JP6641187B2 - Focus position detector and imaging device - Google Patents

Description

  The present invention relates to a focus position detector and an imaging device for detecting a focus position.

  In autofocus technology that automatically detects and adjusts the focus position in an image pickup device, an image plane phase difference type auto focus that detects a defocus amount by arranging a phase difference detection pixel in an image sensor and outputting the phase difference output thereof. Many techniques are known.

  For example, in an image pickup device, two diagonal pixels are discretely arranged as phase difference detection pixels in a structure of a set of four pixels (GGBR, RGBW, etc.) which are color-separated into a plurality of types, and the defocus amount is obtained using the pixel output. A technique for detecting is disclosed (for example, see Patent Document 1).

  By the way, when calculating the defocus amount using the difference between the pixel values of the paired phase difference detection pixels, if crosstalk occurs from a neighboring pixel to the phase difference detection pixel, the defocus amount is calculated. Calculation accuracy decreases.

  Therefore, in order to improve the calculation accuracy of the defocus amount, the theoretical calculation / measurement experiment is performed in advance and a correction table is provided in the signal processing circuit, so that the amount of crosstalk leaking from neighboring pixels to the phase difference detection pixel is obtained. (See, for example, Patent Document 2).

JP 2010-19911A JP 2014-110619 A

  In the technique of Patent Document 1 described above, since the crosstalk is not reduced, in an image sensor having a large crosstalk amount (for example, an image sensor having a large pixel pitch and a small pixel pitch), the F value of the lens of the image pickup optical system is calculated. The accuracy of calculating the defocus amount may be reduced depending on the brightness, image height, array position of the phase difference detection pixels, and the like of the subject.

  Further, in the technique of Patent Document 2, a technique is proposed in which a correction table obtained by performing theoretical calculations and measurement experiments in advance is provided in a signal processing circuit, and crosstalk is corrected to calculate a defocus amount. However, the correction table must correspond to a number of parameters such as the individual recognition / aperture value of the imaging lens, the image height / color of the subject, and the arrangement position of the phase difference detection pixels. There is a problem. In addition, it is necessary to create a large number of image sensors with changed mask arrangements for an experiment for calculating the correction table, and there is a problem in terms of design efficiency.

  For this reason, a technique for reducing the influence of the crosstalk amount leaking into the phase difference detection pixels, improving the calculation accuracy of the defocus amount, and making the focus position detection more accurate in a highly practical manner is desired. .

  The present invention has been made in view of the above-described problems, and provides an in-focus position detector and an imaging device that reduce the influence of the amount of crosstalk, improve the accuracy of calculating the amount of defocus, and increase the accuracy of in-focus position detection. Is to do.

The in-focus position detector of the present invention includes a pair of two diagonal pixels for detecting a phase difference between the imaging pixels in a pixel portion forming a group of four imaging pixels in a predetermined color filter array. An in-focus position detector for detecting an in-focus position of an imaging lens capable of adjusting a focus position by using an imaging element provided with a pixel group of phase difference detection pixels, wherein the in-focus adjustment target area in the pixel unit is Are based on a predetermined correction process using both the pixel values of the first pixel and the second pixel constituting the set of phase difference detection pixels of the two diagonal pixels. A correction pixel value calculation unit that calculates a correction pixel value corrected so as to suppress the amount of crosstalk, and a first pixel group of the set of phase difference detection pixels of the two diagonal pixels in the focus adjustment target area. For each of the pixel and the second pixel The defocus amount calculation means for calculating a defocus amount by performing a correlation operation using the corrected pixel value, wherein the corrected pixel value calculation means as the predetermined correction process, the first to A difference between a pixel value of a pixel and a correction value obtained by dividing a correction coefficient from a value obtained by adding the pixel value of the first pixel and the pixel value of the second pixel, and a lower limit value of the first pixel is defined as the first pixel value. and the corrected pixel value, characterized that you calculate a value lower limit at zero value by subtracting the correction value from the pixel value of the second pixel as the correction pixel value of the second pixel.

  Further, in the in-focus position detector according to the present invention, the correction pixel value calculating means includes, as the predetermined correction processing, a plurality of correction coefficients having different values that are commonly applied in a focus adjustment target area in the pixel unit. Variably set, and the function of calculating the correction pixel value for each of the first pixel and the second pixel in accordance with each of the plurality of correction coefficients. Calculating a plurality of defocus amounts corresponding to each of the plurality of correction coefficients based on the corrected pixel value for each of the pixel and the second pixel; and a final defocus amount based on the plurality of defocus amounts It has the function of calculating

  Further, the image pickup apparatus of the present invention comprises an in-focus position detector of the present invention, and an auto-focus adjusting means for executing an auto-focus adjustment of the imaging lens based on the defocus amount calculated by the in-focus position detector. And the following.

  According to the present invention, since the influence of the crosstalk amount can be reduced and the calculation accuracy of the defocus amount can be improved, the detection accuracy of the focus position can be improved.

It is a block diagram showing a schematic structure of an imaging device of one embodiment by the present invention. FIG. 2 is a diagram schematically illustrating a pixel arrangement of a pixel unit in an imaging device according to an imaging device of an embodiment of the present invention. FIG. 2 is a diagram partially showing an arrangement of phase difference detection pixels in an image sensor according to an imaging device of an embodiment of the present invention. FIG. 2 is a cross-sectional view schematically illustrating a structure of a phase difference detection pixel and an adjacent pixel in an image sensor according to the imaging apparatus of one embodiment of the present invention. FIGS. 4A and 4B are diagrams illustrating light reception of a pupil-divided light beam of an imaging lens with respect to a phase difference detection pixel in an imaging device according to an imaging apparatus according to an embodiment of the present invention. 7A and 7B are diagrams illustrating an imaging model at the time of focusing on a phase difference detection pixel in an image sensor according to an imaging apparatus according to an embodiment of the present invention. FIGS. 7A and 7B are diagrams illustrating an imaging model of a phase difference detection pixel in an imaging device according to an imaging apparatus according to an embodiment of the present invention when a point light source is ahead of an in-focus position. FIGS. 7A and 7B are diagrams illustrating an imaging model of a phase difference detection pixel in an imaging device according to an imaging apparatus according to an embodiment of the present invention when a point light source is behind a focus position. FIG. 2 is a diagram illustrating crosstalk in an image sensor according to an embodiment of the present invention. FIGS. 4A and 4B are diagrams illustrating an imaging model of a phase difference detection pixel in an imaging device according to an imaging apparatus according to an embodiment of the present invention in consideration of crosstalk during focusing. (A) and (b) each describe an imaging model of a phase difference detection pixel in an image sensor according to an embodiment of the present invention in which a point light source considers crosstalk when the point light source is ahead of an in-focus position. FIG. (A) and (b) each illustrate an imaging model in consideration of crosstalk when a point light source is behind a focus position for a phase difference detection pixel in an image sensor according to an imaging apparatus according to an embodiment of the present invention. FIG. 6 is a flowchart relating to automatic focus adjustment control of a focus position detector including crosstalk amount suppression correction processing of Example 1 in the imaging apparatus of one embodiment of the present invention. 7A and 7B are explanatory diagrams each illustrating a crosstalk amount suppression correction process of Example 1 by a focus position detector in an imaging device according to an embodiment of the present invention. 7A and 7B are explanatory diagrams each illustrating a crosstalk amount suppression correction process of Example 1 by a focus position detector in an imaging device according to an embodiment of the present invention. 6 is a flowchart relating to automatic focus adjustment control of a focus position detector including crosstalk amount suppression correction processing of Example 2 in the imaging apparatus of one embodiment of the present invention. (A), (b) is an explanatory view of the crosstalk amount suppression correction processing of Example 2 by the in-focus position detector in the imaging apparatus according to an embodiment of the present invention. 13 is a flowchart relating to automatic focus adjustment control of a focus position detector including crosstalk amount suppression correction processing of Example 3 in the imaging apparatus of one embodiment of the present invention.

  Hereinafter, a focus position detector 32 and an imaging device 50 according to an embodiment of the present invention will be described with reference to the drawings. In the imaging device 50, the focus position detector 32 is configured as a functional unit that detects a focus position (calculates a defocus amount) using a pixel output of a phase difference detection pixel obtained from the image sensor 1.

(Imaging device)
FIG. 1 is a block diagram showing a schematic configuration of an image pickup apparatus 50 according to an embodiment of the present invention. The image pickup apparatus 50 includes an existing typical image pickup device 1 described later and a focus position detector 32 according to the present invention. Thus, as an example, an example in which the television camera is provided with an autofocus system will be described.

  The imaging device 50 is configured to form an image of a subject on the imaging device 1 by the imaging lens 2, perform image processing on an output of the imaging device 1, and output the image data to the display unit 6 and the recording unit 7. It can be divided into an optical system 51, a signal processing system 52, and an input / output system 53. The imaging optical system 51 forms an image of a subject on the imaging device 1 by the imaging lens 2, and the imaging device 1 outputs a digital amount of pixel signals (unit signals corresponding to pixel values).

  The signal processing system 52 receives an operation instruction from the operation unit 8 of the input / output system 53 through the interface (IF) control unit 4, and the IF control unit 4 transmits a control signal related to the received operation to the imaging lens 2 or the digital signal processing. Output to the control unit 3. Further, the IF control unit 4 has a function of not only transmitting a control signal to the digital signal processing control unit 3 but also obtaining various parameters (for example, a gamma value) relating to image processing and outputting the parameters to the operation unit 8. In particular, the IF control unit 4 acquires lens parameters such as the aperture, focal length, and focus from the imaging lens 2 and outputs them to the digital signal processing control unit 3 and the operation unit 8, and the operation unit of the input / output system 53. In the case where the operation instruction from 8 is an automatic focus adjustment, for example, it has a function of outputting an imaging lens control signal relating to a change in lens parameters to the imaging lens 2.

  The digital signal processing control unit 3 controls various operations according to an operation instruction from the IF control unit 4. In particular, the digital signal processing control unit 3 includes a digital signal processing unit 31, a focus position detector 32, and a drive control unit 33. The digital signal processing unit 31 stores the pixel signals of the RGB image pickup pixels obtained from the image pickup device 1 in the memory 5, reads out the pixel signals from the memory 5, performs predetermined image processing, and performs a display output pixel signal (TV A function of composing a display output pixel signal according to the video format of the John camera and outputting the signal to the display unit 6 or the recording unit 7 of the input / output system 53, and a phase difference detection pixel among pixel signals obtained from the image sensor 1. And a function of outputting a signal (pixel value of a phase difference detection pixel) to the focus position detector 32.

  Further, the drive control unit 33 has a function of supplying an image pickup device drive signal for controlling the image pickup operation of the image pickup device 1 to the image pickup device 1.

  The focus position detector 32 calculates a defocus amount, which is a focus shift amount, based on the phase difference detection pixel signal among the pixel signals obtained from the image sensor 1, determines a focus position, and performs automatic focus adjustment. It has a function of outputting information of a control amount to be performed to the IF control unit 4. Thus, the IF control unit 4 can output the imaging lens control signal to the imaging lens 2 and change lens parameters related to automatic focusing.

  Therefore, since the imaging device 50 can perform transmission and reception of electrical signal communication between the imaging lens 2 and the IF control unit 4, the imaging optical system 51 and the signal processing system 52 are connected to the aperture, the focal length, Lens parameters such as focus, exposure information of the image sensor 1, and the like can be shared. Various lens parameters of the imaging lens 2 can be changed by a command (imaging lens control signal) from the IF control unit 4.

  In the example of the imaging device 50 illustrated in FIG. 1, an example is shown in which the input / output system 53 includes the display unit 6, the recording unit 7, and the operation unit 8. The display unit 6, the recording unit 7, or the operation unit 8 such as a television camera for another may be constituted by another device, and a transmission device for communicating with these devices may be provided in the input / output system 53.

(Imaging device)
First, before describing the focus position detector 32 and the imaging device 50 according to an embodiment of the present invention, a configuration example of the imaging device 1 to be used will be described. FIG. 2 schematically illustrates a pixel arrangement of the pixel unit 11 in the image sensor 1 according to the imaging device 50 of the embodiment of the present invention. FIG. 3 partially shows the arrangement of the phase difference detection pixels 113 in the image sensor 1. As the imaging device 1, an existing device can be used. Typically, as shown in FIG. 2, a red (R) color filter (CFR) and a green (G) color filter (CFG) are used. And a color filter (CFB) for blue (B) color is arranged in a set of four pixels in a Bayer array (GGBR) to form the pixel portion 11.

  The image sensor 1 according to the present invention has another color filter array such as an RGBW array in which one of the G color filters is replaced with a white (W) color filter in a filter array of one set of four pixels. It can also be configured. In the present specification, a pixel on which such a color filter is arranged is referred to as an “imaging pixel”.

  By the way, as shown in FIG. 2, in the pixel section 11 of the image sensor 1, a set of two diagonal pixels (shown in FIG. (A pixel, B pixel) phase difference detection pixel 113 is provided, and this phase difference detection pixel 113 functions as an image plane phase difference detection element for detecting a focus. The phase difference detection pixel 113 may be provided with an ND filter that limits only the light amount by a predetermined amount without affecting the color development, but an example in which a color filter is not provided to improve the SN ratio is shown. ing. Then, in the pixel section 11, a set of phase difference detection pixels 113 of two diagonal pixels is discretely arranged at a plurality of positions (see FIG. 1). Further, a light-shielding portion formed by a pair of masks 114 is formed adjacent to the phase difference detection pixel 113 at a diagonal position. Although various shapes are conceivable for the shape of the mask 114, in the present embodiment, the opening shape of the phase difference detection pixel 113 by the mask 114 has a certain suppression effect with an arbitrary shape even if a difference occurs in crosstalk removal performance. Therefore, there is no particular limitation. Therefore, in the present embodiment, the aperture ratio is 50% in the X direction as shown in FIGS. 2 and 3, and the pupil division directions of the A pixel and the B pixel in the phase difference detection pixel 113 are mirror images of each other. The mask 114 is taken as an example.

  Note that the arrangement of the phase difference detection pixels 113 shown in FIGS. 2 and 3 shows an example in which the phase difference detection pixels 113 are arranged at the pixel positions of the R and B color imaging pixels 111 as compared with the set of four pixels. However, it may be arranged at a pixel position of a pair of diagonal G image pickup pixels 111.

  The phase difference detection pixels 113 of a pair of two diagonal pixels are used for calculating a defocus amount as a defocus amount, determining a focus position, and performing automatic focus adjustment. The phase difference detection pixel 113 is configured as a pair of diagonally adjacent two pixels A and B pixels symmetrically adjacent to each other, and the A pixel and the B pixel are pupil division phase difference detection pixels whose phases are shifted from each other in the horizontal and vertical directions. Becomes

  In the specification of the present application, the pupil division direction (horizontal direction in FIGS. 2 and 3) of the A pixel and the B pixel is the X direction, and the direction orthogonal to the X direction on the pixel unit 11 (the vertical direction in FIGS. 2 and 3). ) Is defined as a Y direction, and a direction perpendicular to the XY plane (a direction in front of the planes shown in FIGS. 2 and 3) is defined as a Z direction.

  FIG. 4 is a cross-sectional view of the imaging pixel 111 and the phase difference detection pixel 113 (A pixel and B pixel) of the B and G color rows, taken along the line x-x ′ shown in FIG. The imaging pixels 111 can receive light beams from all pupil regions of the imaging lens 2 through the microlenses 116 by the photodiodes 115 formed on the substrate, passing through the color filters. On the other hand, the opening of the phase difference detection pixel 113 in the pupil division direction (the horizontal direction shown in FIG. 4) is limited by the metal wiring 117 of the image sensor 1 and a mask (metal mask) 114 using a metal electrode or the like. Only light beams from a part of the pupil region of the imaging lens 2 can be received.

  FIGS. 5A and 5B are diagrams illustrating light reception of the pupil-divided light flux of the imaging lens 2 for the phase difference detection pixel 113 in the imaging device 1 according to the imaging device 50 of the present embodiment, respectively. 3 shows the relationship between the cross section of the phase difference detection pixel 113 (A pixel, B pixel) and the pupil of the imaging lens 2. Here, the micro lens 116 is designed so that the pupil of the imaging lens 2 and the photodiode 115 are substantially conjugate. As shown in FIG. 5A, a light beam passing through the left pupil region of the imaging lens 2 in the X direction can reach the photodiode 115 in the B pixel, but is shielded from light by the mask 114 in the A pixel. Conversely, as shown in FIG. 5B, a light beam passing through the right pupil region of the imaging lens 2 can reach the photodiode 115 in the A pixel, but is shielded from light by the mask 114 in the B pixel. As described above, the light beam incident on the phase difference detection pixel 113 can be controlled in association with the pupil position of the imaging lens 2 by the shape of the mask 114.

  By the way, since the mask 114 for forming the phase difference detection pixel 113 and the pupil of the imaging lens 2 are substantially conjugate, the light beam incident on the phase difference detection pixel 113 forms the phase difference detection pixel 113 on the imaging lens 2. It can be considered in the same manner as attaching the light shielding plate S, which is substantially conjugate to the mask 114, to the imaging lens 2. Therefore, in FIGS. 6 to 8, a substantially common light-shielding plate S is provided on the imaging lens 2 side, and a point light source is arranged in the center axis (optical axis in the image forming model) direction of the imaging lens 2 and the imaging device 1. Regarding the image forming model, the image is shown when focused (FIG. 6), when the point light source is ahead of the in-focus position (FIG. 7), and when the point light source is behind the in-focus position (FIG. 8).

  In each of FIGS. 6 to 8, for convenience of explanation, the upper part (a) of the drawing shows the A pixel (phase difference detection pixel A) of the phase difference detection pixel 113, and the lower part (b) shows the phase difference detection pixel. FIG. 3 is an explanatory diagram related to an imaging model of a B pixel (a phase difference detection pixel B) of a pixel 113. FIG. The upper part (a) of the figure shows “(a-1) an imaging model of the phase difference detection pixel A” indicating a physical model of the phase difference detection pixel A, and “(a) indicating an image formed on the image sensor 1 thereof. -2) Image formed by phase difference detection pixel A "and" (a-3) α of phase difference detection pixel A showing an example of pixel values between α-β lines near the central axis of the image sensor 1 The pixel values between -β lines are shown in correspondence with each other. Similarly, also in the lower part (b) of the figure, “(b-1) an imaging model of the phase difference detection pixel B” indicating a physical model of the phase difference detection pixel B, and “an image formed on the image sensor 1” (B-2) Image formed by phase difference detection pixel B "and" (b-3) Phase difference detection pixel B showing an example of pixel values between α-β lines near the central axis of the image sensor 1 Are shown in correspondence with each other.

  At the time of focusing, as shown in FIGS. 6A and 6B, the image formed by each of the phase difference detection pixels A and B adjacent to each other becomes a perfect circle with the same center axis. Accordingly, the pixel values of the phase difference detection pixels A and B are the same.

  On the other hand, in a blurred state when the point light source is ahead of the in-focus position, as shown in FIGS. 7A and 7B, the image formed by the adjacent phase difference detection pixels A and B is on the X-axis ( In the direction of the pupil, the mirror image becomes a semicircle whose diameter is enlarged with respect to the central axis, and the center position of the degree of the spread of the pixel values between the α-β lines is in the X-axis (pupil division) direction as compared with the focusing state shown in FIG. The phase difference detection pixel A is separated from the center by Da and the phase difference detection pixel B is separated by Da (that is, −Da) in the opposite direction. Accordingly, the pixel values of the phase difference detection pixels A and B are also different.

  In the blurred state when the point light source is located behind the in-focus position, as shown in FIGS. 8A and 8B, the image formed by the adjacent phase difference detection pixels A and B is on the X-axis ( In the direction of pupil division), a semicircle having a relationship opposite to that of FIG. 7 in which the mirror image is enlarged with respect to the center axis in the direction is obtained. In comparison, the phase difference detection pixel A is separated by −Da ′ and the phase difference detection pixel B is separated by Da ′ from the center of the X axis (pupil division) direction. In response to this, the pixel values of the phase difference detection pixels A and B are different from each other in a relationship opposite to that of FIG.

  In this way, the image shift amounts (Da shown in FIG. 7 and Da ′ shown in FIG. 7) of the A and B pixels (phase difference detect pixels A and B) of the phase difference detection pixel 113 are calculated, and the shift amounts are calculated. , The defocus amount can be calculated in the manner of triangulation. Also, by grasping whether the B pixel is in the plus direction or the minus direction of the X-axis direction with reference to the A pixel, it is also possible to detect whether the subject is ahead or behind the in-focus position. Yes (the position of A pixel may be grasped based on B pixel). Here, in this example, an example is shown in which pupil division is performed in the X-axis direction. However, even when pupil division is performed in the Y-axis direction or the diagonal direction of the XY plane, the defocus amount can be calculated similarly. Can be.

(Calculation of defocus amount)
Various methods are conceivable as a method of calculating the defocus amount using the phase difference detection pixel 113. One example of the conventional technique is a method of calculating the difference between a pair of images as shown in Expression (1). There is one in which an absolute value is added (for example, see Patent Document 1).

In the equation (1), A _ij and B _ij represent two sets of two diagonal pixels that are discretely arranged in the pixel unit 11 in which the plurality of imaging pixels 111 are arranged in a grid. Represents the i-th pixel value of the j-th column of the detection pixels A and B), where ni is the number of pixels used for the correlation operation, nj is the number of a pair of images to be subjected to the correlation operation in the column direction, and ni and nj are The distance is appropriately set in accordance with the length of the field of view for measuring (field of view for measuring the distance).

  The calculation of the defocus amount based on the expression (1) is to obtain the parallax k at which the correlation operation value COR (k) becomes minimum as the focus position. At this time, it is general to set a threshold value for determining whether or not to perform automatic focus adjustment based on the signal level of the phase difference detection pixel 113 in consideration of the SN ratio of the pixel value. In addition, the number of pixels (ni, nj) used for the correlation calculation is determined based on the detection position of the pixel unit 11, the image height, the lens performance of the imaging lens 2 (lens parameters), and a reference value predetermined based on the signal level thereof. Is determined as the target of the correlation calculation.

  By the way, the example of calculating the defocus amount based on the equation (1) has been described. However, any other method of calculating the defocus amount can detect the phase difference as long as the automatic focusing by the image plane phase difference method is performed. The accuracy (accuracy) of the signal components of the pixel values of the A and B pixels (the phase difference detection pixels A and B) of the pixel 113 is important. For example, if random noise is added to the phase difference detection pixels A and B, the S / N ratio is calculated by increasing the number of ni and nj in Expression (1) and calculating for a plurality of matrices, depending on the subject. And the accuracy of calculating the defocus amount can be improved.

  However, when crosstalk from the neighboring pixels occurs in the phase difference detection pixel 113, the calculation accuracy of the defocus amount decreases. That is, since the crosstalk from the pixels near the phase difference detection pixel 113 is not random, even if the number of ni and nj in the equation (1) is increased, the effect cannot be reduced. Therefore, a separate process for correcting the crosstalk amount is required.

  Here, referring to FIG. 9, regarding the cause of the occurrence of crosstalk in the phase difference detection pixel 113, a G color filter (CFG) constituting a pixel adjacent to the B pixel (phase difference detection pixel B) of the phase difference detection pixel 113. The crosstalk from the imaging pixel 111) will be described as an example. The first cause of the crosstalk is light incident on the adjacent pixel of the phase difference detection pixel B, such as the light ray L1 in the drawing. The light beam L1 generates photoelectrically converted electrons in a deep portion or outside of the photodiode 115, which is an adjacent pixel of the phase difference detection pixel B, and the electrons are diffused so that the photoelectric conversion unit of the phase difference detection pixel B (for example, Leaks into the photodiode 115). The second cause is that the light ray L2 is reflected by the metal wiring 117 of the pixel portion or the like and leaks into the photoelectric conversion portion (photodiode 115) of the phase difference detection pixel B. The crosstalk amount of the light beam L1 changes according to the wavelength of the light beam, and the longer the wavelength, the more easily the light leaks. The greater the angle between the incident light angles of the light beams L1 and L2 with respect to the optical axis of the microlens 116, the greater the effect. Further, the incident light angle varies depending on the aperture value of the imaging lens 2 and the arrangement position of the phase difference detection pixels 113 in the pixel section 11, and is thus affected by the light rays L1 and L2. From FIG. 9, ideally, the light beam on the right side of the imaging lens 2 in the figure should not enter the phase difference detection pixel B, but the pixel value of the phase difference detection pixel B becomes Will also have components. Crosstalk also occurs in the phase difference detection pixel B with respect to the left ray component from the adjacent pixel, but is not shown in the drawing.

  Next, how such crosstalk affects the phase difference detection pixel 113 will be described with reference to FIGS. FIGS. 10 to 12 correspond to FIGS. 6 to 8 described above in which a substantially common light-shielding plate S is provided on the imaging lens 2 side, and the central axes of the imaging lens 2 and the imaging device 1 (in the image forming model, The figure shows a case in which crosstalk is considered for an imaging model in which a point light source is arranged in the (optical axis) direction. When focused (FIG. 10), when the point light source is ahead of the focused position (FIG. 11), respectively, Indicates a position behind the in-focus position (FIG. 12).

  In each of FIGS. 10 to 12, corresponding to FIGS. 6 to 8 described above, the upper part (a) of the drawing shows the A pixel (phase difference detection pixel A) of the phase difference detection pixel 113 and the lower part of the drawing. FIG. 3B is an explanatory diagram related to an imaging model of a B pixel (phase difference detection pixel B) of the phase difference detection pixel 113 in consideration of crosstalk.

  Note that the imaging models shown in FIGS. 10 to 12 show a case where the phase difference detection pixel 113 has leakage due to crosstalk, and the difference from FIGS. 6 to 8 is that there is leakage due to crosstalk. The light shielding plate S in front of the imaging lens 2 in the drawing is not 100% light-shielded, but behaves like a neutral density filter. The transmittance distribution of the light shielding plate S differs depending on the pixel size and the arrangement position of the phase difference detection pixels 113. Here, for the sake of simplicity, the transmittance distribution is assumed to be constant, and only the crosstalk component on the light blocking side is generated, and the crosstalk component of the signal component from the incident light side is ignored. However, in the focus position detector 32 according to the present invention, it is assumed that the transmittance distribution of the light-shielding plate S differs depending on the pixel size and the arrangement position of the phase difference detection pixels 113. The crosstalk amount suppression correction process 3 appropriately suppresses the crosstalk amount even when the crosstalk components on the light shielding side and the incident light side are taken into account.

  First, as shown in FIGS. 10 (a) and 10 (b), at the time of focusing in consideration of crosstalk, it is the same as that at the time of focusing of FIG. 6, and each of the phase difference detection pixels A and B adjacent to each other. Are formed in a perfect circle with the same center axis, and accordingly, the pixel values of the phase difference detection pixels A and B are the same.

  However, in the case where the point light source is in the blurred state in front of the in-focus position, if the light shielding by the light shielding plate S is 100%, the image formed by the phase difference detection pixels A and B will be respectively shown in FIGS. However, considering crosstalk as shown in FIGS. 11A and 11B, the ratio of the transmittance of the light shielding plate S to the semicircle of the signal component in the formed image is considered. The crosstalk component attenuated by the above is imaged as a symmetrical semicircle. For example, as a result of applying Equation (1) between α-β lines in the example of FIG. 7, the parallax k of the phase difference detection pixels A and B should be obtained as k = 2 × Da. Even if Equation (1) is applied under the same condition, the parallax k of the phase difference detection pixels A and B cannot be obtained as k = 2 × Da due to the influence of the crosstalk component.

  Similarly, when the point light source is in the blurred state behind the in-focus position, the cross-talk is taken into consideration as shown in FIGS. The crosstalk component reduced in the ratio of the transmittance of the light shielding plate S to the semicircle of the signal component is imaged as a symmetrical semicircle. For this reason, the disparity k to be obtained by applying the equation (1) in the example of FIG. 7 becomes a different value of the disparity in the example shown in FIG.

  When crosstalk occurs from a neighboring pixel in the phase difference detection pixel 113 based on such a principle, the calculation accuracy of the defocus amount decreases. This crosstalk amount differs depending on the aperture value of the imaging lens 2, the arrangement position of the phase difference detection pixels 113, the image height and color of the subject, and cannot be corrected by a simple fixed correction value. For this reason, the technique of Non-Patent Document 2 requires a large number of correction tables corresponding to a number of parameters such as the individual recognition / aperture value of the imaging lens, the image height / color of the subject, and the arrangement position of the phase difference detection pixels. Although a proposal is made, an enormous number of correction tables are required, which is not practical.

(Focus position detector)
Therefore, in the imaging device 50 of the present embodiment shown in FIG. 1, the crosstalk amount suppression correction processing of each example described below is executed by the focus position detector 32, and then the defocus amount is calculated and calculated. Perform focus position adjustment control.

  More specifically, the focus position detector 32 includes A and B pixels (phase difference detection pixels A, B) constituting a set of two diagonal pixels in the focus adjustment target area in the pixel unit 11. Based on the crosstalk amount suppression correction processing of each embodiment using both the pixel values of B), corrected pixel values corrected so as to suppress the crosstalk amount correlated with each of these pixel values are calculated. I do. Then, the focus position detector 32 uses the corrected pixel value for each of the phase difference detection pixels A and B for the pixel group of the pair of phase difference detection pixels 113 in the diagonal two pixels in the focus adjustment target area. The defocus amount is calculated by executing the correlation calculation processing such as the equation (1).

  First, as a crosstalk amount suppression correction process according to the first embodiment, an example of a process when it is assumed that approximately the same amount of crosstalk components occur in adjacent phase difference detection pixels A and B will be described. Subsequently, as the crosstalk amount suppression correction processing of the second and third embodiments, a processing example when it is assumed that different amounts of crosstalk components occur in the adjacent phase difference detection pixels A and B will be described.

(Example 1)
FIG. 13 is a flowchart relating to the automatic focus adjustment control of the focus position detector 32 including the crosstalk amount suppression correction processing of Example 1 in the imaging device 50 of the present embodiment.

  The processing related to the automatic focus adjustment control of the focus position detector 32 shown in FIG. 13 can be continuously performed at predetermined intervals. Since the pixel signal of the image sensor 1 is stored in the memory 5 each time an image is captured, the in-focus position detector 32 is transmitted from the memory 5 via the digital signal processor 31 to a plurality of locations of the pixel unit 11 of the image sensor 1. Then, a phase difference detection pixel signal (pixel value) of each phase difference detection pixel 113 in the focus adjustment target area arranged in is read (step S1).

  Subsequently, the focus position detector 32 measures the signal level of the read-out phase difference detection pixel signal (Step S2), and determines whether or not the signal level is equal to or larger than a predetermined threshold value in consideration of the SN ratio of the signal. (Step S3). That is, when the pixel value of the phase difference detection pixel 113 is low, the calculation accuracy of the defocus amount is significantly reduced due to the influence of noise, and the focus position of the imaging lens 2 may be adjusted in an incorrect direction. For this reason, a threshold value is provided for the pixel value of the phase difference detection pixel 113 in the focus adjustment target area in the pixel portion, and it is determined in advance whether or not to perform the focus position adjustment. This threshold is set in advance according to the imaging device 1, the method of calculating the defocus amount, and the like.

  Subsequently, when the signal level of the read-out phase difference detection pixel signal is less than the predetermined threshold value, the focus position detector 32 determines that the focus cannot be detected (Step S3: No), and the focus position detection for the current focus position detection is performed. The process ends. On the other hand, when the signal level of the read-out phase difference detection pixel signal is equal to or higher than the predetermined threshold, the focus position detector 32 determines that the focus can be detected (Step S3: Yes), and proceeds to Step S4.

The focus position detector 32 determines the number of pixels (ni, nj) for the correlation operation with respect to the phase difference detection pixel signal for which the focus can be detected (step S4), and the focus adjustment target in the pixel unit 11 as described below. The crosstalk amount suppression correction processing of the first embodiment is executed in the region using Expressions (2) and (3) described later (Step S5), whereby all pairs in the focus adjustment target region in the pixel unit 11 are obtained. With respect to the pixel group of the phase difference detection pixels 113 of a set of two corner pixels, the corrected pixel values D _Aij ′ and D _Bij ′ of the phase difference detection pixels 113 in which the amount of crosstalk is suppressed are obtained. Subsequently, the in-focus position detector 32 calculates the defocus amount by executing a correlation calculation process based on the defocus amount calculation formula such as Expression (1) using the corrected pixel values D _Aij ′ and D _Bij ′. Then, the control unit 4 instructs the IF control unit 4 to execute the automatic focus adjustment of the imaging lens 2 (step S7). Thereby, the IF control unit 4 outputs an imaging lens control signal for moving the position of the imaging lens 2 to the imaging lens 2 and adjusts the focus position. Note that, depending on the method of calculating the defocus amount, a form in which the expression for the suppression of the crosstalk amount and the formula for calculating the defocus amount are integrated and the collective calculation may be performed.

2 and 3, the pixels A and B (phase difference detection pixels A and B) of the phase difference detection pixel 113 are close to each other in a diagonal direction as a pair of two diagonal pixels, and are in the X-axis direction i. _Assuming that the pixel values of the J-th and j-th phase difference detection pixels A and B in the Y-axis direction are D _Aij and D _Bij , respectively, the pixel value D _Aij is _composed of a signal component S _Aij and a crosstalk component N _Aij , and the pixel value D _{Aij Bij} is composed of a signal component S _Bij and a crosstalk component N _Bij .

For example, pixel values between α-β lines of the phase difference detection pixels A and B will be described with reference to FIGS. 14A and 14B. In FIGS. 14A and 14B, the crosstalk component N _Aij , the signal component S _Aij , and the pixel value D _Aij are respectively “N _A ”, “S _A ”, and “D _A ”, and the crosstalk component N _Bij , the signal component S _Bij , and the pixel value D _Bij are respectively. _{"N B", "S B"} is illustrated as _{"D B",} the pixel value _{D Aij,} each actual waveform of _{D Bij S Aij +} N _Aij, the S _{Bij + N} Bij.

For this reason, when the correlation operation value COR (k) is obtained by applying the pixel values D _Aij and D _Bij to A _ij and B _ij in the equation (1), the imaging model shown in FIGS. As described above, the calculation accuracy of the defocus amount decreases.

Therefore, crosstalk amount suppression correction process in Example 1 according to the present invention, the signal component _{S Aij} and the pixel value _{D Aij} consisting crosstalk components _{N Aij,} pixel values consisting of the signal component _{S Bij} and crosstalk component _{N Bij} The correction pixel values D _Aij ′, D _Bij ′ of the phase difference detection pixels A, B are obtained by utilizing the correlation of the proximity with D _Bij in the diagonal direction. Then, when calculating the defocus amount, the corrected pixel values D _Aij ′ and D _Bij ′ are applied to A _ij and B _ij in the equation (1), respectively, to obtain the correlation operation value COR (k), and to obtain the minimum k Determine the value of.

That is, in the crosstalk amount suppression correction processing according to the first embodiment, a value _obtained by subtracting the pixel value D _Bij of the phase difference detection pixel B from the pixel value D _Aij of the phase difference detection pixel A and limiting the lower limit with a zero value is used as the phase difference detection pixel. A corrected pixel value D _Aij ′ of A, and a value _obtained by subtracting the pixel value D _Aij of the phase difference detection pixel A from the pixel value D _Bij of the phase difference detection pixel B and limiting the lower limit with a zero value to the correction pixel of the phase difference detection pixel B It is calculated as the value D _Bij ′.

More specifically, in the present embodiment, _assuming that the crosstalk components N _Aij and N _Bij of the phase difference detection pixels A and B are approximately equivalent, as shown in FIGS. The pixel values D _Aij , D _Bij of the phase difference detection pixels A, B are corrected using Expressions (2) and (3), respectively, to obtain corrected pixel values D _Aij ′, D _Bij ′. In FIG. 15 (a), (b) , the crosstalk component _{N Aij,} the signal component _{S Aij,} and respectively for the pixel value _{D Aij "N A",} and _{"S A", "D A"} cross-talk component N _bij, the signal component _{S bij,} and respectively for the pixel value _{D bij "N B",} illustrated as _{"S B", "D B".} The crosstalk components N _Aij and N _Bij may have different values, respectively, and correspond to a case where the transmittance of the light shielding plate S in the imaging models of FIGS. 10 to 12 is not constant.

Thus, crosstalk amount suppression phase difference detection pixel A, the correction pixel value _{D Aij} of B _{', D Bij'} can be obtained, the correction pixel value _{D Aij ', D Bij'} effectively Figure This can be applied to the imaging models shown in FIGS. When a circuit for reading the pixel values of the phase difference detection pixels A and B is shared, a common noise component generated when passing through the read circuit can be reduced by Expressions (2) and (3). Note that it is not necessary to calculate the corrected pixel values D _Aij ′ and D _Bij ′ for all the pixels i and j (i-th phase difference detection pixels A and B in the X-axis direction and j-th direction in the Y-axis direction). It is sufficient to calculate only the focus adjustment target area in. Then, the defocus amount can be calculated by the calculation method such as Expression (1) using the corrected pixel values D _Aij ′ and D _Bij ′.

Here, regarding the calculation of the correction pixel values D _Aij ′, D _Bij ′ of the first embodiment, the crosstalk components N _Aij , N _Bij of the phase difference detection pixels A, B are considered to be approximately equivalent. . The values of the signal components S _Aij , S _Bij are S _Aij = S _Bij if in focus, and S _Aij ≠ S _Bij if out of focus. However, the crosstalk components N _Aij and N _{Bij tend} to be N _Aij ≒ N _Bij irrespective of whether the _object is focused or not. This is because the phase difference detection pixels A and B of a pair of two diagonal pixels are extremely close to each other, so that the incident light angle with respect to the optical axis of the micro lens 116 which affects the crosstalk can be regarded as approximately the same. This is because all the imaging pixels 111 surrounding the four directions are G color pixels. Also, even when the arrangement of the phase difference detection pixels A and B is not the position of a set of four B pixels and R color pixels as in this example but the G color pixel at a diagonal position, the diagonal position Pixels surrounding the eight directions on one of the phase difference detection pixels 113 are G color pixels × 3, R color pixels × 2, R color pixels × 1, and non-color filter pixels (the other phase difference detection pixels 113) × 1. Since the same relationship is established, the amount of crosstalk affecting each of the pair of phase difference detection pixels A and B on two diagonal pixels can be treated as approximately the same in many cases. Of course, although some subjects there is a possibility that a part of the phase difference detection pixel A, in B the _{N Aij} ≠ _{N Bij, A ij} in equation (1), each correction pixel value _{D Aij} to _{B ij ', D Bij} Is applied to the calculation a plurality of times, and consequently, the calculation accuracy of the correlation operation value COR (k) of the equation (1) is improved as compared with the case where the crosstalk amount suppression correction processing of the first embodiment is not applied. Can be done.

  As described above, according to the focus position detector 32 performing the crosstalk amount suppression correction processing of the first embodiment, it is possible to obtain a pixel value corrected so as to suppress the crosstalk amount in accordance with each image sensor 1. Therefore, the calculation accuracy of the defocus amount can be improved, and the detection accuracy of the focus position can be improved. In particular, it is not necessary to manufacture an image sensor for determining coefficients of a correction table for correcting crosstalk as in the technique of Patent Document 2. According to the in-focus position detector 32 and the imaging apparatus 50 including the same according to the present invention, even when the imaging lens 2 needs to be replaced with the correction table in the technique of Patent Document 2, any processing is performed. It can be handled without changing the method. Further, according to the in-focus position detector 32 according to the present invention, there is no need to hold a large number of correction tables related to the crosstalk amount suppression correction, and the in-image pickup device required by the technique of Patent Document 2 is not required. Can be saved in comparison with the memory and the amount of calculation.

(Example 2)
Next, the crosstalk amount suppression correction processing of Example 2 in the imaging device 50 of the present embodiment will be described. In Equations (2) and (3) according to the above-described first embodiment, the crosstalk components N _A and N _B are considered to be approximately equivalent, respectively, but there is not a small difference between these N _A and N _B. is there. In particular, the configuration in which the A and B pixels of the phase difference detection pixel 113 of one set of two diagonal pixels are in a diagonal array of G color in the imaging element 111 of one set of four pixels unlike the examples shown in FIGS. In the case of the A pixel, the right and left adjacent color filters are R color and the upper and lower adjacent color filters are B color, and the B pixel is the left and right adjacent color filter is B color and the upper and lower adjacent color filter is R color. . In this case, the color filters adjacent to the A and B pixels in the oblique direction are all G colors. For this reason, depending on the color of the subject, the incident angle, and the focus adjustment target area in the pixel unit 11 relating to the calculation of the defocus amount, the corrected pixel value D _{Aij is} obtained by using the equations (2) and (3) as in the first _embodiment. When ', D _Bij ' is calculated, an unnecessary subtraction amount occurs in the signal components S _Aij , S _Bij , and there is a possibility that the SN ratio for the crosstalk components N _Aij , N _Bij may not be sufficiently obtained.

Therefore, in the crosstalk amount suppression correction processing according to the second embodiment, unnecessary subtraction amounts are reduced in the signal components S _Aij and S _Bij in consideration of the case where there is a difference between the crosstalk components N _Aij and N _Bij. In this way, the corrected pixel values D _Aij ″, D _Bij ″ are calculated. FIG. 16 is a flowchart relating to the automatic focus adjustment control of the focus position detector 32 including the crosstalk amount suppression correction processing of Example 2 in the imaging device 50 of the present embodiment. The control flow according to the second embodiment illustrated in FIG. 16 is different in that step S5 illustrated in FIG. 13 relating to the crosstalk amount suppression correction processing is replaced with steps S5a and S5b.

  As shown in FIG. 16, the focus position detector 32 of this example is arranged at a plurality of locations of the pixel unit 11 of the image sensor 1 from the memory 5 via the digital signal processing unit 31, as in the example shown in FIG. The phase difference detection pixel signal (pixel value) of each phase difference detection pixel 113 in the focus adjustment target area is read out (step S1).

  Subsequently, the focus position detector 32 measures the signal level of the read phase difference detection pixel signal in the same manner as in the example shown in FIG. 13 (step S2), and considers the SN ratio of the signal to obtain a predetermined threshold value. It is determined whether or not this is the case (step S3). Subsequently, when the signal level of the read-out phase difference detection pixel signal is less than the predetermined threshold value, the focus position detector 32 determines that the focus cannot be detected (Step S3: No), and the focus position detection for the current focus position detection is performed. The process ends. On the other hand, when the signal level of the read-out phase difference detection pixel signal is equal to or higher than the predetermined threshold, the focus position detector 32 determines that the focus can be detected (Step S3: Yes), and proceeds to Step S4.

The focus position detector 32 determines the number of pixels (ni, nj) of the correlation operation with respect to the phase difference detection pixel signal for which the focus can be detected (step S4). A predetermined correction coefficient x described later is set in the area (step S5a), and the crosstalk amount suppression correction processing of the second embodiment is executed using equations (4) and (5) described later (step S5b). Thus, for the pixel group of the set of phase difference detection pixels 113 of all two diagonal pixels in the focus adjustment target area in the pixel unit 11, the correction pixel value D _Aij ”of the phase difference detection pixel 113 in which the crosstalk amount is suppressed is set. , D _Bij ”are obtained. Subsequently, the in-focus position detector 32 calculates a defocus amount by executing a correlation calculation process based on a defocus amount calculation formula such as Expression (1) using the corrected pixel values D _Aij ″, D _Bij ″. Then, the control unit 4 instructs the IF control unit 4 to execute the automatic focus adjustment of the imaging lens 2 (step S7). Thereby, the IF control unit 4 outputs an imaging lens control signal for moving the position of the imaging lens 2 to the imaging lens 2 and adjusts the focus position. Note that, depending on the method of calculating the defocus amount, a form in which the expression for the suppression of the crosstalk amount and the formula for calculating the defocus amount are integrated and the collective calculation may be performed.

14A based on the pixel values D _A (= S _A + N _A ) and D _B (= S _B + N _B ) of the phase difference detection pixels A and B shown in FIGS. 14A and 14B described above. 17 (a) and 17 (b), the crosstalk component N _Aij , the signal component S _Aij , and the pixel value D _Aij are “N _A ” and “S _A ”, respectively. _a ", and" _{D a} "cross-talk component _{N Bij,} the signal component _{S Bij,} and respectively for the pixel value _{D Bij" N B} "," _{S B} ", as shown as" _{D B} ", pixel value _{D Aij,} actual waveform each _S Aij ₊ N _Aij of _{D Bij,} the S _{Bij + N} Bij.

That is, the amount of crosstalk suppression correction process in Example 2, obtained by adding the pixel value D _Bij of pixel values D _Aij and the phase difference detection pixel B of the phase difference detection pixel A from the pixel value D _Aij of the phase difference detection pixel A a value obtained by lower-limit correction value obtained by dividing the correction coefficient x to the value in the difference to a zero value as the corrected pixel value D _{Aij "of} the phase difference detection pixel a, the correction value from the pixel value D _Bij of the phase difference detection pixel B _Is calculated as a corrected pixel value D _Bij ″ of the phase difference detection pixel B.

The difference from the first embodiment is that by subtracting (D _Aij + D _Bij ) / x from the pixel values D _Aij and D _Bij on the basis of Expressions (4) and (5) to remove negative components, The point is that the corrected pixel values D _Aij ″, D _Bij ″ are obtained. The value of x, by selecting the _{{(x-1) · N} A -N B}, the closer to 0 the value of _{{(x-1) · N} B -N A}, suitably crosstalk It is possible to obtain corrected pixel values D _Aij ″, D _Bij ″ with suppressed amounts, and as a result of an experimental study, it is preferable that the value of x is a fixed value larger than 1 and 3 or less. Note that the value of x is a fixed value in the focus adjustment target area in the pixel unit 11 in one calculation of the defocus amount, and is not changed by the values of i and j.

Depending on the value of x, the values of the corrected pixel values D _Aij ″, D _Bij ″ may be apparently small, but in Expressions (4) and (5), when x = 2, Expressions (2) and ( This is the same as _setting the pixel values D _Aij and D _Bij in 3) to １ ／. Therefore, similarly to the first embodiment, it is possible to suppress the crosstalk amount having a correlation with each other for the pixel values of the phase difference detection pixels A and B, and to have a correlation with the phase difference detection pixels A and B. Assuming that there is a difference in crosstalk, the difference between N _Aij and N _Bij can also be reduced by weighting and subtracting with the value of x. That is, depending on the value of x, the corrected pixel values D _Aij ″, D _Bij ″ are represented by the ratio of each value of the signal components D _Aij , D _Bij to the corresponding value of the crosstalk components N _Aij , N _Bij. The S / N ratio is relatively improved, and as a result, the calculation accuracy of the defocus amount can be improved.

  By the way, in addition to the method of experimentally examining the value of x and determining it as a fixed value, a method of adaptively determining the value of x for each focus adjustment target area in the pixel portion can be adopted. That is, since the incident light angle with respect to the optical axis of the microlens 116 of the i-th phase difference detection pixels A and B in the Y-axis direction with respect to the optical axis of the microlens 116 can be considered to be the same, After the determination, the color component ratio of the subject in the focus adjustment target area in the pixel portion and the color filter arrangement of the imaging pixels 111 adjacent to the phase difference detection pixels A and B can be adaptively determined and determined. Since the color filter array is actually fixed, the value of x can be determined according to the color component ratio of the subject in the focus adjustment target area in the pixel section after the focus adjustment target area in the pixel section is determined. However, even in this case, it is preferable that the value of x is a fixed value greater than 1 and 3 or less.

According to the second embodiment, since all the functions and effects of the first embodiment are included, and the difference between the crosstalk components N _Aij and N _Bij can be reduced, the calculation accuracy of the defocus amount can be improved. Can be improved.

(Example 3)
Next, automatic focus adjustment control including crosstalk amount suppression correction processing of Example 3 in the imaging device 50 of the present embodiment will be described. In the above-described second embodiment, an example has been described in which the value of x in Expressions (4) and (5) is set to one and the defocus amount is calculated in the focus adjustment target area in the pixel portion. On the other hand, in the automatic focus adjustment control including the crosstalk amount suppression correction processing according to the third embodiment, the value of x is changed within one focus adjustment target area in the pixel unit 11 during one automatic focus adjustment control operation. To calculate the corrected pixel values D _Aij ′ and D _Bij ′ to calculate the corresponding defocus amounts for a plurality of times, and calculate the final defocus amounts based on the defocus amounts calculated for the plurality of times. decide.

That is, the crosstalk amount suppression correction processing in the third embodiment variably sets a plurality of correction coefficients x having different values that are commonly applied in the focus adjustment target area in the pixel unit 11, and sets the phase difference detection pixels A and B. The correction pixel values D _Aij ′ and D _Bij ′ for each of them are calculated according to each of the plurality of correction coefficients x. The calculation of the defocus amount in the third embodiment is based on the correction pixel values D _Aij ′ and D _Bij ′ for each of the phase difference detection pixels A and B, and calculates a plurality of defocus amounts corresponding to each of the plurality of correction coefficients x. A defocus amount is calculated, and a final defocus amount is calculated based on the plurality of defocus amounts.

More specifically, in the automatic focus adjustment control including the crosstalk amount suppression correction processing according to the third embodiment, in order to further improve the accuracy in consideration of the case where there is a difference between the crosstalk components N _Aij and N _Bij , the value of x is set. Is changed and the number of times of calculating the defocus amount (y times) is arbitrarily set to a fixed number, and the value of x is set to a fixed value larger than 1 and equal to or smaller than 3 as in the second embodiment. Is preferred. In the third embodiment, the final defocus amount is determined by calculating the defocus amount for y times.

Therefore, in the present embodiment, the correction pixel values D _Aij ″ and D _Bij ″ for y times are calculated. FIG. 18 is a flowchart relating to the automatic focus adjustment control of the focus position detector 32 including the crosstalk amount suppression correction processing of Example 3 in the imaging device 50 of the present embodiment. The control flow according to the third embodiment illustrated in FIG. 18 is different from the control flow relating to the automatic focus adjustment control illustrated in FIG. 16 in that the defocus amount is calculated y times in accordance with the value of the correction coefficient x that is variably set, and the final defocus amount is calculated. The difference is that processing (steps S6a, S6b, S6c) for determining the defocus amount is added.

  As shown in FIG. 18, the in-focus position detector 32 of the present example is provided with a digital signal processing unit 31 from the memory 5 via the digital signal processing unit 31, and a plurality of locations of the pixel unit 11 of the image sensor 1 as in the examples shown in FIGS. Then, a phase difference detection pixel signal (pixel value) of each phase difference detection pixel 113 in the focus adjustment target area arranged in is read (step S1).

  Subsequently, the focus position detector 32 measures the signal level of the read phase difference detection pixel signal in the same manner as in the examples shown in FIGS. 13 and 16 (step S2), and considers the SN ratio of the signal. It is determined whether it is equal to or more than a predetermined threshold (step S3). Subsequently, when the signal level of the read-out phase difference detection pixel signal is less than the predetermined threshold value, the focus position detector 32 determines that the focus cannot be detected (Step S3: No), and the focus position detection for the current focus position detection is performed. The process ends. On the other hand, when the signal level of the read-out phase difference detection pixel signal is equal to or higher than the predetermined threshold, the focus position detector 32 determines that the focus can be detected (Step S3: Yes), and proceeds to Step S4.

  The focus position detector 32 determines the number of pixels (ni, nj) for the correlation operation with respect to the phase difference detection pixel signal for which focus detection is possible (step S4). It is determined whether or not the value of the defocus amount has already been calculated through steps S6 and S6a to be described later and stored in the memory 5 as the value of the y defocus amount (step S6b). When the value of the defocus amount for y times is stored in the memory 5 (step S6b: Yes), the process proceeds to step S6c, and when the value is less than y times (step S6b: No), the process proceeds to step S5a. .

Subsequently, the focus position detector 32 sets a correction coefficient x that changes to a different value y times in the focus adjustment target area in the pixel unit 11 when calculating the defocus amount to be performed y times each time. Then (step S5a), the crosstalk amount suppression correction processing using the above-described equations (4) and (5) is executed (step S5b), whereby all the pairs in the focus adjustment target area in the pixel unit 11 are executed. With respect to the pixel group of the phase difference detection pixels 113 of a set of two corner pixels, the correction pixel values D _Aij ″, D _Bij ″ of the phase difference detection pixels 113 in which the crosstalk amount is suppressed using the correction coefficient x are obtained.

Subsequently, the focus position detector 32 performs a correlation calculation process based on a defocus amount calculation formula such as Expression (1) using the corrected pixel values D _Aij ″ and D _Bij ″ based on the correction coefficient x. Is calculated (step S6), and the value of the defocus amount is stored in the memory 5 (step S6a).

In the focus adjustment target area in the pixel unit 11, y values of the defocus amount calculated using the correction pixel values D _Aij ″ and D _Bij ″ based on the correction coefficient x of y different values are stored in the memory 5. Are repeated (step S6b), and when the y defocus amount values are stored in the memory 5 (step S6b: Yes), the process proceeds to step S6c.

  If y values of the defocus amounts are stored in the memory 5 in one focus adjustment target area in the pixel unit 11 during one automatic focus adjustment control operation (step S6b: Yes), the focus position is set. The detector 32 determines a final defocus amount using the y defocus amount values (step S6c).

  Subsequently, the focus position detector 32 instructs the IF control unit 4 to execute automatic focusing of the imaging lens 2 based on the final defocus amount (Step S7). Thereby, the IF control unit 4 outputs an imaging lens control signal for moving the position of the imaging lens 2 to the imaging lens 2 and adjusts the focus position. Note that, depending on the method of calculating the defocus amount, a form in which the expression for the suppression of the crosstalk amount and the formula for calculating the defocus amount are integrated and the collective calculation may be performed.

As a specific value of the correction coefficient x that changes to a different value y times within one focus adjustment target area in the pixel unit 11 during one auto focus control operation, for example, the value of x is 1.7, The correction pixel values D _Aij ″ and D _Bij ″ are obtained over three times of 2.0 and 2.3, and three corresponding defocus amounts are calculated, and the three defocus amounts are calculated. To determine the final defocus amount.

  As a method of determining the final defocus amount using the y defocus amount values, for example, a method of averaging these y values, or a distance of moving the imaging lens 2 among the y values Or a method of selecting and determining the defocus amount corresponding to the one that minimizes the color components in the focus adjustment target area in the pixel portion and the color filter array of the imaging pixel 111 adjacent to the phase difference detection pixel 113. A method may be adopted in which a weighted average weighted in consideration of the relationship is used.

According to the third embodiment, all the functions and effects of the first and second embodiments are included, the difference between the crosstalk components N _Aij and N _Bij is reduced, and the calculation accuracy of the defocus amount is _improved . It can be further improved.

  As described above, the present invention has been described with the example of the specific embodiment. However, the present invention is not limited to the above-described example, and can be variously modified without departing from the technical idea thereof. For example, in the imaging device 1, an example is shown in which the mask 114 is configured to be adapted to the microlens 116 and shield light in accordance with the split pupil direction. However, even without the microlens 116, the mask 114 is moved in the split pupil direction. It can be configured to shield light accordingly.

  According to the present invention, a pixel value corrected so as to suppress the amount of crosstalk can be obtained in accordance with each image sensor, so that it is useful for a camera that performs automatic focus adjustment.

REFERENCE SIGNS LIST 1 imaging element 2 imaging lens 3 digital signal processing control unit 4 IF control unit 5 memory 6 display unit 7 recording unit 8 operation unit 11 pixel unit 31 digital signal processing unit 32 in-focus position detector 33 drive control unit 50 imaging device 51 imaging Optical system 52 Signal processing system 53 Input / output system 111 Imaging pixel 113 Phase difference detection pixel (A pixel, B pixel)
114 Mask 115 Photodiode 116 Microlens 117 Metal wiring

Claims (3)

A pixel group forming a group of four imaging pixels in a predetermined color filter array is provided with a group of two diagonal pixels for detecting a phase difference between the imaging pixels. A focusing position detector that detects a focusing position of an imaging lens whose focal position can be adjusted using the imaging device,
Based on a predetermined correction process using both the pixel values of the first pixel and the second pixel constituting the set of phase difference detection pixels of the diagonal two pixels in the focus adjustment target area in the pixel portion, Correction pixel value calculation means for calculating a correction pixel value corrected to suppress the crosstalk amount having a correlation with each other for each pixel value,
For a pixel group of the set of phase difference detection pixels of the two diagonal pixels in the focus adjustment target area, performing a correlation calculation process using the correction pixel value for each of the first pixel and the second pixel. Defocus amount calculating means for calculating a defocus amount by
The correction pixel value calculation unit may include, as the predetermined correction process, a correction coefficient for a value obtained by adding a pixel value of the first pixel and a pixel value of the second pixel from a pixel value of the first pixel. A value obtained by subtracting the corrected values and limiting the lower limit with a zero value as the corrected pixel value of the first pixel, and subtracting the corrected value from the pixel value of the second pixel and limiting the lower limit with a zero value to the second pixel the focus position detector characterized that you calculated as corrected pixel value. The correction pixel value calculation means variably sets a plurality of the correction coefficients having different values to be commonly applied in a focus adjustment target area in the pixel unit as the predetermined correction processing, and A function of calculating the correction pixel value for each of the two pixels according to each of the plurality of correction coefficients;
The defocus amount calculation means calculates a plurality of defocus amounts corresponding to each of the plurality of correction coefficients based on the corrected pixel values for each of the first pixel and the second pixel, and calculates the plurality of defocus amounts. and having a function of calculating the final defocus amount on the basis of the focus amount, the focus position detector according to claim 1. A focus position detector according to claim 1 or 2 ,
Based on the defocus amount calculated by the in-focus position detector, an automatic focus adjustment unit that performs automatic focus adjustment of the imaging lens;
An imaging device comprising: