Image sensor

Abstract

To appropriately perform focus detection according to a split-pupil phase difference method, in pixels arranged at a position where an image height of a photoelectric conversion region is high.SOLUTION: In an imaging element, a plurality of pixels including a microlens and a first photoelectric conversion unit and a second photoelectric conversion unit and configured to perform photoelectric conversion of first and second light fluxes passing through first and second regions of a pupil of an optical system, respectively, are arranged in a region into which light from the optical system enters. The pixel includes a first pixel in which the microlens and the first and second photoelectric conversion units have a first positional relationship corresponding to a position of a first exit pupil of the optical system, and a second pixel in which the microlens and the first and second photoelectric conversion units have a second positional relationship corresponding to a position of a second exit pupil of the optical system. A plurality of first pixel rows with a plurality of the first pixels in a first direction are formed in a second direction, and a second pixel row with a plurality of the second pixels in the first direction is arranged in place of some of the first pixel rows.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to an imaging device. [Background technology]

[0002] A technique is known in which pixels each having two photoelectric conversion units are arranged in a photoelectric conversion region and focus detection is performed using a pupil division phase difference method (see Patent Document 1). In such a technique, detection accuracy deteriorates in pixels arranged at positions where the image height of the photoelectric conversion region is high. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2002-314062 Summary of the Invention

[0004] An image sensor according to one aspect of the invention includes a microlens, a first photoelectric conversion unit, and a second photoelectric conversion unit. A plurality of pixels are arranged in a region where light from the optical system is incident, and each pixel photoelectrically converts a first and a second light flux that has passed through a first and a second region of a pupil of the optical system. The pixels include a first pixel in which the microlens and the first and the second photoelectric conversion units have a first positional relationship corresponding to a position of a first exit pupil of the optical system, a second pixel in which the microlens and the first and the second photoelectric conversion units have a second positional relationship corresponding to a position of a second exit pupil of the optical system, and a third pixel in which the microlens and the first and the second photoelectric conversion units have a third positional relationship corresponding to a position of a third exit pupil of the optical system. and a third pixel having a third positional relationship corresponding to the position of the first pixel in the first region, wherein a plurality of first pixel rows, in which the first pixels are lined up in the first direction, are arranged in a second direction in both a first region within the region and a second region within the region that is located closer to the end in the first direction than the first region, a second pixel row, in which the second pixels are lined up in the first direction, is arranged only in the second region of the first region and the second region, a third pixel row, in which the third pixels are lined up in the first direction, is arranged only in the second region of the first region and the second region, and at least one of the plurality of first pixel rows arranged in the second region is arranged between the second pixel row and the third pixel row. An image sensor according to one aspect of the invention includes a microlens, a first photoelectric conversion unit, and a second photoelectric conversion unit. A plurality of pixels are arranged in a region where light from the optical system is incident, the pixels performing photoelectric conversion on first and second light fluxes that have passed through first and second regions of a pupil of the optical system. The pixels include a first pixel in which the microlens and the first and second photoelectric conversion units have a first positional relationship corresponding to a position of a first exit pupil of the optical system, and a second pixel in which the microlens and the first and second photoelectric conversion units have a second positional relationship corresponding to a position of a second exit pupil of the optical system. A first pixel row in which the first pixels are arranged in the first direction is arranged in a plurality of rows in a second direction in both a first region and a second region located closer to an end of the first direction than the first region. A second pixel row in which the second pixels are arranged in the first direction is arranged in a plurality of rows in a second direction only in the second region out of the first region and the second region. The second pixel rows adjacent to each other in the second direction are In the first direction The distances from the center of the regions vary. According to one aspect of the invention, an imaging element includes a microlens, a first photoelectric conversion unit, and a second photoelectric conversion unit, and a plurality of pixels for photoelectrically converting first and second light beams that have passed through first and second regions of a pupil of an optical system are arranged in a region where light from the optical system is incident, and the pixels include a first pixel in which the microlens and the first and second photoelectric conversion units have a first positional relationship corresponding to a position of a first exit pupil of the optical system, a second pixel in which the microlens and the first and second photoelectric conversion units have a second positional relationship corresponding to a position of a second exit pupil of the optical system, and a third pixel in which the microlens and the first and second photoelectric conversion units have a second positional relationship corresponding to a position of a second exit pupil of the optical system. and a third pixel having a third positional relationship corresponding to the position of a third exit pupil of the optical system, wherein a first pixel row, in which the first pixels are arranged in the first direction, is arranged in a plurality of rows in a second direction in both a first region within the region and a second region within the region that is located closer to the end in the first direction than the first region, a second pixel row, in which the second pixels are arranged in the first direction, is arranged only in the second region of the first region and the second region, and a third pixel row, in which the third pixels are arranged in the first direction, is arranged only in the second region of the first region and the second region, and the second pixel row and the third pixel row are arranged non-periodically in the second direction. [Brief description of the drawings]

[0005] [Figure 1] FIG. 2 is a block diagram illustrating a camera according to a first embodiment. [Diagram 2] FIG. 2 is a diagram illustrating a schematic configuration of an imaging element. [Diagram 3] FIG. 2 is a diagram illustrating an example of a pixel arrangement of an imaging element. [Figure 4] FIG. 2 is a circuit diagram illustrating a pixel. [Diagram 5] FIG. 2 is a diagram illustrating an example of a-series and b-series signals. [Figure 6] FIG. 2 is a diagram illustrating a photoelectric conversion region of an image sensor. [Figure 7] 1 is a cross-sectional view of an image sensor for explaining a focus detection light beam by a pupil division phase difference method. [Figure 8] 1 is a cross-sectional view of an image sensor for explaining a focus detection light beam by a pupil division phase difference method. [Figure 9] FIG. 13 is a diagram for explaining an area in which a short distance corresponding ML row and a long distance corresponding ML row are provided in the first modification of the first embodiment. [Figure 10] FIG. 11 is a diagram for explaining an area in which a short distance corresponding ML row and a long distance corresponding ML row are provided in a second modification of the first embodiment. [Figure 11] 13 is an enlarged view of a microlens of a pixel arranged at a position where the image height of an image sensor according to the second embodiment is high. FIG. [Figure 12] 13 is an enlarged view of a microlens of a pixel arranged at a position where the image height of an image sensor according to the second embodiment is high. FIG. [Figure 13] FIG. 13 is an enlarged view of a microlens of a pixel arranged at a position where the image height of an image sensor is high according to a third modified example of the second embodiment. [Figure 14] FIG. 13 is an enlarged view of a microlens of a pixel arranged at a position where the image height of an image sensor is high according to a third modified example of the second embodiment. [Figure 15] FIG. 2 is a diagram illustrating a first aspect of the arrangement of a short distance corresponding ML row and a long distance corresponding ML row. [Figure 16] FIG. 13 is a diagram illustrating a second aspect of the arrangement of the short distance corresponding ML row and the long distance corresponding ML row. [Figure 17] FIG. 13 is a diagram illustrating a third aspect of the arrangement of short distance corresponding ML rows and long distance corresponding ML rows. [Figure 18] FIG. 13 is a diagram illustrating a fourth aspect of the arrangement of short distance corresponding ML rows and long distance corresponding ML rows. [Figure 19] FIG. 13 is a diagram illustrating a fifth aspect of the arrangement of short distance corresponding ML rows and long distance corresponding ML rows. [Figure 20] FIG. 13 is a diagram illustrating a sixth aspect of the arrangement of short distance corresponding ML rows and long distance corresponding ML rows. [Figure 21] FIG. 13 is a diagram illustrating a seventh aspect of the arrangement of short distance corresponding ML rows and long distance corresponding ML rows. [Figure 22]FIG. 13 is a diagram illustrating an eighth aspect of the arrangement of short distance corresponding ML rows and long distance corresponding ML rows. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0006] (First embodiment) FIG. 1 is a block diagram illustrating a digital camera 1 (hereinafter referred to as camera 1) equipped with a focus detection device using a solid-state image sensor 3 (hereinafter referred to as image sensor 3) according to the first embodiment. In the image sensor 3, pixels having multiple photoelectric conversion units for one microlens are arranged across the entire photoelectric conversion area where an image is generated in the image sensor 3. The photoelectric conversion signal read out for each pixel is used for focus detection using the pupil division phase difference method.

[0007] Generally, in the central part of the photoelectric conversion region, light is incident on the photoelectric conversion unit approximately perpendicularly, whereas in the peripheral part located outside the central part of the photoelectric conversion region (the region with a higher image height than the central part), light is incident on the photoelectric conversion unit obliquely. For this reason, pixels suitable for the case where light is incident obliquely are provided in advance in the peripheral part of the photoelectric conversion region.

[0008] However, when the distance from the exit pupil to the image sensor 3 changes due to the characteristics of the photographing lens used in the camera 1, the angle of the light that is obliquely incident on the peripheral portion changes. Therefore, in this embodiment, a plurality of pixels suitable for cases where light is obliquely incident are provided in the peripheral portion of the photoelectric conversion area in correspondence with a plurality of different photographing lenses. This makes it possible to appropriately perform focus detection using the pupil division phase difference method in the peripheral portion of the photoelectric conversion area even when the characteristics of the photographing lenses are different.

[0009] The following detailed description will be given with reference to the drawings. In the first embodiment, the camera 1 is an interchangeable lens type camera such as a single-lens reflex type or a mirrorless type, but the camera does not have to be an interchangeable lens type camera. For example, the camera may be configured as an imaging device such as a lens-integrated camera equipped with a zoom lens, or a camera mounted on a mobile terminal such as a smartphone. Also, the imaging device may be configured as a video camera that captures moving images, a mobile camera, or the like, not limited to still images.

[0010] <Camera configuration> A photographing lens 2 is attached to the camera 1 as an imaging optical system. The photographing lens 2 has a focusing lens and an aperture. The focusing lens and aperture of the photographing lens 2 are controlled by a lens control unit 2a that receives instructions from a microprocessor 9. The photographing lens 2 forms an optical image (subject image) on the imaging surface of an image sensor 3. The photographing lens 2 is also called an imaging optical system.

[0011] The imaging element 3 has a plurality of pixels. As described later, each of the plurality of pixels has two photoelectric conversion units that perform photoelectric conversion on the light that has passed through the photographing lens 2 and entered the imaging element 3 to generate electric charges. The two photoelectric conversion units each output a signal based on the electric charges generated by the photoelectric conversion. The imaging element 3 is controlled by an imaging control unit 4 that receives instructions from a microprocessor 9. Signals output from the plurality of pixels of the imaging element 3 are processed through a signal processing unit 5 and an A / D conversion unit 6, and then temporarily stored in a memory 7. The lens control unit 2a, the imaging control unit 4, the memory 7, the microprocessor 9, a focus calculation unit (focus detection processing unit) 10, a recording unit 11, an image compression unit 12, an image processing unit 13, and the like are connected to the bus 8. The imaging element 3 may have a stacked configuration including some or all of the signal processing unit 5, the A / D conversion unit 6, and the memory 7. The imaging element 3 may have a stacked configuration in which at least one of the signal processing unit 5, the A / D conversion unit 6, and the memory 7 is stacked with a plurality of pixels.

[0012] An operation signal is input from an operation unit 9a such as a release button to the microprocessor 9. The microprocessor 9 sends instructions to each block based on the operation signal from the operation unit 9a, and controls the camera 1.

[0013] The focus calculation unit 10 calculates the focus adjustment state of the photographing lens 2 by a pupil division type phase difference detection method based on signals from pixels of the image sensor 3. The focus calculation unit 10 calculates the focus position of the focus lens for focusing the image of the photographing lens 2 on the imaging surface of the image sensor 3 based on signals based on charges generated by first and second photodiodes S1 and S2 of the pixel 20 described later. Specifically, the focus calculation unit 10 detects the image shift amount of the image formed by a plurality of light beams passing through different regions of the pupil of the photographing lens 2, and calculates the defocus amount based on the detected image shift amount. The defocus amount is the amount of shift between the imaging surface on which the image of the photographing lens 2 is formed and the imaging surface of the image sensor 3. Since the calculation of the defocus amount by the phase difference detection method is publicly known, a detailed description will be omitted. The focus calculation unit 10 calculates the movement amount of the focus lens to the in-focus position based on the calculated defocus amount. The microprocessor 9 instructs the lens control unit 2a to move the focusing lens and sends the calculated movement amount of the focus lens to the lens control unit 2a. This causes the focus adjustment to be performed automatically. The focus calculation unit 10, the microprocessor 9, and the lens control unit 2a operate as a focus adjustment unit.

[0014] The image processing unit 13 performs a predetermined image processing on the signal from the imaging element 3 stored in the memory 7 to generate image data. The image processing unit 13 operates as an image generating unit. The image compression unit 12 compresses the image data after the image processing in a predetermined format. The recording unit 11 records the compressed image data in a predetermined file format on the recording medium 11a and reads out the image data recorded on the recording medium 11a. The recording medium 11a is composed of a memory card that is detachable from the recording unit 11, etc.

[0015] Moreover, the image processing unit 13 generates image data for displaying an image on the display unit 14. The display unit 14 displays an image based on the image data generated by the image processing unit 13. The images displayed by the display unit 14 include reproduced images (still images, moving images) based on image data recorded on the recording medium 11a, and monitor images (also called live view images) acquired by the imaging element 3 at a predetermined interval (for example, 60 fps).

[0016] <Image sensor overview> 2 is a diagram illustrating a schematic configuration of the image sensor 3. The image sensor 3 has a plurality of pixels 20 arranged in a matrix and a peripheral circuit for outputting a signal from each pixel 20. Generally, the smallest unit constituting an image is called a "pixel," but in this embodiment, a configuration for generating a signal of the smallest unit constituting an image is called a "pixel." The photoelectric conversion region 31 indicates a region in which the pixels 20 are arranged in a matrix. In the example of Fig. 2, the photoelectric conversion region 31 is illustrated as having an area of ​​16 pixels, 4 rows horizontally and 4 columns vertically, but the actual number of pixels is much greater than that illustrated in Fig. 2.

[0017] FIG. 3 is a diagram illustrating an example of a pixel arrangement of the imaging element 3. A microlens ML and a color filter (not shown) are provided in the pixel 20. For each pixel 20, one of three color filters having different spectral characteristics, for example, R (red), G (green), and B (blue), is provided. The R color filter mainly transmits light in the red wavelength range. The G color filter mainly transmits light in the green wavelength range. The B color filter mainly transmits light in the blue wavelength range. The G color filter transmits light in a wavelength range shorter than that of the R color filter. The B color filter transmits light in a wavelength range shorter than that of the G color filter. As a result, each pixel 20 has different spectral characteristics depending on the color filter provided.

[0018] In the imaging element 3, pixel rows (also referred to as RG rows) in which pixels 20 having R and G color filters (hereinafter referred to as pixel 20R and pixel 20G, respectively) are alternately arranged, and pixel rows (also referred to as GB rows) in which pixels 20 having G and B color filters (hereinafter referred to as pixel 20G and pixel 20B, respectively) are alternately arranged, are repeatedly arranged in a two-dimensional manner. In the first embodiment, the pixels 20R, 20G, and 20B are arranged according to a Bayer array. In the following description, when the pixel 20 is referred to without the R, G, or B designation, it includes all of the pixel 20R, the pixel 20G, and the pixel 20B.

[0019] Each pixel 20 is provided with two photoelectric conversion units. In general, when two photoelectric conversion units are provided per pixel, there are two cases where the two photoelectric conversion units are arranged in the horizontal direction, i.e., in the row direction (also referred to as horizontal division), and two photoelectric conversion units are arranged in the vertical direction, i.e., in the column direction (also referred to as vertical division). In the first embodiment, horizontally divided pixels 20 are arranged over the entire photoelectric conversion region 31. However, in a predetermined region, vertically divided pixels 20 may be arranged instead of the horizontally divided pixels 20.

[0020] 3, one of the two photoelectric conversion units of each pixel 20 is shown hatched. The hatched photoelectric conversion unit indicates that a first signal (to be described later) is generated based on the charge generated in the photoelectric conversion unit. The non-hatched photoelectric conversion unit indicates that a second signal (to be described later) is generated based on the charge generated in the photoelectric conversion unit. Each pixel 20 performs photoelectric conversion in two photoelectric conversion units in accordance with a control signal from a peripheral circuit, and outputs a signal based on the charge generated in the photoelectric conversion units.

[0021] 2 again. The peripheral circuitry is composed of, for example, a vertical scanning circuit 21, a horizontal scanning circuit 22, control signal lines 23 and 24 connected thereto, first and second vertical signal lines 25a and 25b that receive signals from the pixels 20, a constant current source 26 connected to the first and second vertical signal lines 25a and 25b, a correlated double sampling circuit (CDS circuit) 27, a horizontal signal line 28 that receives a signal output from the CDS circuit 27, an output amplifier 29, and the like. In this embodiment, two vertical signal lines, i.e., the first and second vertical signal lines 25a and 25b, are provided for one pixel column consisting of a plurality of pixels 20 arranged in the column direction.

[0022] The vertical scanning circuit 21 and the horizontal scanning circuit 22 output predetermined control signals in response to instructions from the imaging control unit 4. Each pixel 20 is driven by a control signal output from the vertical scanning circuit 21, and outputs a signal based on an electric charge generated by photoelectric conversion to first and second vertical signal lines 25a and 25b. The signal output from the pixel 20 is subjected to noise removal by a CDS circuit 27, and is output to the outside via a horizontal signal line 28 and an output amplifier 29 in response to a control signal from the horizontal scanning circuit 22.

[0023] <Configuration with two photodiodes PD per pixel> 4 is a circuit diagram illustrating pixels 20 arranged in M ​​columns (arranged vertically) in FIG. 3, i.e., for example, pixel 20G in the Nth row and pixel 20R in the N+1th row. Each pixel 20 has two photodiodes S1 and S2 as a photoelectric conversion unit inside (behind) the microlens ML and the color filter. That is, each pixel 20 has a first photodiode S1 arranged on the left side of the pixel 20 and a second photodiode S2 arranged on the right side of the pixel 20. Therefore, a light beam that has passed through a first region of the pupil of the photographing lens 2 is incident on the first photodiode S1 of each pixel 20, and a light beam that has passed through a second region of the pupil of the photographing lens 2 is incident on the second photodiode S2.

[0024] In this embodiment, for example, the first photodiode S1, the second photodiode S2, and a readout unit that reads out a signal based on the charge generated by the first and second photodiodes S1 and S2 are collectively called a "pixel." An example will be described in which the readout unit includes a transfer transistor, an FD region, an amplification transistor, and a selection transistor, which will be described later, but the scope of the readout unit does not necessarily have to be as shown in this example.

[0025] In each pixel 20, as described above, the first and second photodiodes S1 and S2 are incident with light that has passed through different regions of the pupil of the photographing lens 2, i.e., the first and second regions. The first and second photodiodes S1 and S2 each perform photoelectric conversion on the incident light to generate electric charges. The electric charges generated in the first photodiode S1 are transferred to a first FD (floating diffusion) region FD1 via a first transfer transistor Tx-1. The electric charges generated in the second photodiode S2 are transferred to a second FD region FD2 via a second transfer transistor Tx-2.

[0026] The first FD region FD1 accumulates the received charge and converts it into a voltage. A signal according to the potential of the first FD region FD1 is amplified by the first amplification transistor AMP1. The first FD region FD1 and the first amplification transistor AMP1 operate as a first signal generation unit. Then, the generated signal is read out via the first vertical signal line 25a (output unit) as a signal of a row selected by the first selection transistor SEL1 that selects a row. The first reset transistor RES1 operates as a reset unit that resets the potential of the first FD region FD1. In this embodiment, the signal generated based on the charge generated in the first photodiode S1 is referred to as the first signal.

[0027] Similarly, the second FD region FD2 accumulates the received charge and converts it into a voltage. A signal according to the potential of the second FD region FD2 is amplified by the second amplification transistor AMP2. The second FD region FD2 and the second amplification transistor AMP2 operate as a second signal generation unit. Then, the generated signal is read out via the second vertical signal line 25b (output unit) as a signal of a row selected by the second selection transistor SEL2 that selects a row. The second reset transistor RES2 operates as a reset unit that resets the potential of the second FD region FD2. In this embodiment, the signal generated based on the charge generated in the second photodiode S2 is referred to as the second signal.

[0028] Although an example has been given in which the first signal based on the charges generated in the first photodiode S1 is read out via the first vertical signal line 25a and the second signal based on the charges generated in the second photodiode S2 is read out via the second vertical signal line 25b, a configuration in which the first signal based on the charges generated in the first photodiode S1 and the second signal based on the charges generated in the second photodiode S2 are read out in a time-division manner via a single vertical signal line, i.e., the first vertical signal line 25a or the second vertical signal line 25b, may be used by using a common FD region between the first photodiode S1 and the second photodiode S2.

[0029] <Control signal> In the pixel 20, a first transfer transistor Tx-1 that transfers the charge generated in the first photodiode S1 is turned on by a first control signal φTx1, and a second transfer transistor Tx-2 that transfers the charge generated in the second photodiode S2 is turned on by a second control signal φTx2. The first selection transistor SEL1 for row selection, which outputs a first signal to the first vertical signal line 25a, is turned on by a first control signal φSEL1. The first reset transistor RES1, which resets the potential of the first FD region FD1, is turned on by a first control signal φRES1. In addition, the second selection transistor SEL2 for row selection, which outputs the second signal to the second vertical signal line 25b, is turned on by the second control signal φSEL2. The second reset transistor RES2, which resets the potential of the second FD region FD2, is turned on by the second control signal φRES2.

[0030] <Focus adjustment> The camera 1 of this embodiment generates a pair of focus detection signals used for focus detection (detection of the in-focus position) based on the first signal and the second signal of the pixel 20G read from, for example, a pixel row (GB row) in which the pixels 20G and 20B are alternately arranged among the pixel rows corresponding to the focus area. The focus area is an area in which the focus calculation unit 10 detects the amount of image shift as phase difference information, and is also called a focus detection area, a ranging point, or an autofocus (AF) point.

[0031] For example, when an operation signal indicating that the release button has been half-pressed is input from the operation unit 9a, the microprocessor 9 instructs the imaging control unit 4 to perform imaging for focus adjustment. In imaging for focus adjustment, a control signal for focus adjustment is supplied from the vertical scanning circuit 21 and the horizontal scanning circuit 22 to a pixel row to be read out from the imaging element 3, thereby performing readout for focus adjustment. Readout for focus adjustment refers to supplying a first control signal φTx1, etc. to each pixel 20G in the pixel row to read out a first signal based on the charge generated in the first photodiode S1 from the pixel 20G, and supplying a second control signal φTx2, etc. to each pixel 20G in the pixel row to read out a second signal based on the charge generated in the second photodiode S2 from the pixel 20G.

[0032] By the readout for focus adjustment, the first signal and the second signal read out from each pixel 20G are stored in the memory 7. The multiple first signals A1, A2, ..., An (referred to as a-series signals) stored in the memory 7 and the multiple second signals B1, B2, ..., Bn (referred to as b-series signals) stored in the memory 7 represent the intensity distribution of an image formed by multiple light beams that have passed through different regions of the pupil of the photographing lens 2.

[0033] 5 is a diagram illustrating an a-series signal consisting of a plurality of first signals and a b-series signal consisting of a plurality of second signals. In FIG. 5, n a-series signals are represented by hatched circles. Also, n b-series signals are represented by white circles. The a-series signals and b-series signals from the pixel 20G are read out from every other column in FIG. 3. The vertical dashed lines in FIG. 5 correspond to pixel columns. The focus calculation unit 10 calculates the image shift amount of multiple images by performing image shift detection calculation processing (correlation calculation processing, phase difference detection processing) based on the a-series signals and the b-series signals, and calculates the defocus amount by multiplying the image shift amount by a predetermined conversion coefficient.

[0034] Next, the microprocessor 9 judges whether the defocus amount calculated by the focus calculation unit 10 is within a tolerance. If the defocus amount exceeds the tolerance, the microprocessor 9 judges that the focus is not achieved, and sends a lens drive instruction to the lens control unit 2a. The lens control unit 2a moves the focusing lens to a position (focus position) that brings the defocus amount within the tolerance. On the other hand, if the defocus amount is within the tolerance, the microprocessor 9 judges that the focus is achieved, and does not send a lens drive instruction.

[0035] In the above-described camera 1, the image data used for focus adjustment is generated based on the first signal and the second signal of the pixel 20G read from a pixel row (GB row) in which the pixels 20G and 20B are alternately arranged. The image data used for focus adjustment is not limited to the first signal and the second signal of the pixel 20G, and may be generated based on the first signal and the second signal of the pixel 20B. Moreover, the image data used for focus adjustment may be generated based on the first signal and the second signal of the pixel 20G or the pixel 20R read from a pixel row (RG row) in which the pixels 20R and 20G are alternately arranged.

[0036] <Image data generation> The camera 1 of this embodiment generates image data relating to a subject image based on a first signal and a second signal read from each pixel 20 of the photoelectric conversion region 31 (FIG. 2). When an operation signal indicating that the release button has been fully pressed is input from the operation unit 9a, the microprocessor 9 instructs the imaging control unit 4 to capture an image for recording. In capturing an image for recording, image control signals are supplied from the vertical scanning circuit 21 and the horizontal scanning circuit 22 to each pixel row of the image sensor 3 to perform image reading. The image reading refers to supplying a first control signal φTx1, etc. to each pixel 20 in the pixel row to read out a first signal based on the charge generated in the first photodiode S1 from the pixel 20, and supplying a second control signal φTx2, etc. to each pixel 20 in the pixel row to read out a second signal based on the charge generated in the second photodiode S2 from the pixel 20.

[0037] By reading out the signal for the image, the first signal and the second signal read out from each pixel 20 are stored in the memory 7. The image processing unit 13 adds up the first signal and the second signal stored in the memory 7 for each pixel 20 to generate a signal for the image, and further performs gradation processing, color interpolation processing, etc. to generate image data relating to the subject image.

[0038] <Scaling> In this embodiment, the pixels 20 are arranged at equal intervals in the photoelectric conversion region 31. That is, the intervals in the row and column directions of the first and second photodiodes S1 and S2 of the pixels 20 are equal across the entire photoelectric conversion region 31. By arranging them at equal intervals, the wiring positions and wiring densities around the first and second photodiodes S1 and S2 can be made equal in each pixel 20, so that variations in the electrical characteristics (e.g., parasitic capacitance, etc.) of each pixel 20 can be suppressed.

[0039] On the other hand, in this embodiment, the center position of the microlens ML that collects light on the pixel 20 is shifted from the center position of the pixel 20 depending on the position of the pixel 20 arranged in the photoelectric conversion region 31. The amount of shift is increased as the image height of the position where the pixel 20 is arranged is higher (as it is farther away from the center of the photoelectric conversion region 31). In this example, the center positions of the first and second photodiodes S1 and S2 are set to the center positions of the pixel 20. Gradually changing the amount of shift between the center positions of the microlens ML and the pixel 20 in the photoelectric conversion region 31 is called scaling. The reason why the amount of shift between the center positions of the pixel 20 and the microlens ML is changed depending on the position of the pixel 20 is to allow light that has passed through different regions of the pupil of the photographing lens 2, i.e., the first and second regions of the pupil, to be appropriately incident on the first and second photodiodes S1 and S2. A more detailed explanation will be given with reference to FIG. 6.

[0040] FIG. 6 is a diagram for explaining the photoelectric conversion region 31 of the image sensor 3. As described above, the pixels 20 are arranged at equal intervals throughout the photoelectric conversion region 31. Then, the center positions of the microlenses ML are shifted from the center positions of the pixels 20 by scaling suitable for light incident through the photographing lens 2 having standard characteristics (for example, focal length f=50 mm in 35 mm format). In the example of FIG. 6, the direction in which the microlenses ML are shifted is the row direction (horizontal direction), which is the direction toward the center of the photoelectric conversion region 31. In this embodiment, a pixel row in which the microlenses ML are shifted to conform to the standard characteristics is called a middle distance corresponding ML row. By configuring the pixel rows throughout the photoelectric conversion region 31 as the middle distance corresponding ML row, in all the pixels 20, the light that has passed through the first and second regions of the pupil of the photographing lens 2 having standard characteristics is appropriately incident on the first and second photodiodes S1 and S2.

[0041] In this embodiment, even when a photographing lens 2 having characteristics different from the standard characteristics is used, a pixel row in which microlenses ML are arranged with different scaling from the middle distance corresponding ML row is provided so that light passing through the first and second regions of the pupil of the photographing lens 2 is appropriately incident on the first and second photodiodes S1 and S2. Specifically, instead of the middle distance corresponding ML row, a short distance corresponding ML row and a long distance corresponding ML row are provided in, for example, a 30% region on the left and right ends of the photoelectric conversion region 31, i.e., a part of the region with a high image height. In this embodiment, a pixel row in which microlenses ML are arranged by shifting them to match the characteristics of a wide-angle lens having a focal length shorter than the focal length f=50 mm is called a short distance corresponding ML row. Also, a pixel row in which microlenses ML are arranged by shifting them to match the characteristics of a telephoto lens having a focal length longer than the focal length f=50 mm is called a long distance corresponding ML row.

[0042] The lower part of FIG. 6 shows an enlarged view of the microlens ML of the pixel 20 arranged in the peripheral part of the photoelectric conversion region 31, for example, in the region 31b where the image height is high. In this embodiment, a pixel row (GB row) in which the pixels 20G and the pixels 20B are alternately arranged is used for focus detection, so a part of the GB row is replaced with a short distance corresponding ML row and a long distance corresponding ML row. In the enlarged view, the upper two rows and the lower two rows sandwiching the short distance corresponding ML row are medium distance corresponding ML rows. In the medium distance corresponding ML row, the distance between the line CS passing through the center of the pixel 20 and the line CL (middle) passing through the center of the microlens ML is G. That is, the shift amount of the center position of the microlens ML from the center position of the pixel 20 is G. In FIG. 6, the pixel 20G is shown in white, the pixel 20B is shown in diagonal lines, and the pixel 20R is shown in hatching.

[0043] On one hand, in the short-distance corresponding ML row, the distance between the line CS passing through the center of the pixel 20 and the line CL(short) passing through the center of the microlens ML is Gs. That is, the amount of displacement of the center position of the microlens ML with respect to the center position of the pixel 20 is Gs. The displacement amount Gs in the short-distance corresponding ML row is larger than the displacement amount G in the medium-distance corresponding ML row so that G < Gs holds. The reason will be described with reference to FIGS. 7 and 8.

[0044] <Central portion of the image sensor> FIG. 7 is a cross-sectional view of the medium-distance corresponding ML row of the image sensor 3 for explaining the focus detection light beam by the pupil division phase difference method. In FIG. 7, the case where light is incident substantially perpendicularly on the pixel 20 in the central portion of the photoelectric conversion region 31 in FIG. 6, that is, the region 31a where the image height is low, is shown. In FIG. 7, five pixels 20Ga, 20Bb, 20Gc, 20Bd, and 20Ge among the pixels 20 in the medium-distance corresponding ML row are exemplified. The pixels 20Ga to 20Ge each have microlenses MLa to MLe, a first photodiode S1a, a second photodiode S2a, a first photodiode S1b, a second photodiode S2b, a first photodiode S1c, a second photodiode S2c, a first photodiode S1d, a second photodiode S2d, a first photodiode S1e, and a second photodiode S2e.

[0045] In the central portion of the photoelectric conversion region 31, light is incident substantially perpendicularly on the pixel 20. For this reason, the center position of the microlens ML is arranged to coincide with the center positions of the first and second photodiodes S1 and S2. That is, it is not necessary to displace the center position of the microlens ML with respect to the center position of the pixel 20. Also, in the central portion of the photoelectric conversion region 31 where light is incident substantially perpendicularly on the pixel 20, it is not necessary to provide the short-distance corresponding ML row and the long-distance corresponding ML row. According to FIG. 7, the line CS passing through the center of the pixel 20 and the line CL(medium) passing through the center of the microlens ML coincide.

[0046] The planar shapes of the first photodiodes S1a, S1b, S1c, S1d, and S1e are projected onto an area 91 common to all of the first photodiodes S1a, S1b, S1c, S1d, and S1e on a ranging pupil plane 90 that is separated from the microlenses MLa to MLe by the ranging pupil distance d.

[0047] Furthermore, the planar shapes of the second photodiodes S2a, S2b, S2c, S2d, and S2e are projected onto an area 92 common to all of the second photodiodes S2a, S2b, S2c, S2d, and S2e on a ranging pupil plane 90 spaced apart from the microlenses MLa to MLe by the ranging pupil distance d. The areas 91 and 92 are symmetrical with respect to a line Ax passing through the optical axis. In this case, light passing through the first and second areas 91 and 92 on the ranging pupil plane 90 is appropriately incident on the first and second photodiodes S1 and S2 of the pixel 20.

[0048] A pair of areas 91, 92 on the ranging pupil plane 90 onto which the above-mentioned planar shapes are respectively projected are referred to as ranging pupils. For ease of explanation, the ranging pupil plane 90 is assumed to be located at substantially the same position as the exit pupil plane of the optical system of the photographing lens 2. In other words, the ranging pupil distance d from the microlenses MLa-MLe to the ranging pupil plane 90 is assumed to be substantially equal to the exit pupil distance from the microlenses MLa-MLe to the exit pupil plane of the optical system of the photographing lens 2. For this reason, light that has passed through the first and second areas 91, 92 on the ranging pupil plane 90 corresponds to light that has passed through the first and second areas of the pupil of the photographing lens 2.

[0049] Taking pixels 20Gc and 20Bd among pixels 20Ga to 20Ge as examples, pixel 20Gc receives a focus detection light beam 81 that passes through a ranging pupil 91 and a microlens MLc via a first photodiode S1c, and receives a focus detection light beam 82 that passes through a ranging pupil 92 and a microlens MLc via a second photodiode S2c.

[0050] In addition, the pixel 20Bd receives the focus detection light beam 71 that passes through the ranging pupil 91 and the microlens MLd by the first photodiode S1d, and receives the focus detection light beam 72 that passes through the ranging pupil 92 and the microlens MLd by the second photodiode S2d.

[0051] The same is true for the pixels 20Ga, 20Bb, and 20Ge other than the pixels 20Gc and 20Bd in Fig. 7. As described above, in the central part of the photoelectric conversion region 31, the focus detection light beams that have passed through the ranging pupils 91 and 92, which correspond to the light that has passed through the first and second regions of the pupil of the photographing lens 2, are appropriately incident on the first and second photodiodes S1 and S2, without shifting the central position of the microlens ML with respect to the central position of the pixel 20.

[0052] As a result, in the medium distance corresponding ML row, the first photodiode S1 in the multiple pixels 20 outputs a first signal corresponding to the intensity of an image formed by the focus detection light beam corresponding to the ranging pupil 91. Also, the second photodiode S2 in the multiple pixels 20 outputs a second signal corresponding to the intensity of an image formed by the focus detection light beam corresponding to the ranging pupil 92. As a result, pupil division for focus detection can be performed with high accuracy.

[0053] <Periphery of the image sensor> FIG. 8 is a cross-sectional view of the ML row corresponding to the middle distance of the image sensor 3 for explaining the focus detection light flux by the pupil division phase difference method. FIG. 8 shows a case where light is obliquely incident on the pixels 20 in the peripheral part of the photoelectric conversion area 31 in FIG. 6, that is, the area 31b located at a position where the image height is higher than that of the central part. FIG. 8 illustrates five pixels 20Gg to 20Gk out of the pixels 20 in the ML row corresponding to the middle distance. The pixels 20Gg to 20Gk each have microlenses MLg to MLk, a first photodiode S1g, a second photodiode S2g, a first photodiode S1h, a second photodiode S2h, a first photodiode S1i, a second photodiode S2i, a first photodiode S1j, a second photodiode S2j, a first photodiode S1k, and a second photodiode S2k.

[0054] In the peripheral portion of the photoelectric conversion region 31, light is obliquely incident on the pixel 20. For this reason, the center position of the microlens ML is shifted from the center positions of the first and second photodiodes S1 and S2. In other words, the amount of shift between the line CS passing through the center of the pixel 20 and the line CL (center) passing through the center of the microlens ML is G.

[0055] According to FIG. 8, the planar shapes of the first photodiodes S1g, S1h, S1i, S1j, and S1k are projected onto an area 91 common to all of the first photodiodes S1g, S1h, S1i, S1j, and S1k on a ranging pupil plane 90 that is separated from the microlenses MLg to MLk by the ranging pupil distance d.

[0056] The planar shapes of the second photodiodes S2g, S2h, S2i, S2j, and S2k are projected onto a region 92 common to all of the second photodiodes S2g, S2h, S2i, S2j, and S2k on a ranging pupil plane 90 spaced apart from the microlenses MLg to MLk by the ranging pupil distance d. The regions 91 and 92 are symmetrical with respect to a line Ax passing through the optical axis. In this case, light passing through the first and second regions 91 and 92 on the ranging pupil plane 90 is appropriately incident on the first and second photodiodes S1 and S2 of the pixel 20.

[0057] 7, a pair of regions 91, 92 on the ranging pupil plane 90 onto which the above-mentioned planar shapes are respectively projected are referred to as ranging pupils. For ease of explanation, the ranging pupil distance d from the microlenses MLg-MLk to the ranging pupil plane 90 is considered to be substantially equal to the exit pupil distance from the microlenses MLg-MLk to the exit pupil plane of the optical system of the photographing lens 2. For this reason, light that has passed through the first and second regions 91, 92 on the ranging pupil plane 90 corresponds to light that has passed through the first and second regions of the pupil of the photographing lens 2.

[0058] Taking pixels 20Bh and 20Gi as examples among pixels 20Gg to 20Gk, pixel 20Bh receives a focus detection light beam 171 that passes through the ranging pupil 91 and microlens MLh via the first photodiode S1h, and receives a focus detection light beam 172 that passes through the ranging pupil 92 and microlens MLh via the second photodiode S2h.

[0059] In addition, the pixel 20Gi receives a focus detection light beam 181 that passes through the ranging pupil 91 and the microlens MLi through the first photodiode S1i, and receives a focus detection light beam 182 that passes through the ranging pupil 92 and the microlens MLi through the second photodiode S2i.

[0060] The same is true for pixels 20Gg, 20Bj, and 20Gk other than pixels 20Bh and 20Gi in Fig. 8. As described above, in the peripheral portion of photoelectric conversion region 31, by shifting the center position of the microlens ML with respect to the center position of the pixel 20 in the middle distance corresponding ML row, the focus detection light flux that has passed through the ranging pupils 91 and 92, which corresponds to the light that has passed through the first and second regions of the pupil of the photographing lens 2, is appropriately incident on the first and second photodiodes S1 and S2.

[0061] As a result, in the medium distance corresponding ML row, the first photodiode S1 in the multiple pixels 20 outputs a first signal corresponding to the intensity of an image formed by the focus detection light beam corresponding to the ranging pupil 91. Also, the second photodiode S2 in the multiple pixels 20 outputs a second signal corresponding to the intensity of an image formed by the focus detection light beam corresponding to the ranging pupil 92. As a result, pupil division for focus detection can be performed with high accuracy.

[0062] In general, the exit pupil distance from the microlenses MLg-MLk to the exit pupil plane of the optical system of the photographing lens 2 varies depending on the characteristics of the interchangeable lens 2. For this reason, when the focusing pupil distance d, which is considered to be substantially equal to the above-mentioned exit pupil distance, is shorter than the example in FIG. 8, light is incident on the pixel 20 at an even more oblique angle than in FIG. 8. Therefore, as the short distance corresponding ML row illustrated in FIG. 6, a pixel row is provided in which the microlenses ML are arranged with a shift amount Gs larger than the shift amount G in the medium distance corresponding ML row. In the short distance corresponding ML row, the distance between the line CS passing through the center of the pixel 20 and the line CL (short) passing through the center of the microlens ML is Gs (G <Gs)である。

[0063] Conversely, when the focusing pupil distance d is longer than the example in FIG. 8, light that is closer to being perpendicular to the pixel 20 than that in FIG. 8 (closer to the example in FIG. 7) is incident on the pixel 20. Therefore, as the long distance corresponding ML row illustrated in FIG. 6, a pixel row is provided in which the microlenses ML are arranged with a shift amount GL that is smaller than the shift amount G in the medium distance corresponding ML row. Although the illustration of the long distance corresponding ML row is omitted, the distance between the line CS passing through the center of the pixel 20 and the line CL (long) passing through the center of the microlens ML is GL (GL <G)である。

[0064] By providing a short-distance corresponding ML line and a long-distance corresponding ML line, when the distance measurement pupil distance d becomes shorter than the example in FIG. 8, in the short-distance corresponding ML line, the first photodiodes S1 in the plurality of pixels 20 each output a first signal corresponding to the intensity of the image formed by the focus detection light beam corresponding to the distance measurement pupil 91. Also, the second photodiodes S2 in the plurality of pixels 20 each output a second signal corresponding to the intensity of the image formed by the focus detection light beam corresponding to the distance measurement pupil 92. Thus, the pupil division for focus detection can be accurately performed.

[0065] Also, when the distance measurement pupil distance d becomes longer, in the long-distance corresponding ML line, the first photodiodes S1 in the plurality of pixels 20 each output a first signal corresponding to the intensity of the image formed by the focus detection light beam corresponding to the distance measurement pupil 91. Also, the second photodiodes S2 in the plurality of pixels 20 each output a second signal corresponding to the intensity of the image formed by the focus detection light beam corresponding to the distance measurement pupil 92. Thus, the pupil division for focus detection can be accurately performed.

[0066] In the above description, an example has been described in which in the region 31b located on the left side of the center of the photoelectric conversion region 31 in FIG. 6, the center position of the microlens ML is shifted to the left from the center positions of the first and second photodiodes S1 and S2. In the region 31d located on the right side of the center of the photoelectric conversion region 31 in FIG. 6, the center position of the microlens ML is shifted to the right from the center positions of the first and second photodiodes S1 and S2. Although not shown, the shift amount G in the medium-distance corresponding ML line, the shift amount Gs in the short-distance corresponding ML line, and the shift amount GL in the long-distance corresponding ML line have the relationship of GL < G < Gs, similar to the above-described example.

[0067] According to the first embodiment described above, the following operational effects can be obtained. (1) The image sensor 3 includes a microlens ML, a first photodiode S1, and a second photodiode S2. A plurality of pixels 20 that photoelectrically convert the first and second light beams that have passed through the first and second regions of the pupil of the photographing lens 2 are arranged in an area where light from the photographing lens 2 is incident. The pixels 20 have a first positional relationship between the microlens ML and the first and second photodiodes S1 and S2 that corresponds to the position of the first exit pupil of the photographing lens 2 (the microlens ML and the first and second photodiodes S1 and S2 are scaled to correspond to the center position of the pixel 20 by scaling suitable for light incident through the photographing lens 2 having standard characteristics). The intermediate distance corresponding ML row includes a first pixel in which the microlens ML and the first and second photodiodes S1 and S2 have a second positional relationship corresponding to the position of the second exit pupil of the photographing lens 2 (the center position of the microlens ML is shifted relative to the center position of the pixel 20 by scaling different from that of the intermediate distance corresponding ML row), and a plurality of intermediate distance corresponding ML rows in which a plurality of first pixels are lined up in a first (horizontal) direction are formed in a second (vertical) direction, and a plurality of second pixels are lined up in the first direction, for example a short distance corresponding ML row is arranged in place of a portion of the intermediate distance corresponding ML row. With this configuration, even when a photographing lens 2 having characteristics different from the standard characteristics is used, light that has passed through different regions of the pupil of the photographing lens 2, i.e., the first and second regions of the pupil, can be appropriately incident on the first and second photodiodes S1 and S2 in the short distance compatible ML row. As a result, pupil division for focus detection can be performed with high accuracy.

[0068] (2) The pixel 20 of the image sensor 3 includes a third pixel in which the microlens ML and the first and second photodiodes S1 and S2 have a third positional relationship corresponding to the position of the third exit pupil of the photographing lens 2 (further, the center position of the microlens ML is shifted from the center position of the pixel 20 by different scaling), and a plurality of third pixels are arranged in the first direction, for example, a long distance corresponding ML row is arranged in place of a part of the medium distance corresponding ML row, and the medium distance corresponding ML row is between the short distance corresponding ML row and the long distance corresponding ML row. Since this configuration is adopted, even when a photographing lens 2 having further different characteristics is used, light that has passed through different regions of the pupil of the photographing lens 2, i.e., the first and second regions of the pupil, can be appropriately incident on the first and second photodiodes S1 and S2 in the long distance corresponding ML row. As a result, pupil division for focus detection can be performed with high accuracy.

[0069] (3) The lengths of the short distance corresponding ML rows and the long distance corresponding ML rows of the image sensor 3 are shorter than the length in the first (horizontal) direction of the area into which light from the photographic lens 2 is incident. With this configuration, it is possible to reduce the number of second pixels and third pixels arranged in place of first pixels, and to leave many first pixels.

[0070] (4) The short distance corresponding ML row and the long distance corresponding ML row of the image sensor 3 are arranged in an area farther away from the center of the area where light from the photographing lens 2 is incident than a predetermined value, in other words, in an area where the image height is high. With this configuration, particularly in an area where light is likely to be incident obliquely on the pixel 20, light that has passed through the first and second areas of the pupil of the photographing lens 2 can be made to appropriately enter the first and second photodiodes S1 and S2 in the short distance corresponding ML row or the long distance corresponding ML row. As a result, pupil division for focus detection can be performed with high accuracy.

[0071] The following modifications are also within the scope of the present invention, and one or more of the modifications may be combined with the above-described embodiment. (Variation 1) In the first embodiment, as illustrated in Fig. 6, the lengths of the short-distance corresponding ML rows and long-distance corresponding ML rows provided in place of some of the middle-distance corresponding ML rows are made common to all pixel rows in the photoelectric conversion region 31. That is, in a part of the left and right end 30% area of ​​the photoelectric conversion region 31, the middle-distance corresponding ML rows are replaced with the short-distance corresponding ML rows and the long-distance corresponding ML rows. In contrast, in the first modification of the first embodiment, the range (horizontal length) of the left and right end areas of the photoelectric conversion region 31 where the short-distance corresponding ML rows and the long-distance corresponding ML rows are provided is changed.

[0072] Fig. 9 is a diagram illustrating the left and right end regions of the photoelectric conversion region 31 in which a short distance corresponding ML row and a long distance corresponding ML row are provided instead of a middle distance corresponding ML row in the first modification of the first embodiment. In Fig. 9, the lengths of the short distance corresponding ML row and the long distance corresponding ML row that replace a part of the middle distance corresponding ML row are made non-uniform in the photoelectric conversion region 31. For example, the middle distance corresponding ML row is replaced with the short distance corresponding ML row and the long distance corresponding ML row in a part of the left and right end 25% region of the photoelectric conversion region 31, or the middle distance corresponding ML row is replaced with the short distance corresponding ML row and the long distance corresponding ML row in a part of the left and right end 35% region of the photoelectric conversion region 31.

[0073] As described above, when generating image data, the image processing unit 13 generates a signal for an image by adding the first signal and the second signal read out from each pixel 20 of the imaging element 3 for each pixel 20. That is, not only the pixels 20 included in the medium distance corresponding ML row, but also the pixels 20 included in the short distance corresponding ML row and the long distance corresponding ML row are used to generate the signal for an image.

[0074] However, the image signal based on the first signal and the second signal photoelectrically converted by the pixel 20 included in the short distance corresponding ML row does not strictly match the image signal based on the first signal and the second signal photoelectrically converted by the pixel 20 included in the medium distance corresponding ML row. Furthermore, the signal for the image based on the first signal and the second signal photoelectrically converted by the pixel 20 included in the long distance corresponding ML row does not strictly match the signal for the image based on the first signal and the second signal photoelectrically converted by the pixel 20 included in the medium distance corresponding ML row.

[0075] For the above reasons, if the lengths of the short-distance corresponding ML rows and long-distance corresponding ML rows that replace the medium-distance corresponding ML rows in the photoelectric conversion region 31 are made the same for all pixel rows in the photoelectric conversion region 31, there is a risk that the boundary where the medium-distance corresponding ML row changes to the short-distance corresponding ML row in the horizontal direction, or the boundary where the medium-distance corresponding ML row changes to the long-distance corresponding ML row in the horizontal direction, will stand out in the image generated based on the first signal and the second signal.

[0076] However, according to variant example 1 of the first embodiment, the lengths of the short-distance corresponding ML rows and the long-distance corresponding ML rows that replace part of the medium-distance corresponding ML rows are made non-uniform, i.e., uneven, within the photoelectric conversion area 31, so that the boundary where the medium-distance corresponding ML row changes to the short-distance corresponding ML row in the horizontal direction, or the boundary where the medium-distance corresponding ML row changes to the long-distance corresponding ML row in the horizontal direction, can be made less noticeable in the image generated based on the first signal and the second signal.

[0077] As described above, according to the first modification of the first embodiment, the following advantageous effects can be obtained. That is, a plurality of short-distance corresponding ML rows are arranged in the imaging element 3, and the arrangement of adjacent short-distance corresponding ML rows in the second (vertical) direction differs in distance from the center of the area where light from the photographing lens 2 is incident. With this configuration, it is possible to make the boundary where the medium-distance corresponding ML row changes to the short-distance corresponding ML row in the first (horizontal) direction and the boundary where the short-distance corresponding ML row changes to the medium-distance corresponding ML row in the first (horizontal) direction less noticeable in the image generated based on the first signal and the second signal.

[0078] Similarly, a plurality of long distance corresponding ML rows are arranged in the imaging element 3, and adjacent long distance corresponding ML rows in the second (vertical) direction are arranged at different distances from the center of the area where light from the photographing lens 2 is incident. With this configuration, it is possible to make the boundary where the medium distance corresponding ML row changes to the long distance corresponding ML row in the first (horizontal) direction and the boundary where the long distance corresponding ML row changes to the medium distance corresponding ML row in the first (horizontal) direction less noticeable in the image generated based on the first signal and the second signal.

[0079] (Variation 2) In a second modification of the first embodiment, the row spacing of the short-distance corresponding ML row and the long-distance corresponding ML row that replace a part of the medium-distance corresponding ML row is made non-uniform in the photoelectric conversion region 31. Fig. 10 is a diagram for explaining a region where the short-distance corresponding ML row and the long-distance corresponding ML row are provided in place of the medium-distance corresponding ML row in a second modification of the first embodiment. In Fig. 10, the spacing between the short-distance corresponding ML row and the long-distance corresponding ML row that replace a part of the medium-distance corresponding ML row, and the spacing between the long-distance corresponding ML row and the short-distance corresponding ML row are made non-uniform in the column direction (vertical direction) of the photoelectric conversion region 31.

[0080] For example, the number of intermediate-distance corresponding ML rows between the short-distance corresponding ML rows and the long-distance corresponding ML rows, and the number of intermediate-distance corresponding ML rows between the long-distance corresponding ML rows and the short-distance corresponding ML rows are each irregularly changed. Note that irregular change means that the number of intermediate-distance corresponding ML rows between the long-distance corresponding ML rows and the short-distance corresponding ML rows is non-periodically changed in the column direction (vertical direction) of the photoelectric conversion region 31. However, in variant example 2, the irregular change also includes changing the number of medium-distance corresponding ML rows between the long-distance corresponding ML rows and the short-distance corresponding ML rows, for example, by preparing 5 to 8 patterns and periodically repeating these 5 to 8 patterns in the column direction (vertical direction) of the photoelectric conversion region 31.

[0081] According to variant example 2 of the first embodiment, with regard to the short-distance corresponding ML rows and long-distance corresponding ML rows that replace some of the medium-distance corresponding ML rows, the number of medium-distance corresponding ML rows between the short-distance corresponding ML rows and the long-distance corresponding ML rows, and the number of medium-distance corresponding ML rows between the long-distance corresponding ML rows and the short-distance corresponding ML rows are irregularly changed in the column direction (vertical direction) of the photoelectric conversion region 31, so that the boundary where the medium-distance corresponding ML rows change to the short-distance corresponding ML rows in the horizontal direction, or the boundary where the medium-distance corresponding ML rows change to the long-distance corresponding ML rows in the horizontal direction, can be made less noticeable in the image generated based on the first signal and the second signal.

[0082] As described above, according to the second modification of the first embodiment, the following advantageous effects can be obtained. That is, the short-distance corresponding ML rows and the long-distance corresponding ML rows of the imaging element 3 are arranged non-periodically in the second (vertical) direction, replacing the middle-distance corresponding ML rows. With this configuration, it is possible to make the boundary where the middle-distance corresponding ML rows change to the short-distance corresponding ML rows in the first (horizontal) direction, or the boundary where the middle-distance corresponding ML rows change to the long-distance corresponding ML rows in the first (horizontal) direction less noticeable in the image generated based on the first signal and the second signal.

[0083] Note that Modification 1 of the first embodiment and Modification 2 of the first embodiment may be combined. That is, the lengths of the short-distance corresponding ML rows and the long-distance corresponding ML rows that replace a part of the medium-distance corresponding ML rows may be made uneven (irregularly changed) in the photoelectric conversion region 31, and the number of medium-distance corresponding ML rows between the short-distance corresponding ML rows and the long-distance corresponding ML rows, and the number of medium-distance corresponding ML rows between the long-distance corresponding ML rows and the short-distance corresponding ML rows may be irregularly changed in the column direction (vertical direction) of the photoelectric conversion region 31.

[0084] Second embodiment The second embodiment differs from the first embodiment in that the short distance corresponding ML rows and the long distance corresponding ML rows which are replaced with parts of the medium distance corresponding ML rows are replaced in sets of three adjacent rows. In the second embodiment, the camera 1 may or may not be an interchangeable lens type, similar to the first embodiment. Further, it may be configured as an imaging device such as a smartphone or a video camera.

[0085] FIG. 11 is an enlarged view of the microlens ML of the pixel 20 disposed in the peripheral portion of the photoelectric conversion region 31 in FIG. 6, for example, in the region 31b at a position with a high image height, in the imaging element 3 according to the second embodiment. Also in the second embodiment, a pixel row (GB row) in which the pixel 20G and the pixel 20B are alternately arranged is used for focus detection, and a part of the GB row is replaced with a short-distance corresponding ML row or a long-distance corresponding ML row.

[0086] In FIG. 11, three sets of short-distance corresponding ML rows are provided. The upper three rows and the lower three rows sandwiching the three short-distance corresponding ML rows are each a medium-distance corresponding ML row. In the medium-distance corresponding ML row, the distance between the line CS passing through the center of the pixel 20 and the line CL (medium) passing through the center of the microlens ML is G. That is, the amount of displacement of the center position of the microlens ML with respect to the center position of the pixel 20 is G. In FIG. 11, the pixel 20G is shown in white, the pixel 20B is shown by diagonal lines, and the pixel 20R is shown by hatching.

[0087] In the short-distance corresponding ML row in FIG. 11, the distance between the line CS passing through the center of the pixel 20 and the line CL (short) passing through the center of the microlens ML is Gs. That is, the amount of displacement of the center position of the microlens ML with respect to the center position of the pixel 20 is Gs. The amount of displacement Gs in the short-distance corresponding ML row is larger than the amount of displacement G in the medium-distance corresponding ML row so that G < Gs holds. The reason is as described in the first embodiment.

[0088] In addition, the pixel row (GB row) in which the pixel 20G and the pixel 20B are alternately arranged, which is arranged in one of the three short distance corresponding ML rows in FIG. 11, is used for focus detection. The two pixel rows adjacent to the boundary between the short distance corresponding ML row and the medium distance corresponding ML row are not used for focus detection. The reason for this is that the shift amount of the microlens ML between the two pixel rows adjacent to the above-mentioned boundary is different, so it cannot be denied that a small distortion occurs in the microlens ML due to the manufacturing reason of the microlens ML. If it is assumed that the microlens ML is distorted, even if the shift amount of the center position of the microlens ML from the center position of the pixel 20 is strictly controlled as described above, the distortion of the microlens ML will reduce the accuracy of pupil division for focus detection. In order to avoid such a risk, the two pixel rows adjacent to the boundary between the short distance corresponding ML row and the medium distance corresponding ML row are not used for focus detection.

[0089] For the same reason, the middle distance corresponding ML rows in Fig. 11 that may be used for focus detection are a pixel row (GB row) in which pixels 20G and pixels 20B are alternately arranged, and a pixel row (RG row) in which pixels 20R and pixels 20G are alternately arranged, which do not contact the boundary with the short distance corresponding ML row. In this case, image data used for focus adjustment is generated based on the first signal and the second signal of pixel 20G, pixel 20B, or pixel 20R in the middle distance corresponding ML row.

[0090] The same is true for the long distance corresponding ML row. Fig. 12 is an enlarged view of the microlens ML of the pixel 20 arranged in the region 31c (Fig. 6) where the image height of the photoelectric conversion region 31 of the image sensor 3 according to the second embodiment is high. In Fig. 12, the long distance corresponding ML rows are provided in a set of three rows. The upper three rows and the lower three rows sandwiching the three long distance corresponding ML rows are medium distance corresponding ML rows. In the medium distance corresponding ML row, the shift amount of the center position of the microlens ML with respect to the center position of the pixel 20 is G. In FIG. 12 as well, the pixel 20G is shown in white, the pixel 20B is shown in diagonal lines, and the pixel 20R is shown in hatching.

[0091] In the long-distance corresponding ML row of FIG. 12, the distance between the line CS passing through the center of the pixel 20 and the line CL(long) passing through the center of the microlens ML is GL. That is, the amount of shift of the center position of the microlens ML with respect to the center position of the pixel 20 is GL. The shift amount GL in the long-distance corresponding ML row is smaller than the shift amount G in the medium-distance corresponding ML row so that GL < G holds. This reason is as described in the first embodiment.

[0092] Also, among the long-distance corresponding ML rows in FIG. 12, the one used for focus detection is a pixel row (GB row) in which the pixels 20G and 20B are alternately arranged, which is arranged in one of the three long-distance corresponding ML rows. The two pixel rows adjacent to the boundary between the long-distance corresponding ML row and the medium-distance corresponding ML row are not used for focus detection. This reason is the same as in the case of the short-distance corresponding ML row described above.

[0093] For the same reason, among the medium-distance corresponding ML rows in FIG. 12, those that can be used for focus detection are a pixel row (GB row) in which the pixels 20G and 20B are alternately arranged and not adjacent to the boundary with the long-distance corresponding ML row, and a pixel row (RG row) in which the pixels 20R and 20G are alternately arranged. In this case, image data used for focus adjustment is generated based on the first signal and the second signal of the pixels 20G, 20B, or 20R in the medium-distance corresponding ML row.

[0094] In the above description, an example in which the center position of the microlens ML is shifted to the left from the center positions of the first and second photodiodes S1 and S2 in the regions 31b and 31c located on the left side of the center of the photoelectric conversion region 31 in FIG. 6 has been described. In the regions 31d and 31e located on the right side of the center of the photoelectric conversion region 31 in FIG. 6, the center position of the microlens ML is shifted to the right from the center positions of the first and second photodiodes S1 and S2. The shift amount G in the medium-distance corresponding ML row, the shift amount Gs in the short-distance corresponding ML row, and the shift amount GL in the long-distance corresponding ML row have the relationship of GL < G < Gs.

[0095] According to the second embodiment described above, in addition to the effects of the first embodiment, the following effects can be obtained. (1) The short distance corresponding ML rows and the long distance corresponding ML rows of the image sensor 3 are arranged in place of the medium distance corresponding ML rows at a predetermined ratio of the total number of rows in the second (vertical) direction. With this configuration, for example, it is possible to arrange a predetermined number of pixel rows in succession in the second (vertical) direction, in which the shift amount of the center position of the microlens ML is the same from the center position of the pixel 20, so that the pixel rows at the boundary where the shift amount changes are not used for focus detection. As a result, pupil division for focus detection can be performed with high accuracy.

[0096] (2) The short distance corresponding ML rows and the long distance corresponding ML rows of the image sensor 3 are each arranged in three rows in the second (vertical) direction. With this configuration, the center pixel row of the three rows is used for focus detection, thereby enabling accurate pupil division for focus detection.

[0097] (3) The short distance corresponding ML rows and long distance corresponding ML rows of the image sensor 3 are arranged periodically in the second (vertical) direction in place of the medium distance corresponding ML rows. With this configuration, the short distance corresponding ML rows and long distance corresponding ML rows can be arranged so as not to be biased toward the upper or lower part of the photoelectric conversion region 31. As a result, pupil division for focus detection can be performed without bias in the photoelectric conversion region 31.

[0098] The following modifications are also within the scope of the present invention, and one or more of the modifications may be combined with the above-described embodiment. (Variation 3) Modification 3 of the second embodiment differs from the second embodiment in that the short distance corresponding ML rows and the long distance corresponding ML rows that are replaced with parts of the medium distance corresponding ML rows are each replaced in sets of four adjacent rows. The camera 1 in the third modification of the second embodiment may or may not be of an interchangeable lens type, as in the second embodiment, and may be configured as an imaging device such as a smartphone or a video camera.

[0099] FIG. 13 is an enlarged view of a microlens ML of a pixel 20 disposed in a region 31b at a position where the image height of the photoelectric conversion region 31 in FIG. 6 is high in the imaging device 3 according to Modification 3 of the second embodiment. In Modification 3 of the second embodiment, both a pixel row (GB row) in which pixels 20G and pixels 20B are alternately arranged and a pixel row (RG row) in which pixels 20R and pixels 20G are alternately arranged are used for focus detection, and a part of the GB row and the RG row is replaced with a short-distance corresponding ML row or a long-distance corresponding ML row.

[0100] In FIG. 13, four short-distance corresponding ML rows are provided as a set. Three upper rows and four lower rows sandwiching the four short-distance corresponding ML rows are each a medium-distance corresponding ML row. In the medium-distance corresponding ML row, the distance between a line CS passing through the center of the pixel 20 and a line CL (medium) passing through the center of the microlens ML is G. That is, the amount of displacement of the center position of the microlens ML with respect to the center position of the pixel 20 is G. In FIG. 13, the pixel 20G is shown in white, the pixel 20B is shown by diagonal lines, and the pixel 20R is shown by hatching.

[0101] In the short-distance corresponding ML row in FIG. 13, the distance between a line CS passing through the center of the pixel 20 and a line CL (short) passing through the center of the microlens ML is Gs. That is, the amount of displacement of the center position of the microlens ML with respect to the center position of the pixel 20 is Gs. The amount of displacement Gs in the short-distance corresponding ML row is larger than the amount of displacement G in the medium-distance corresponding ML row so that G < Gs holds. The reason for this is as described in the first embodiment.

[0102] Also, among the short-distance corresponding ML rows in FIG. 13, those used for focus detection are a pixel row (GB row) in which pixels 20G and pixels 20B are alternately arranged and a pixel row (RG row) in which pixels 20R and pixels 20G are alternately arranged, which are arranged in two of the four short-distance corresponding ML rows. The two pixel rows in contact with the boundary between the short-distance corresponding ML row and the medium-distance corresponding ML row are not used for focus detection. The reason for this is as described in the second embodiment.

[0103] For the same reason, among the medium-distance corresponding ML lines in FIG. 13, the pixel rows (GB rows) in which pixel 20G and pixel 20B are alternately arranged and the pixel rows (RG rows) in which pixel 20R and pixel 20G are alternately arranged, which are not in contact with the boundary of the short-distance corresponding ML lines, can be used for focus detection. In this case, image data used for focus adjustment is generated based on the first signal and the second signal of pixel 20G, pixel 20B, or pixel 20R in the medium-distance corresponding ML lines.

[0104] The same applies to the long-distance corresponding ML lines. FIG. 14 is an enlarged view of the microlens ML of pixel 20 arranged in region 31c (FIG. 6) at a position with a high image height in the photoelectric conversion region 31 of image sensor 3 according to Modification 3 of the second embodiment. In FIG. 14, four sets of long-distance corresponding ML lines are provided. The three upper rows and the four lower rows sandwiching the four long-distance corresponding ML lines are medium-distance corresponding ML lines respectively. In the medium-distance corresponding ML lines, the amount of displacement of the center position of the microlens ML with respect to the center position of pixel 20 is G. In FIG. 14 as well, pixel 20G is shown in white, pixel 20B is shown with diagonal lines, and pixel 20R is shown with hatching.

[0105] In the long-distance corresponding ML lines in FIG. 14, the distance between the line CS passing through the center of pixel 20 and the line CL(long) passing through the center of the microlens ML is GL. That is, the amount of displacement of the center position of the microlens ML with respect to the center position of pixel 20 is GL. The amount of displacement GL in the long-distance corresponding ML lines is smaller than the amount of displacement G in the medium-distance corresponding ML lines so that GL < G holds. The reason is the same as that described in the first embodiment.

[0106] Also, among the long-distance corresponding ML lines in FIG. 14, the pixel rows (GB rows) in which pixel 20G and pixel 20B are alternately arranged and the pixel rows (RG rows) in which pixel 20R and pixel 20G are alternately arranged, which are arranged in two of the four long-distance corresponding ML lines, can be used for focus detection. The two pixel rows in contact with the boundary between the long-distance corresponding ML lines and the medium-distance corresponding ML lines are not used for focus detection. The reason is the same as in the case of the short-distance corresponding ML lines described above.

[0107] For the same reason, among the medium-distance corresponding ML lines in FIG. 14, the pixel rows (GB rows) in which pixels 20G and 20B are alternately arranged and the pixel rows (RG rows) in which pixels 20R and 20G are alternately arranged, which are not in contact with the boundary with the long-distance corresponding ML lines, can be used for focus detection. In this case, image data used for focus adjustment is generated based on the first signal and the second signal of pixels 20G, 20B, or 20R in the medium-distance corresponding ML lines.

[0108] In the above description, an example in which the center position of the microlens ML is shifted to the left from the center positions of the first and second photodiodes S1 and S2 in regions 31b and 31c located to the left of the center of the photoelectric conversion region 31 in FIG. 6 has been described. In regions 31d and 31e located to the right of the center of the photoelectric conversion region 31 in FIG. 6, the center position of the microlens ML is shifted to the right from the center positions of the first and second photodiodes S1 and S2. The shift amount G in the medium-distance corresponding ML lines, the shift amount Gs in the short-distance corresponding ML lines, and the shift amount GL in the long-distance corresponding ML lines have the relationship GL < G < Gs.

[0109] As described above, according to Modification 3 of the second embodiment, the following operational effects can be obtained. That is, the short-distance corresponding ML lines and the long-distance corresponding ML lines of the imaging device 3 are each arranged in four rows in the second (vertical) direction. With such a configuration, by using the two pixel rows in the middle of the four rows for focus detection, pupil division for focus detection can be performed with high accuracy.

[0110] (Third Embodiment) In the third embodiment, the positions of the short-distance corresponding ML lines and the long-distance corresponding ML lines arranged in the photoelectric conversion region 31 will be described. Specifically, variations in the arrangement of the short-distance corresponding ML lines and the long-distance corresponding ML lines that replace a part of the medium-distance corresponding ML lines will be exemplified. Eight modes from (Example 1) to (Example 8) will be exemplified below, but any arrangement may be adopted, and a part of these arrangements may be appropriately changed.

[0111] (Example 1) Fig. 15 is a diagram for explaining a first mode of arrangement of the short-distance compatible ML row and the long-distance compatible ML row. Fig. 15 shows the order in which the medium-distance compatible ML row, the short-distance compatible ML row, and the long-distance compatible ML row are arranged when focusing on the region 31x of the photoelectric conversion region 31 in Fig. 6, for example. (Example 1) corresponds to the above-mentioned first embodiment. For this reason, only the GB row used for focus detection among the medium-distance compatible ML rows is replaced with the short-distance compatible ML row or the long-distance compatible ML row.

[0112] In FIG. 15, among the pixel rows (GB rows) in which the pixels 20G and 20B of the medium-distance corresponding ML row are alternately arranged, the short-distance corresponding ML rows are arranged in rows No. 1, No. 13, No. 25, etc. As a result, the short-distance corresponding ML rows appear repeatedly sandwiched between the 11 medium-distance corresponding ML rows. After repeating such an arrangement for 96 rows, from the 97th row onwards, the long-distance corresponding ML rows are arranged in rows No. 97, No. 109, etc. among the pixel rows (GB rows) in which the pixels 20G and 20B of the medium-distance corresponding ML row are alternately arranged. As a result, the long-distance corresponding ML rows appear repeatedly sandwiched between the 11 medium-distance corresponding ML rows. In this way, according to (Example 1), an arrangement in which short-distance corresponding ML rows appear repeatedly every 11 rows and an arrangement in which long-distance corresponding ML rows appear repeatedly every 11 rows are repeated in a cycle of 96 rows.

[0113] (Example 2) Fig. 16 is a diagram for explaining a second mode of arrangement of the short-distance compatible ML row and the long-distance compatible ML row. Like Fig. 15, Fig. 16 shows the order in which the medium-distance compatible ML row, the short-distance compatible ML row, and the long-distance compatible ML row are arranged when focusing on the region 31x (Fig. 6) of the photoelectric conversion region 31. (Example 2) is an arrangement assuming a case in which the RB row is also used for focus detection in addition to the GB row. For this reason, the GB row and the RB row used for focus detection among the medium-distance compatible ML rows are replaced with the short-distance compatible ML row or the long-distance compatible ML row.

[0114] 16, among pixel rows (GB rows) in which pixels 20G and 20B of the medium distance corresponding ML row are alternately arranged, short distance corresponding ML rows are arranged in rows No. 1, No. 13, No. 25, etc. Also, among pixel rows (RG rows) in which pixels 20R and 20G of the medium distance corresponding ML row are alternately arranged, short distance corresponding ML rows are arranged in rows No. 2, No. 14, No. 26, etc. As a result, two short distance corresponding ML rows adjacent in the column direction (vertical direction) appear repeatedly with ten medium distance corresponding ML rows in between.

[0115] After repeating such an arrangement for 96 rows, from the 97th row onwards, among the pixel rows (GB rows) in which the pixels 20G and 20B of the medium distance corresponding ML row are alternately arranged, the long distance corresponding ML rows are arranged in rows No. 97, 109, etc. Furthermore, among the pixel rows (RG rows) in which the pixels 20R and 20G of the medium distance corresponding ML row are alternately arranged, the long distance corresponding ML rows are arranged in rows No. 98, 110, etc. As a result, two long distance corresponding ML rows adjacent in the column direction (vertical direction) appear repeatedly with 10 medium distance corresponding ML rows in between. Thus, according to (Example 2), an arrangement in which two consecutive short-distance corresponding ML rows appear every 10 rows and an arrangement in which two consecutive long-distance corresponding ML rows appear every 10 rows are repeated in a cycle of 96 rows.

[0116] (Example 3) Fig. 17 is a diagram for explaining a third mode of arrangement of the short-distance compatible ML row and the long-distance compatible ML row. Like Fig. 15, Fig. 17 shows the order in which the medium-distance compatible ML row, the short-distance compatible ML row, and the long-distance compatible ML row are arranged when focusing on the region 31x (Fig. 6) of the photoelectric conversion region 31. (Example 3) is an arrangement assuming a case in which the RB row is also used for focus detection in addition to the GB row. For this reason, the GB row and the RB row used for focus detection among the medium-distance compatible ML rows are replaced with the short-distance compatible ML row or the long-distance compatible ML row.

[0117] 17, among pixel rows (GB rows) in which pixels 20G and 20B of the medium distance corresponding ML row are alternately arranged, short distance corresponding ML rows are arranged in rows No. 1, No. 21, No. 41, .... Also, among pixel rows (RG rows) in which pixels 20R and 20G of the medium distance corresponding ML row are alternately arranged, short distance corresponding ML rows are arranged in rows No. 6, No. 26, No. 46, ....

[0118] Moreover, among the pixel rows (GB rows) in which the pixels 20G and 20B of the medium distance corresponding ML row are alternately arranged, the long distance corresponding ML row is arranged in row No. 11, row No. 31, .... Furthermore, among the pixel rows (RG rows) in which the pixels 20R and 20G of the medium distance corresponding ML row are alternately arranged, the long distance corresponding ML row is arranged in row No. 16, row No. 36, ....

[0119] In this way, according to (Example 3), an arrangement in which short-distance corresponding ML rows appear repeatedly every four rows and an arrangement in which long-distance corresponding ML rows appear repeatedly every four rows are repeated in a cycle of 10 rows.

[0120] (Example 4) Fig. 18 is a diagram for explaining a fourth mode of arrangement of the short-distance compatible ML row and the long-distance compatible ML row. Like Fig. 15, Fig. 18 shows the order in which the medium-distance compatible ML row, the short-distance compatible ML row, and the long-distance compatible ML row are arranged when focusing on the region 31x (Fig. 6) of the photoelectric conversion region 31. (Example 4) is an arrangement assuming a case in which the RB row is also used for focus detection in addition to the GB row. For this reason, the GB row and the RB row used for focus detection among the medium-distance compatible ML rows are replaced with the short-distance compatible ML row or the long-distance compatible ML row.

[0121] 18, among pixel rows (GB rows) in which pixels 20G and 20B of the medium distance corresponding ML row are alternately arranged, short distance corresponding ML rows are arranged in rows No. 1, No. 13, No. 25, No. 37, etc. Also, among pixel rows (RG rows) in which pixels 20R and 20G of the medium distance corresponding ML row are alternately arranged, short distance corresponding ML rows are arranged in rows No. 4, No. 16, No. 28, No. 40, etc.

[0122] Moreover, among the pixel rows (GB rows) in which the pixels 20G and 20B of the medium distance corresponding ML row are alternately arranged, the long distance corresponding ML rows are arranged in rows No. 7, No. 19, No. 31, No. 43, etc. Furthermore, among the pixel rows (RG rows) in which the pixels 20R and 20G of the medium distance corresponding ML row are alternately arranged, the long distance corresponding ML rows are arranged in rows No. 10, No. 22, No. 34, No. 46, etc.

[0123] In this way, according to (Example 4), an arrangement in which short-distance corresponding ML rows appear repeatedly every third row and an arrangement in which long-distance corresponding ML rows appear repeatedly every third row are repeated in a cycle of six rows.

[0124] (Example 5) FIG. 19 is a diagram for explaining a fifth mode of arrangement of short-distance compatible ML rows and long-distance compatible ML rows. Like FIG. 15, FIG. 19 shows the order in which the medium-distance compatible ML rows, short-distance compatible ML rows, and long-distance compatible ML rows are arranged when focusing on region 31x (FIG. 6) of photoelectric conversion region 31. (Example 5) corresponds to the above-mentioned second embodiment. For this reason, the short-distance compatible ML rows are arranged in a set of three rows, for example, as shown in rows Nos. 12 to 14. In addition, the long-distance compatible ML rows are arranged in a set of three rows, for example, as shown in rows Nos. 96 to 98.

[0125] In Fig. 19, three adjacent short-distance ML rows in the column direction (vertical direction) are repeated with nine medium-distance ML rows in between. After repeating this arrangement for 96 rows, from the 97th row onwards, three adjacent long-distance ML rows in the column direction (vertical direction) are repeated with nine medium-distance ML rows in between. Thus, according to (Example 5), an arrangement in which three consecutive short-distance corresponding ML rows appear every nine rows, and an arrangement in which three consecutive long-distance corresponding ML rows appear every nine rows, are repeated in a cycle of 96 rows. In addition, in (Example 5), as described above, neither of the two pixel rows adjacent to the boundary between the short distance corresponding ML row and the medium distance corresponding ML row, nor the two pixel rows adjacent to the boundary between the long distance corresponding ML row and the medium distance corresponding ML row are used for focus detection.

[0126] (Example 6) Fig. 20 is a diagram for explaining a sixth mode of arrangement of short-distance compatible ML rows and long-distance compatible ML rows. Fig. 20, like Fig. 15, shows the order in which the medium-distance compatible ML rows, short-distance compatible ML rows, and long-distance compatible ML rows are arranged when focusing on region 31x (Fig. 6) of photoelectric conversion region 31. (Example 6) corresponds to Modification 3 of the second embodiment described above. For this reason, for example, as shown in row Nos. 12 to 15, the short-distance compatible ML rows are arranged in sets of four. Also, for example, as shown in row Nos. 96 to 99, the long-distance compatible ML rows are arranged in sets of four.

[0127] In Fig. 20, four adjacent short-distance corresponding ML rows in the column direction (vertical direction) are repeatedly arranged with eight medium-distance corresponding ML rows in between. After repeating such an arrangement for 96 rows, from the 97th row onwards, four adjacent long-distance corresponding ML rows in the column direction (vertical direction) are repeatedly arranged with eight medium-distance corresponding ML rows in between. Thus, according to (Example 6), an arrangement in which short-distance corresponding ML rows appear repeatedly in four consecutive rows every eight rows, and an arrangement in which long-distance corresponding ML rows appear repeatedly in four consecutive rows every eight rows, are repeated in a cycle of 96 rows. In addition, in (Example 6), as described above, neither of the two pixel rows adjacent to the boundary between the short distance corresponding ML row and the medium distance corresponding ML row, nor the two pixel rows adjacent to the boundary between the long distance corresponding ML row and the medium distance corresponding ML row are used for focus detection.

[0128] (Example 7) FIG. 21 is a diagram for explaining a seventh mode of arrangement of the short-distance compatible ML row and the long-distance compatible ML row. Like FIG. 15, FIG. 21 shows the order in which the medium-distance compatible ML row, the short-distance compatible ML row, and the long-distance compatible ML row are arranged when focusing on the region 31x (FIG. 6) of the photoelectric conversion region 31. (Example 7) corresponds to the above-mentioned second embodiment. For this reason, the short-distance compatible ML rows are arranged in a set of three rows, for example, as shown in rows Nos. 24 to 26. Also, the long-distance compatible ML rows are arranged in a set of three rows, for example, as shown in rows Nos. 12 to 14.

[0129] In FIG. 21, three short distance corresponding ML rows adjacent in the column direction (vertical direction) and three long distance corresponding ML rows adjacent in the column direction (vertical direction) appear alternately with nine medium distance corresponding ML rows in between. In addition, in (Example 7), as described above, neither of the two pixel rows adjacent to the boundary between the short distance corresponding ML row and the medium distance corresponding ML row, nor the two pixel rows adjacent to the boundary between the long distance corresponding ML row and the medium distance corresponding ML row are used for focus detection.

[0130] (Example 8) Fig. 22 is a diagram for explaining an eighth mode of arrangement of short-distance compatible ML rows and long-distance compatible ML rows. Fig. 22, like Fig. 15, shows the order in which the medium-distance compatible ML rows, short-distance compatible ML rows, and long-distance compatible ML rows are arranged when focusing on region 31x (Fig. 6) of photoelectric conversion region 31. (Example 8) corresponds to Modification 3 of the second embodiment described above. For this reason, for example, as shown in row Nos. 24 to 27, the short-distance compatible ML rows are arranged in sets of four. Also, for example, as shown in row Nos. 12 to 15, the long-distance compatible ML rows are arranged in sets of four.

[0131] In FIG. 22, four short distance corresponding ML rows adjacent in the column direction (vertical direction) and four long distance corresponding ML rows adjacent in the column direction (vertical direction) appear alternately with eight medium distance corresponding ML rows in between. In addition, in (Example 8), as described above, neither of the two pixel rows adjacent to the boundary between the short distance corresponding ML row and the medium distance corresponding ML row, nor the two pixel rows adjacent to the boundary between the long distance corresponding ML row and the medium distance corresponding ML row are used for focus detection.

[0132] According to the third embodiment described above, in addition to the effects of the first embodiment, the following effects can be obtained. That is, the short distance corresponding ML rows and the long distance corresponding ML rows of the image sensor 3 are arranged in place of the medium distance corresponding ML rows at the ratios shown in the above examples out of the total number of rows in the second (vertical) direction. With this configuration, for example, it is possible to prevent pixel rows at the boundary where the shift amount changes from being used for focus detection, and it is possible to arrange the short distance corresponding ML rows and the long distance corresponding ML rows so as not to be biased toward the upper or lower part of the photoelectric conversion region 31. As a result, pupil division for focus detection can be performed accurately without bias in the photoelectric conversion region 31.

[0133] Although various embodiments and modifications have been described above, the present invention is not limited to these. Other aspects that are conceivable within the scope of the technical idea of ​​the present invention are also included within the scope of the present invention. [Explanation of symbols]

[0134] 1. Camera 2...Photographic lens 3. Image sensor 9...Microprocessor 10…Focus calculation section 13...Image processing unit 20, 20G, 20R, 20B…pixels 21...Vertical scanning circuit 22...Horizontal scanning circuit 23, 24...Control signal lines 25…Vertical signal line AMP: Amplification transistor FD…FD area ML...Micro lens S1, S2…Photodiodes SEL...Select transistor Tx-1, Tx-2...Transfer transistor

Claims

1. a plurality of pixels each including a microlens, a first photoelectric conversion unit, and a second photoelectric conversion unit, each performing photoelectric conversion on a first light flux and a second light flux that have passed through a first region and a second region of a pupil of an optical system, are disposed in a region where light from the optical system is incident; the pixels include a first pixel having a first positional relationship between the microlens and the first and second photoelectric conversion units corresponding to a position of a first exit pupil of the optical system, a second pixel having a second positional relationship between the microlens and the first and second photoelectric conversion units corresponding to a position of a second exit pupil of the optical system, and a third pixel having a third positional relationship between the microlens and the first and second photoelectric conversion units corresponding to a position of a third exit pupil of the optical system, a first pixel row in which a plurality of the first pixels are arranged in the first direction is arranged in a second direction in both a first region in the region and a second region located closer to an end in the first direction than the first region; a second pixel row in which a plurality of the second pixels are arranged in the first direction is disposed only in the second region out of the first region and the second region; a third pixel row in which a plurality of the third pixels are arranged in the first direction is disposed only in the second region out of the first region and the second region; an imaging element, wherein at least one pixel row of the plurality of first pixel rows arranged in the second region is arranged between the second pixel row and the third pixel row;

2. 2. The imaging device according to claim 1, an imaging element in which the length of the second pixel row and the length of the third pixel row are shorter than the length of the region in the first direction;

3. 3. The imaging device according to claim 2, an imaging element in which a plurality of the second pixel rows are arranged in the second region, and the second pixel rows adjacent to each other in the second direction are at different distances from the center of the region in the first direction;

4. 4. The imaging device according to claim 3, The imaging element includes a plurality of third pixel rows arranged in the second region, and adjacent third pixel rows in the second direction are at different distances from the center of the region in the first direction.

5. 5. The imaging device according to claim 1, The second pixel row and the third pixel row are arranged at a predetermined ratio of a total number of rows in the second direction of the imaging element.

6. 6. The imaging device according to claim 5, The second pixel row and the third pixel row are periodically arranged in the second direction.

7. 6. The imaging device according to claim 5, The imaging element, in which the second pixel row and the third pixel row are non-periodically arranged in the second direction.

8. 6. The imaging device according to claim 5, The second pixel row and the third pixel row are each an imaging element in which at least three rows are arranged in the second direction.

9. a plurality of pixels each including a microlens, a first photoelectric conversion unit, and a second photoelectric conversion unit, each performing photoelectric conversion on a first light flux and a second light flux that have passed through a first region and a second region of a pupil of an optical system, are disposed in a region where light from the optical system is incident; the pixels include a first pixel having a first positional relationship between the microlens and the first and second photoelectric conversion units corresponding to a position of a first exit pupil of the optical system, and a second pixel having a second positional relationship between the microlens and the first and second photoelectric conversion units corresponding to a position of a second exit pupil of the optical system, a first pixel row in which a plurality of the first pixels are arranged in the first direction is arranged in a second direction in both a first region in the region and a second region located closer to an end in the first direction than the first region; a plurality of second pixel rows, in which the second pixels are arranged in the first direction, are arranged only in the second region out of the first region and the second region; The second pixel rows adjacent to each other in the second direction have different distances from the center of the region in the first direction.

10. 10. The imaging device according to claim 9, the pixels include a third pixel having a third positional relationship in which the microlens and the first and second photoelectric conversion units correspond to a position of a third exit pupil of the optical system; a third pixel row in which a plurality of the third pixels are arranged in the first direction is disposed only in the second region out of the first region and the second region; The third pixel rows adjacent to each other in the second direction are at different distances from the center of the region in the first direction.

11. a plurality of pixels each including a microlens, a first photoelectric conversion unit, and a second photoelectric conversion unit, each performing photoelectric conversion on a first light flux and a second light flux that have passed through a first region and a second region of a pupil of an optical system, are disposed in a region where light from the optical system is incident; the pixels include a first pixel having a first positional relationship between the microlens and the first and second photoelectric conversion units corresponding to a position of a first exit pupil of the optical system, a second pixel having a second positional relationship between the microlens and the first and second photoelectric conversion units corresponding to a position of a second exit pupil of the optical system, and a third pixel having a third positional relationship between the microlens and the first and second photoelectric conversion units corresponding to a position of a third exit pupil of the optical system, a first pixel row in which a plurality of the first pixels are arranged in the first direction is arranged in a second direction in both a first region in the region and a second region located closer to an end in the first direction than the first region; a second pixel row in which a plurality of the second pixels are arranged in the first direction is disposed only in the second region out of the first region and the second region; a third pixel row in which a plurality of the third pixels are arranged in the first direction is disposed only in the second region out of the first region and the second region; The imaging element, in which the second pixel row and the third pixel row are non-periodically arranged in the second direction.

Images