2 schematically shows the shape of the digital camera 1 shown in Fig. 1 as viewed from the right in the figure, with the front surface 4A of the main body 4 facing left in the figure, the back surface 4C facing right in the figure, and the top surface 4B facing upward in the figure. In addition, the photographing lens group 2 attached to the front surface 1A of the main body 4 is arranged on the left side of the main body 4 in the figure.

[0014] The photographing lens group 2 is attached to the front surface 4A of the body 4 of the digital camera 1, and is made up of a zoom lens group 6 that changes magnification and a focus lens group 7 that adjusts focus. The light receiving surface of the image sensor 102 is arranged inside the photographing lens group 2 via an aperture / shutter 103. Data obtained from the image sensor 102 is sent to an image processing unit 107 and an image plane phase difference pixel processing unit 115.

[0015] The image processing unit 107 performs image processing and stores the image data in the image memory 108. An image display circuit 109 is connected to the data path of the image memory 108. The microprocessor 110 displays an image on the image display unit 114 via the image display circuit 109, and writes or reads the image to or from a memory card 118 via a card interface 117. The image plane phase difference pixel processing unit 115 calculates the phase difference from the phase difference pixel data to determine the defocus amount. The image plane phase difference pixel processing unit 115 corresponds to the focus position detection device according to this embodiment.

[0016] The microprocessor 110 is also connected to an image sensor control circuit 111 to control the image sensor 102. The microprocessor 110 is connected to a lens drive mechanism 104 and an encoder 105 to control the drive of the zoom lens group 6 and the focus lens group 7. The microprocessor 110 is connected to an aperture / shutter 103 to control the exposure amount and exposure time. The release button 3 arranged on the top surface 4B of the body 4 of the digital camera 1 and the operation button group 116 arranged on the back surface 4C are also connected to the microprocessor 110, and can obtain information on button operations by the user.

[0017] <Example of pixel configuration of image sensor> FIG. 3 is a diagram showing an example of the arrangement of the image pixels 203 and the phase difference pixels ZA and ZB in the image sensor 102. In FIG. 3, the horizontal arrangement in the diagram is expressed as "columns," and the vertical arrangement is expressed as "rows." FIG. 3 representatively shows 8 columns x 6 rows of pixels. Also, in FIG. 3, the gray areas represent the light receiving area of ​​each pixel. A color filter is attached on each pixel.

[0018] The pixel group 201 is one unit of the image pixels 203 and includes four pixels arranged in two columns and two rows. Of the four pixels, pixel R is equipped with a red color filter, pixels Gr and Gb are equipped with green color filters, and pixel B is equipped with a blue color filter. In the example of Figure 3, pixel R is arranged in the upper left of the pixel group 201, pixel Gr is arranged in the upper right, pixel Gb is arranged in the lower left, and pixel B is arranged in the lower right.

[0019] The image pixels 203 are configured by repeating such pixel groups 201 along the column and row directions. This is the basic pixel configuration of the image pixels 203, and this arrangement is generally called a Bayer array.

[0020] In the example of Fig. 3, the phase difference pixel is arranged in the fourth row 202 from the top, numbered "3," in this Bayer array. As shown in Fig. 3, in the row 202, phase difference pixels ZA (first phase difference pixels) and phase difference pixels ZB (second phase difference pixels) are arranged alternately along the column direction in a portion of the image pixels 203 that would originally be pixels B. The color filters on the phase difference pixels ZA and ZB are green, not blue.

[0021] The light-receiving area of ​​the phase detection pixel ZA is the left half of the pixel, and the right half is light-shielded. The phase detection pixel ZA receives light beams that have passed through the left pupil partial area of ​​the photographing lens (as viewed from the imaging surface). The phase detection pixel ZB does the opposite, forming a pair. In other words, the light-receiving area of ​​the phase detection pixel ZB is the right half of the pixel, and the left half is light-shielded. The phase detection pixel ZB receives light beams that have passed through the right pupil partial area of ​​the photographing lens (as viewed from the imaging surface).

[0022] <Issues with conventional focus position detection> The image plane phase difference pixel processing unit 115 calculates the phase difference and determines the defocus amount from the data of the phase difference pixels ZA and ZB received from the image sensor 102 having the pixel configuration shown in Fig. 3. Here, with conventional automatic focusing using image plane phase difference, there is room for improvement in the focus detection accuracy for subjects with a large number of high-frequency components.

[0023] Fig. 4 is a diagram showing an example of a subject 601 with a large amount of high-frequency components. Fig. 4 illustrates an example of the subject 601, in which the left half in the x direction (horizontal direction in the figure) of the screen is black and the right half is white. Fig. 4 also illustrates an AF area 602 in the form of a dotted rectangle in the center of the screen.

[0024] The object 601 illustrated in Fig. 4 is a step chart that changes suddenly from black to white at the center of the screen along the x direction of the screen. When such an object 601 is expressed using the configuration of the image sensor 102 shown in Fig. 3, it is a typical example of an object pattern with high frequency in the column direction of the phase difference pixels ZA and ZB, that is, an "object with many high frequency components."

[0025] Fig. 5 is a diagram showing an example of a focus position detection result when the subject 601 with a large amount of high frequency components shown in Fig. 4 is in focus (focused) using a conventional automatic focusing method. The vertical axis of Fig. 5 represents the distance measurement result (defocus amount) (µm). The horizontal axis of Fig. 5 represents the subject position, which represents the position of the subject 601 in Fig. 4 in the phase difference pixel column direction (x direction). The distance measurement result in Fig. 5 represents the result when the subject is shifted in the x direction in units smaller than one pixel of the image sensor 102.

[0026] The distance measurement result is ideally 0 μm for any object position because the object is in focus. However, the results of the conventional method shown in Figure 5 show that large distance measurement errors occur depending on the object position. In the example of Figure 5, a distance measurement error of over 200 μm occurs. The reason for this error is that the phase difference pixel data string is sparse, and the edge pattern of the object 601 cannot be accurately represented by the data string of the phase difference pixels ZA and ZB.

[0027] In response to such a conventional problem, in this embodiment, the phase difference pixel data between the phase difference pixels ZA and ZB is estimated, thereby correcting the phase difference pixel data string so that it becomes dense.

[0028] <Phase difference pixel estimation method> Fig. 6 is a diagram showing the concept of a method for estimating phase difference pixel data between phase difference pixels ZA and ZB. Fig. 6 shows three rows of image pixels 203 including row 202 in Fig. 3. As described above, row 202 includes phase difference pixels ZA and ZB.

[0029] Here, the row i and column j in which pixel Z is located are denoted by Z i,j The pixel data string of the phase difference pixel ZA before estimation is expressed as {ZA 3,1 ,ZA 3,5 ,}, the pixel data string of the phase difference pixel ZB is {ZB 3,3 ,ZB 3,7 ,...}, resulting in a data string with every four pixels.

[0030] In this embodiment, the phase difference pixel data between the phase difference pixels ZA and ZB is estimated so that such a phase difference pixel data string becomes dense. Specifically, as described above, the light receiving area of ​​the phase difference pixel ZA is the left half of the pixel, and the right half is shielded from light. Therefore, as shown in FIG. 6, the estimated pixel ZB' (second estimated phase difference pixel) of the phase difference pixel ZB is estimated on the assumption that the estimated pixel ZB' is located in the right half of the same pixel where the phase difference pixel ZA is located. Similarly, as described above, the light receiving area of ​​the phase difference pixel ZB is the right half of the pixel, and the left half is shielded from light. Therefore, as shown in FIG. 6, the estimated pixel ZA' (first estimated phase difference pixel) of the phase difference pixel ZA is estimated on the assumption that the estimated pixel ZA' is located in the left half of the same pixel where the phase difference pixel ZB is located.

[0031] By this estimation process, the pixel data string of the estimated phase difference pixel ZA is {ZA 3,1 ,ZA´ 3,3 ,ZA 3,5 ,ZA´ 3,7 ,}, and the pixel data string of the phase difference pixel ZB is {ZB´ 3,1 ,ZB 3,3 ,ZB´ 3,5 ,ZB 3,7 ,···}, resulting in a dense data string with every two pixels.

[0032] The estimated pixel ZA′ of the phase difference pixel ZA can be calculated using the following equation (1).

[0033]

number

[0034] The estimated pixel ZB' of the phase difference pixel ZB can be calculated using the following equation (2).

[0035]

number

[0036] In this way, by making an estimation using neighboring pixel data on the same column as the estimation pixel, rather than using neighboring pixel data shifted in the phase difference pixel column direction (horizontal direction in FIG. 4), it is possible to make an estimation with a relatively high degree of accuracy even for subject patterns with high frequency in the phase difference pixel column direction, that is, for “subjects with many high frequency components.”

[0037] Furthermore, by correcting with a correction factor F that has been adjusted and secured in advance, the effects of manufacturing variations, luminous flux vignetting, and lens aberration can be absorbed, enabling highly accurate estimation.

[0038] Here, the necessity of the correction factor F in the above estimation process will be described with reference to FIG. 7. FIG. 7 is a diagram showing an example of phase difference pixel data and image pixel data of a certain row when the image sensor 102 is exposed to uniform light through a photographing lens. Here, "a certain row" refers to a row that includes phase difference pixels ZA and ZB, such as row 202 shown in FIG. 3. The horizontal axis in each of FIGS. 7(a), (b), and (c) indicates the column number of each pixel in a certain row of the image sensor. The horizontal axis is equivalent to the x-axis of the image sensor 102, and the center of the horizontal axis corresponds to the center of the image sensor.

[0039] 7(a) shows graphs A and B of the characteristics of the phase difference pixel data (ZA, ZB), and graph C of the characteristics of the image pixel data (upper and lower Gr average). The vertical axis of FIG. 7(a) indicates the data value of each pixel. In FIG. 7(a), graph A is shown by a solid line, graph B by a dotted line, and graph C by a thick solid line. Here, the "upper and lower Gr average" refers to the average value of data of two pixels Gr arranged in the same column as the phase difference pixels ZA and ZB, which are arranged in rows above and below the row including the phase difference pixels ZA and ZB, as shown in FIG. 3, for example.

[0040] As shown in FIG. 7(a), characteristic C of the image pixel data (average of upper and lower Gr) is greatest at the center of the image sensor 102 and tends to decrease in value toward the edge of the image sensor 102. Characteristic A of the phase difference pixel data (ZA) tends to be higher on the left side of the image sensor 102 and lower on the right side. Characteristic B of the phase difference pixel data (ZB) tends to be higher on the right side of the image sensor 102 and lower on the left side. These characteristics are caused by the influence of general shading (peripheral light falloff).

[0041] 7(b) shows graph A+B of the sum of the characteristics of the phase difference pixel data (ZA) and the characteristics of the phase difference pixel data (ZB). The vertical axis of FIG. 7(b) also shows the data value of each pixel, as in FIG. 7(a). FIG. 7(b) also shows graph C of the characteristics of the same image pixel data (average of upper and lower Gr) as FIG. 7(a).

[0042] As shown in FIG. 7(b), the characteristic A+B of the phase difference pixel data (ZA+ZB) is greatest at the center of the image sensor 102 and tends to decrease toward the edge of the image sensor 102. This tendency of characteristic A+B is similar to the tendency of characteristic C of the image pixel data, but the data values ​​themselves do not match those of characteristic C. As described above, the light receiving areas of the phase difference pixels ZA and ZB are basically half the light receiving area of ​​the image pixel 203, as shown in FIG. 3. Therefore, adding characteristic A of the phase difference pixel data (ZA) and characteristic B of the phase difference pixel data (ZB) together is expected to produce data approximately equivalent to the image pixel data C. However, in reality, this is not ideal because the light receiving area cannot be precisely halved, there are manufacturing variations, light beam vignetting, and lens aberrations. This results in differences in data values, as shown in FIG. 7(b).

[0043] 7(c) shows a graph D of the ratio (proportion F) of the characteristic A+B of the total phase difference pixel data (ZA+ZB) to the characteristic C of the image pixel data (average of upper and lower Gr). The vertical axis of FIG. 7(c) shows the proportion F of each pixel.

[0044] As shown in FIG. 7(c), the ratio F tends to be almost the same value along the x direction of the image sensor 102. In this example, the ratio F is approximately 0.67, which is a fairly large deviation. In other words, unless this ratio F is taken into consideration as the correction factor F during estimation, a large estimation error will occur. Therefore, in this embodiment, as in the above equations (1) and (2), the estimation accuracy is improved by multiplying the correction factor F in calculating the estimated pixels ZA′ and ZB′ of the phase difference pixels ZA and ZB.

[0045] Strictly speaking, the ratio F (=correction factor F) also varies slightly depending on the pixel position. In the example of FIG. 7(c), the ratio F is slightly higher in the center of the image sensor 102 and slightly lower in the periphery of the image sensor 102. The ratio F also varies when the row in which the phase detection pixels ZA and ZB are arranged changes, and also when the focus position changes. For this reason, it is necessary to know the ratio F (=correction factor F) for each condition and position.

[0046] 8 is a diagram showing an example of a focus position detection result using pixel data strings of the phase difference pixels ZA and ZB after estimation by the phase difference pixel estimation method of this embodiment. The outline of FIG. 8 is similar to that of the comparative example shown in FIG.

[0047] As shown in Fig. 8, when the phase difference pixel data is estimated, the large ranging error that occurred in the comparative example of Fig. 5 is suppressed. The reason for this is that the edge pattern of the subject 601 can be relatively accurately represented by accurately estimating the gaps in the phase difference pixel data string. Therefore, by performing automatic focusing using the pixel data strings of the phase difference pixels ZA and ZB estimated by the phase difference pixel estimation method of this embodiment, it is possible to estimate with relatively high accuracy even subject patterns that are high frequency in the phase difference pixel string direction, that is, "subjects with many high frequency components."

[0048] Fig. 9 is a flowchart of an example of the pre-adjustment process. Each process in the flowchart of Fig. 9 is carried out in an adjustment process immediately after the digital camera 1 is assembled in a factory, for example.

[0049] In step S700, the digital camera 1 is fixed in front of a uniform light subject, and the entire image sensor 102 is set to receive uniform light. At this time, the brightness of the subject, exposure time, and exposure sensitivity are set so that pixel data does not become saturated.

[0050] In step S701, the microprocessor 110 moves the focus lens group 7 and the zoom lens group 6 via the lens drive mechanism 104 to arbitrary initial positions.

[0051] In step S702, the image sensor control circuit 111 acquires data of all phase difference pixels ZA and ZB and image pixel data in the vicinity thereof. Here, as described with reference to FIGS. 3 and 6, "image pixel data in the vicinity of phase difference pixels ZA and ZB" refers to two pixels Gr that are arranged in a row with a row number one smaller than the row of the corresponding phase difference pixels ZA and ZB (the row adjacent above in the example of FIG. 3) and a row with a row number one larger than the row of the corresponding phase difference pixels ZA and ZB (the row adjacent below in the example of FIG. 3), and that are arranged in the same column number. The data acquired in this step is output to the image plane phase difference pixel processing unit 115.

[0052] In step S703, the image plane phase difference pixel processing unit 115 calculates the ratio F (=correction factor F) of each of the phase difference pixels ZA and ZB acquired in step S702. The ratio F on the phase difference pixel ZA can be calculated using the following formula (3). As shown in FIG. 3, there is no data of the phase difference pixel ZB on the phase difference pixel ZA, so the average of the data of the two phase difference pixels ZB arranged on both sides of the phase difference pixel ZA in the same row is used.

[0053]

number

[0054] Similarly, the ratio F on the phase difference pixel ZB can be calculated using the following formula (4).

[0055]

number

[0056] In step S704, the image plane phase difference pixel processing unit 115 stores the ratio F calculated for each of the phase difference pixels ZA and ZB in step S703 in an adjustment value memory in the digital camera 1. At this time, since storing the ratio F for each of the phase difference pixels ZA and ZB requires a large amount of memory, it is also possible to obtain an approximation curve in which the horizontal axis represents the positions of the phase difference pixels ZA and ZB and the vertical axis represents the ratio F, and store only the coefficients of the approximation curve to save memory.

[0057] In step S705, if it is necessary to calculate the ratio F for other lens positions and store it in the adjustment value memory (YES in S705), the process proceeds to step S706, where the microprocessor 110 moves the focus lens group 7 and the zoom lens group 6 to the next lens positions via the lens drive mechanism 104, and then the processing from step S702 onwards is repeated. At this time, if the processing is repeated for all lens positions, an extremely long time and adjustment value memory would be required. Therefore, it is possible to narrow down the processing to only representative lens positions and find the ratio F for the lens positions in between by interpolation.

[0058] On the other hand, if the calculation process of the ratio F has been completed for all of the predetermined lens positions in step S705 (NO in S705), this control flow ends.

[0059] Fig. 10 is a flowchart of an example of imaging plane phase difference detection processing. Each process in the flowchart of Fig. 10 is performed when capturing an image with the digital camera 1. The processing in the flowchart of Fig. 10 corresponds to the focus position detection method according to the embodiment.

[0060] In step S800, the photographing lens group 2 of the digital camera 1 is aimed at a subject for which distance measurement is desired, and the pattern of the target subject is exposed onto the phase difference pixels ZA and ZB of the image sensor .

[0061] In step S801, the image plane phase difference pixel processing unit 115 acquires data of the phase difference pixels ZA and ZB in step S800 and image pixel data in the vicinity thereof from the image sensor 102. At this time, pixel data outside the AF area 602 where distance measurement is not performed is unnecessary, so it is more efficient to limit the data to pixel data within the AF area 602.

[0062] In step S802, the image plane phase difference pixel processing unit 115 (estimation unit) estimates estimated pixels ZA′ and ZB′ between the phase difference pixels ZA and ZB. In this step, as described with reference to FIG. 6, the estimated pixels ZA′ and ZB′ are calculated using the above-described equations (1) and (2).

[0063] In step S803, the image plane phase detection pixel processing unit 115 (focus detection unit) performs phase difference detection using the pixel data sequence (ZA, ZA', ZA, ZA', ...) of the phase detection pixel ZA, whose data volume has doubled after the estimation process in step S802, and the pixel data sequence (ZB, ZB', ZB, ZB', ...) of the phase detection pixel ZB, whose data volume has also doubled, to calculate defocus (amount of defocus). The phase difference detection method is a very common method and will not be explained here, but in brief, the amount of deviation between the two data sequences is found by scanning the correlation value and estimating the subpixels. Then, the amount of deviation between the two data sequences is multiplied by a coefficient determined by the lens optical system to find defocus. When the processing of step S803 is completed, this control flow ends.

[0064] 6 and step S802 in Fig. 10, the image plane phase difference pixel processing unit 115 serving as the point position detection device according to this embodiment performs estimation processing to estimate a plurality of estimated pixels ZA', ZB' arranged between a plurality of phase difference pixels ZA, ZB based on data of two image pixels Gr that are included in rows adjacent to the row in which the plurality of phase difference pixels ZA, ZB are arranged and that are arranged in the same column as the plurality of phase difference pixels ZA, ZB in the pixel array of the image sensor 102. In addition, the image plane phase difference pixel processing unit 115 determines, in advance, the ratio between the average characteristic of pixel data along the two rows including the two image pixels Gr (graph C in Fig. 7(b)) and the sum characteristic of pixel data of the phase difference pixels ZA, ZB (graph A+B in Fig. 7(b)), and uses this ratio as a correction factor F in the estimation processing.

[0065] With this configuration, even in the case of an object with a large number of high-frequency components, such as the object 601 illustrated in FIG. 4, the estimated pixels ZA' and ZB' can be estimated with high accuracy. Furthermore, by using the correction factor F in the estimation process, it becomes possible to perform relatively accurate estimation even if the light receiving area of ​​the phase difference pixel is not exactly half, there is manufacturing variation, there is light beam vignetting, or there is lens aberration. As a result, it is possible to improve the accuracy of focus detection for an object with a large number of high-frequency components.

[0066] The present embodiment has been described above with reference to specific examples. However, the present disclosure is not limited to these specific examples. Design modifications to these specific examples made by a person skilled in the art as appropriate are also included within the scope of the present disclosure as long as they comprise the features of the present disclosure. The elements of the above-described specific examples, as well as their arrangement, conditions, shape, etc., are not limited to those exemplified and can be modified as appropriate. The elements of the above-described specific examples can be combined in various ways as appropriate, as long as no technical contradictions arise. [Explanation of symbols]

[0067] 1. Digital camera (imaging device) 115 Image plane phase difference pixel processing unit (focus position detection device, estimation unit, focus detection unit) ZA Phase detection pixel (first phase detection pixel) ZB Phase detection pixel (second phase detection pixel) ZA' estimated pixel (first estimated phase difference pixel) ZB' estimated pixel (second estimated phase difference pixel) F correction factor [Prior art documents] [Patent documents]

[0068] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-038414

Claims

1. a focus position detection device that detects a focus position based on pixel signal values ​​from an image sensor that includes image pixels that receive an object image through a photographing lens group and photoelectrically convert the image, and a pair of first and second phase difference pixels that receive and photoelectrically convert light beams that pass through a pair of pupil regions of the photographing lens group, an estimation unit that performs an estimation process of estimating a plurality of first estimated phase difference pixels to be arranged between each of the plurality of first phase difference pixels based on data of two image pixels that are included in a row adjacent to a row in which the plurality of sets of the first phase difference pixel and the second phase difference pixel are arranged and that are arranged in the same column as each of the plurality of second phase difference pixels in a pixel array of the imaging element, and estimating a plurality of second estimated phase difference pixels to be arranged between each of the plurality of second phase difference pixels based on data of two image pixels that are included in a row adjacent to a row in which the plurality of sets of the first phase difference pixel and the second phase difference pixel are arranged and that are arranged in the same column as each of the plurality of first phase difference pixels; a focus detection unit that detects a phase difference and calculates a defocus amount to a just-in-focus position using a data string of the first phase difference pixels to which the plurality of first estimated phase difference pixels estimated by the estimation unit are added and a data string of the second phase difference pixels to which the plurality of second estimated phase difference pixels are added; Equipped with the estimation unit grasps, through advance adjustment, a ratio between an average of characteristics of pixel data along two rows including the two image pixels and a sum of characteristics of pixel data of the plurality of first phase difference pixels and characteristics of pixel data of the plurality of second phase difference pixels, and uses the ratio as a correction rate in the estimation process. Focal position detection device.

2. An imaging device equipped with the focal position detection device according to claim 1.

3. A focus position detection method for detecting a focus position based on pixel signal values ​​from an image sensor in which image pixels that receive and photoelectrically convert an object image via a photographing lens group and pairs of first and second phase difference pixels that receive and photoelectrically convert light beams that pass through paired pupil regions of the photographing lens group are arranged, the method comprising: an estimation step of performing an estimation process of estimating a plurality of first estimated phase difference pixels to be arranged between each of the plurality of first phase difference pixels based on data of two image pixels that are included in a row adjacent to a row in which the plurality of sets of the first phase difference pixel and the second phase difference pixel are arranged and that are arranged in the same column as each of the plurality of second phase difference pixels in a pixel array of the imaging element, and estimating a plurality of second estimated phase difference pixels to be arranged between each of the plurality of second phase difference pixels based on data of two image pixels that are included in a row adjacent to the row in which the plurality of sets of the first phase difference pixel and the second phase difference pixel are arranged and that are arranged in the same column as each of the plurality of first phase difference pixels; a focus detection step of detecting a phase difference and calculating a defocus amount to a just-in-focus position using a data string of the first phase difference pixels to which the plurality of first estimated phase difference pixels estimated in the estimation step have been added and a data string of the second phase difference pixels to which the plurality of second estimated phase difference pixels have been added; Including, the estimation step grasps a ratio between an average of characteristics of pixel data along two rows including the two image pixels and a sum of characteristics of pixel data of the plurality of first phase difference pixels and characteristics of pixel data of the plurality of second phase difference pixels by advance adjustment, and uses the ratio as a correction rate in the estimation process; Focus position detection method.

Images