JP5896603B2 - Imaging apparatus and image processing apparatus - Google Patents

Description

The present invention relates to an imaging apparatus and an image processing apparatus , and more particularly to an imaging apparatus using an imaging element having focus detection pixels.

  2. Description of the Related Art Conventionally, there is an image pickup apparatus in which some pixels of an image pickup element used in an image pickup apparatus such as a digital still camera or a digital video camera are replaced with pixels for focus detection using a phase difference method (see, for example, Patent Document 1). In this focus detection pixel, a pupil is provided between the microlens and the photodiode by providing a light shielding layer having an opening for allowing light that has passed through different exit pupil regions of the imaging optical system to enter the photoelectric conversion unit. The divided focus detection signals are acquired. Then, only the signals from focus detection pixels divided into left and right or up and down are collected to form two pupil separated images, and the defocus amount is calculated based on the deviation amount of the pupil separated images. To do.

  The shift amount of the projection position of the image divided into pupils with respect to a certain defocus amount is determined by the baseline length between the two divided pupils. The longer the baseline length, the larger the shift amount and the focus detection accuracy is improved.

  An advantage of using some pixels of the image sensor as focus detection pixels is that focus detection by the phase difference method can be performed even during moving image shooting. On the other hand, it is required that the focus detection pixels also have proper exposure at the exposure at which the imaging pixels have proper exposure and can be read out without being distinguished from the imaging pixels.

JP 2000-156823 A JP 2007-103590 A

  However, in the case of dividing the pupil by providing a light shielding layer between the microlens and the photodiode, if the aperture is increased, the amount of received light can be increased, but the base line length cannot be increased.

  On the other hand, if the opening is made narrow and slit-shaped in order to increase the baseline length, the amount of light that can be received decreases, and the SN ratio deteriorates. As a result, there is a problem that the error of the correlation calculation for determining the image shift amount due to the influence of noise becomes large and the accuracy of focus detection is lowered. In addition, the accuracy of focus detection also deteriorates due to the deterioration of linearity due to a large deviation from the center of the dynamic range of the image sensor.

  In order to increase the baseline length, the pupil aperture is narrowed, and in order to match the sensitivity with the imaging pixels, Patent Document 2 proposes that the focus detection pixels be configured to be larger than the imaging pixels. Has been.

  However, the method disclosed in Patent Document 2 has a problem that an area that is lost due to focus detection pixels is missing. The video signal at the focus detection pixel position is created by interpolating from surrounding pixels using the same interpolation method as the defective pixel and signal processing. However, if the defect is large, the interpolation accuracy deteriorates and becomes conspicuous as a video signal. There is a problem.

  The present invention has been made in view of the above problems, and it is an object of the present invention to increase the base length for focus detection and to have sensitivity equivalent to that of a shooting pixel in a focus adjustment pixel. And

To achieve the above object, an imaging apparatus of the present invention is covered by a filter that first light receiving region which is a part of the area on one side than the center of the light receiving region transmits light of a first wavelength band Rutotomoni The light receiving regions other than the first light receiving region are covered with a filter that transmits light in a second wavelength band that is a partial wavelength band of the first wavelength band, and a plurality of discretely arranged a first focus detection pixel, the second light receiving region different from the first light receiving area with respect to the center is covered by a filter which transmits light of the first wavelength band Rutotomoni, the second light-receiving The light receiving region other than the region is covered with a filter that transmits light in the second wavelength band, and a plurality of second focus detection pixels arranged discretely and a light receiving region in the second wavelength band a filter that transmits light, the first wavelength An image sensor including a plurality of normal pixels covered by a filter that is different from a filter that transmits light in the region, and a signal output from each of the first focus detection pixel and the second focus detection pixel and a value, the first focus detection pixels, and the first focus detection pixel based on an output signal value from a plurality of the normal pixels arranged around the second focus detection pixels, respectively, and, And calculating means for obtaining a phase difference from the signal value obtained by performing the defective pixel correction process on the second focus detection pixel .

  According to the present invention, it is possible to increase the base line length for focus detection and to have the same sensitivity as that of the imaging pixel in the focus adjustment pixel.

1 is a block diagram illustrating a configuration of an imaging apparatus according to the present invention. FIG. 3 is a schematic cross-sectional view of a focus detection pixel and a normal pixel in the first embodiment. The figure which shows the spectral transmittance of a color filter. FIG. 3 is a diagram illustrating an example of an arrangement of color filters in the first embodiment. FIG. 3 is a layout diagram of focus detection pixels according to the first embodiment. The figure which shows an example of the pupil separation image based on the image signal output from the focus detection pixel. 10 is a flowchart of defocus amount calculation processing. FIG. 10 is a diagram illustrating an example of an arrangement of color filters according to a second embodiment. FIG. 10 is a schematic cross-sectional view of a focus detection pixel and a normal pixel in the second embodiment. FIG. 10 is a diagram illustrating an example of an arrangement of color filters according to a third embodiment.

  The best mode for carrying out the present invention will be described below in detail with reference to the accompanying drawings.

  FIG. 1 is a block diagram showing a configuration of an imaging apparatus such as a digital still camera or a digital video camera according to the present invention.

  In the configuration shown in FIG. 1, an optical image formed on the image sensor 11 by the photographing lens 10 is converted into an electric signal and output to the A / D converter 12. In the present embodiment, the image sensor 11 includes a normal pixel for acquiring an image signal and a focus detection pixel for performing phase difference type focus detection as described later. The A / D converter 12 converts the analog signal output from the image sensor 11 into a digital signal (image signal) and outputs the digital signal to the separation calculation unit 13.

  The separation calculation unit 13 stores the position of the focus detection pixel of the image sensor 11 and performs a calculation of a method to be described later at the timing when the focus detection pixel data is input. As a result, a pupil-divided focus detection signal corresponding to light transmitted through different areas of the exit pupil of the photographing lens 10 is obtained and output to the pupil separation image memory 17. At the same time, an image signal is generated by interpolation processing similar to that for a defective pixel, and is output to the signal processing unit 14 instead of the focus detection signal. Note that the image signal from the normal pixel is output to the signal processing unit 14 as it is.

  The signal processing unit 14 creates a luminance signal and a color difference signal from the image signal output from the separation calculation unit 13, and the image memory 15 stores the luminance signal and the color difference signal created by the signal processing unit 14.

  On the other hand, the pupil separation image memory 17 accumulates the focus detection signals separated by the separation calculation unit 13, and the CPU 16 performs a correlation operation on the pupil separation images accumulated in the pupil separation image memory 17 to shift the image projection position. Is detected and the defocus amount is calculated.

  A lens, a lens driving unit, a power source, and the like are not shown in the figure, but are not essential to the present invention, and thus description thereof is omitted.

<First Embodiment>
FIG. 2 is a schematic cross-sectional view of focus detection pixels (first pixels) and normal pixels of the image sensor 11 according to the first embodiment of the present invention. In FIG. 2, the center pixel is a focus detection pixel, and the left and right pixels are normal pixels.

  In FIG. 2, 101 and 108 are photoelectric conversion units, 102 is a wiring layer, 104 is a normal pixel color filter, 105 is a microlens, and 106 and 107 are focus detection pixel color filters. Each normal pixel is provided with a color filter having a single characteristic. Here, R (red), G (green), and B (blue) filters are arranged at a constant period. On the other hand, the focus detection pixels in the first embodiment are provided with color filters having different characteristics in the areas divided in the left and right (horizontal direction), and the spectral transmittance of one of the color filters is the spectrum of the other. The characteristic is that the transmittance is included. In the above description, the focus detection pixel is an area divided in the horizontal direction, but may be an area divided in two in the vertical direction or the oblique direction.

  Here, the characteristics of the filter mounted on the focus detection pixel will be described in detail with reference to FIG. FIG. 3 is a diagram showing the spectral transmittance, where the horizontal axis indicates the wavelength and the vertical axis indicates the transmittance. A curve 201 is a characteristic of a red (R) filter having a high transmittance of R (red), a curve 202 is a characteristic of a green (G) filter having a high transmittance of G (green), and a curve 203 is a transmittance of B (blue). Shows the characteristics of a high blue (B) filter. A curve 204 indicates the characteristics of the yellow (Ye) filter. The Ye filter shows a high transmittance in a portion where the transmittance of both the G filter and the R filter is high. That is, the Ye filter transmits the R wavelength and the G wavelength. As a combination having similar characteristics, the magenta (Mg) filter transmits R and B wavelengths, and the cyan (Cy) filter transmits G and B wavelengths.

  FIG. 4 is a diagram showing an example of the color filter arrangement in the first embodiment, in which R, G, and B primary color filters are arranged in a so-called Bayer arrangement. A pixel 301 (first focus detection pixel) and a pixel 311 (second focus detection pixel) adjacent to the two pixels shown in the center of FIG. 4 are focus detection pixels, and are divided into left and right (horizontal directions). Each region is covered with two different types of filters. Here, in the focus detection pixel 301, the left half is an R filter (a filter that transmits light in the second wavelength band), and the right half (first light receiving region) is a Ye filter (light in the first wavelength band). It is covered with a transparent filter. The right half of the focus detection pixel 311 is an R filter (a filter that transmits light in the second wavelength band), and the left half (second light receiving region) is a Ye filter (transmits light in the first wavelength band). Covered with a filter). From these focus detection pixels 301 and 311, a signal in which a signal corresponding to the light transmitted through the half-opening R filter and a signal corresponding to the light transmitted through the half-opening Ye filter is mixed is read out. Since the Ye filter has both R and G spectral transmittances, the signals output from the focus detection pixels 301 and 311 are signals corresponding to light transmitted through the full aperture R filter and the half aperture G filter. It can be regarded as equivalent. For this reason, the R signal of the full apertures of the focus detection pixels 301 and 311 is obtained by a technique of interpolating defective pixels from surrounding pixels, and if subtracted, a signal corresponding to the pupil-divided half aperture G signal is obtained. Can be separated.

  Next, a description will be given of a method in which the separation calculation unit 13 separates the signals obtained by pupil division in the focus detection pixels 301 and 311. Here, the focus detection pixel 301 will be described as an example. A plurality of R filter pixels 302 to 305 exist around the focus detection pixel 301. Pixels 302 to 305 are all aperture pixels without a member that restricts the aperture on the pixel surface. On the other hand, since the Ye filter transmits the R wavelength in the focus detection pixel 301, the sensitivity of R can be regarded as the same full aperture.

  Therefore, the signal value of the R signal of the entire aperture of the focus detection pixel 301 is calculated using a conventional defective pixel interpolation technique. Specifically, an average value of the signal value of the pixel 302, the signal value of the pixel 303, the signal value of the pixel 304, and the signal value of the pixel 305 is used. A value obtained by subtracting the average value of the pixels 302 to 305 from the signal value of the focus detection pixel 301 is the signal value of the G signal that has passed through the Ye filter of the focus detection pixel 301.

Expressing this as an expression,
Pixel 301 full aperture R signal value = (pixel 302 + pixel 303 + pixel 304 + pixel 305) / 4
Pixel 301 pupil division G signal value = pixel 301−pixel 301 full aperture R signal value

It is.
Since the G component is the component that contributes most to the luminance signal when creating the video signal, focus detection using the pupil-divided G signal has a more accurate value.

  FIG. 5 is a diagram showing an example of the arrangement of focus detection pixels in the image sensor 11, and a set of focus detection pixels 301 and 311 is interspersed with normal pixels and scattered at a constant ratio. In FIG. 5, a column in which G filters and B filters are alternately arranged has a focus detection pixel 301 having the filter arrangement shown in FIG. 4 and a focus detection pixel having a filter arrangement opposite to the focus detection pixel 301. Pixel 311 is included. In addition, focus detection pixels 312 and 313 are included in a column in which R filters and G filters are alternately arranged. In the focus detection pixel 312 (first focus detection pixel), the left half transmits a B filter (a filter that transmits light in the second wavelength band), and the right half transmits a cyan (Cy) filter (transmits the first wavelength band). Filter). The focus detection pixel 313 (second focus detection pixel) is covered with a Cy filter on the left half and a B filter on the right half. As described above, since the Cy filter transmits G and B wavelengths, the pupil is divided by replacing the R filter with the B filter and replacing the Ye filter with the Cy filter and performing the same processing as described above. G signal can be detected.

  FIG. 6 is a diagram showing an example of a pupil separation image stored in the pupil separation image memory 17, and reference numerals 801 and 802 denote pupil-divided pupil separation images obtained from left and right focus detection pixels, respectively. The CPU 16 obtains a deviation between the projection positions of the two pupil separated images.

  FIG. 7 is a flowchart showing the procedure of the phase difference method defocus amount calculation processing by the CPU 16.

  First, in S11, a level difference between two pupil separation images generated by shading is corrected. For example, correction is performed by applying a gain so that the average value or peak value of two separated pupil images are aligned. Next, correlation calculation is performed in S12. Specifically, the process of obtaining the sum of the absolute values of the differences by superimposing the two images is performed while shifting the overlapping position, and the position where the absolute value of the difference is the smallest is obtained. Next, in S13, sub-pixel matching is performed using a method called parabolic fitting based on values before and after the position where the absolute value of the difference becomes small. This improves the accuracy of the image shift amount. Finally, the defocus amount is obtained by multiplying the image shift amount obtained in S14 by a predetermined coefficient determined by the baseline length.

  As described above, according to the first embodiment, a specific aperture pupil component is separated by separating a specific color component by calculation with surrounding pixels.

  In the first embodiment described above, the average value of the red values of the surrounding four pixels is used as the R signal value of the entire aperture for obtaining the G signal value obtained by pupil division from the focus detection pixels. However, the present invention is not limited to this, and if the defective pixel correction process, for example, an adaptive process using the correlation of surrounding pixel values or a color that refers to a reference pixel that is not the same color. A process using the ratio of The better the defective pixel correction process, the better the pupil division pixel value separation performance.

  In the above-described example, the case where the pixel is divided into left and right (horizontal direction) has been described, but the pixel may be divided into upper and lower (vertical direction).

<Second Embodiment>
Next, a second embodiment of the present invention will be described. In the first embodiment described above with reference to FIG. 4, the focus adjustment pixels are configured to be covered half by R filter and Ye filter, or B filter and Cy filter. On the other hand, in the second embodiment, a long baseline length is secured by arranging the Ye filter (or Cy filter) in a slit shape at a position deviated from the center of the light receiving region. Since the configuration other than the filter is the same as that of the first embodiment described above, the description thereof is omitted.

  FIG. 8 is a diagram showing an example of the color filter arrangement in the second embodiment. Like FIG. 4 described in the first embodiment, R, G, and B primary color filters are arranged in a so-called Bayer arrangement. Has been. The difference from FIG. 4 is that the Ye filters in the focus adjustment pixels 401 and 402 are formed in a slit shape at symmetrical positions, and regions other than the Ye filter (regions other than the first light receiving region and second light receiving regions). The region other than the region) is covered with the R filter. Such focus adjustment pixels 401 and 402 are discretely arranged in the image sensor 11 as shown in FIG.

  FIG. 9 is a schematic sectional view of focus detection pixels and normal pixels of the image sensor 11 according to the second embodiment. In FIG. 9, the center pixel is a focus detection pixel, and the left and right pixels are normal pixels.

  In FIG. 9, reference numerals 501 and 502 denote photoelectric conversion units, and 507 denotes a wiring layer. In the second embodiment, a structure called a “back surface sensor” is shown in which the wiring layer 507 is configured on the back side of the photoelectric conversion units 501 and 502. 505 is a microlens, 506 is a color filter for normal pixels, and 503 and 504 are color filters for focus detection pixels. Here, the color filter 503 is an R filter (a filter that transmits light in the second wavelength band, and the color filter 504 is a Ye filter (a filter that transmits light in the first wavelength band).

  In the above-described example, the focus detection pixels are covered with the R filter and the slit-shaped Ye filter. However, as in the first embodiment, when focus detection pixels are arranged in the R pixels in the Bayer array, the B filter and the slit-shaped Cy filter are replaced.

  With the above configuration, it is possible to secure a longer base line length as compared with the configuration of the first embodiment.

<Third Embodiment>
Next, a third embodiment of the present invention will be described. Compared with the first embodiment described above, the third embodiment covers normal pixels with Mg, Cy, Ye, and G complementary color filters instead of R, G, and B primary color filters, and focus adjustment pixels. Is covered with a half-opening R filter and a Ye filter. With such a filter configuration, the method of separating the pupil-divided signal is different from the first and second embodiments described above, and this will be described below.

  FIG. 10 is a diagram showing an example of the color filter array in the third embodiment. As shown in FIG. 10, the arrangement is such that rows in which Cy filters and Ye filters are alternately arranged, G filters, and rows in which Mg filters are alternately arranged are repeated. A pixel 601 (first focus detection pixel) and a pixel 611 adjacent to the two pixels shown in the center of FIG. 10 are focus detection pixels (second focus detection pixels). In the focus detection pixel 601, the left half (first light receiving region) is an R filter (filter that transmits light in the second wavelength band), and the right half is Ye filter (filter that transmits light in the first wavelength band). Covered by. The right half of the focus detection pixel 611 is an R filter (a filter that transmits light in the second wavelength band), and the left half (second light receiving region) is a Ye filter (transmits light in the first wavelength band). Covered with a filter).

In the complementary color arrangement, the luminance and color difference are as follows: luminance Y = {(G + Cy) + (Mg + Ye)} / 2
Color difference R−Y = {(Mg + Ye) − (G + Cy)} = 2R−G
Color difference B−Y = {(Mg + Cy) − (G + Ye)} = 2B−G

Asking. And when approximating the primary color signal G = Y− (R−Y) − (B−Y)
R = n (R−Y) −m {Y− (B−Y)}
B = p (B−Y) −q {Y− (R−Y)}
It becomes. Note that n, m, p, and q are coefficients.
The R value of all the apertures of the focus detection pixels 601 and 611 is obtained from the surrounding normal pixels by the above formula. The processing after the signal values of the R signals of all the apertures of the focus detection pixels 601 and 611 are obtained is the same as in the first embodiment described above.

  Thus, it is possible to use a complementary color filter. Further, as described in the second embodiment, the R filter of FIG. 10 may be formed in a slit shape.

  In the first to third embodiments described above, the case where the half of the light receiving region of the focus detection pixel or the slit-like region is covered with a filter having transmission characteristics other than the filter used in the normal pixel has been described. . However, the present invention is not limited to this, and may be a half of the light receiving region and a slit-like region as long as the light transmitted through different regions of the exit pupil of the photographing lens 10 can be separated. , It can be a partial region on one side from the center.

<Fourth Embodiment>
In the first and second embodiments described above, pupil division is performed by covering the focus detection pixels with the R filter and the Ye filter. On the other hand, in the fourth embodiment, the Ye filter portion (a region other than the first light receiving region and a region other than the second light receiving region) is not covered with the color filter. In that case, it can be considered that the focus detection pixel is equivalent to transmitting red of the full aperture and cyan of the partial aperture. As a result, the same effect as in the first and second embodiments can be obtained.

  In the third embodiment described above, the same effect can be obtained by adopting a configuration in which the Ye filter portion is not covered with the color filter.

<Fifth Embodiment>
In the first to fourth embodiments, the normal pixels are each covered with a single characteristic color filter, and the image processing apparatus outputs a color image signal as a result of signal processing. In contrast, in the fifth embodiment, a color filter is not attached to a normal pixel. A black and white video signal having no color information is output as an image signal.

  The focus detection pixel of the fifth embodiment performs pupil division by creating a portion covered with the color filter and a portion not covered with the color filter. In that case, the color filter non-attached pixel value of the entire aperture is calculated by the defective pixel interpolation technique, and the value is subtracted from the focus detection pixel value. The signal value obtained as a result is a negative value, but the correlation calculation similar to that described in the first embodiment can be performed by inverting the sign of the obtained image.

<Sixth Embodiment>
In the above embodiment, the color filter having the same characteristics is used for left and right pupil separation. However, the left and right pupil separation color filters do not have to have the same characteristics, and the G signal pupil separation image is separated by a similar operation in the combination of the B filter and the Cy filter as well as the combination of the R filter and the Ye filter. Arithmetic is possible. That is, the same effect is obtained when R + Ye is a right pupil-separated signal and B + Cy is a left pupil-separated signal.

<Seventh Embodiment>
In the above embodiment, the left and right filter shapes are symmetrical with respect to the light receiving center of the pixel, but it is possible to suppress left and right variations in shading characteristics by gently shifting the center with respect to the lens optical axis center. is there. It is also effective to provide several types of light receiving areas separated by filters to deal with manufacturing variations of microlenses, deviation of the center of the optical axis of the lens from the center of the image sensor, and deviation of the pupil position due to the lens of the interchangeable lens. is there. In the above embodiment, a normal image is obtained in a mosaic arrangement, but a stripe arrangement may be used. In the above embodiment, a single-plate image sensor is used, but the same effect can be obtained with a system that obtains a color signal with two or three plates.

<Eighth Embodiment>
In the above-described embodiment, the arrangement of typical RGB primary color filters, Cy, Ye, and Mg complementary color filters has been described as an example. However, a mixed filter arrangement such as R, Ye, Cy, and B has the same effect. The necessary structural requirements for the pupil separation pixel include a first filter corresponding to the wavelength of the front aperture pixel that can be interpolated from the surrounding pixels, and a second filter that includes that wavelength and also passes a different wavelength. It is a thing.

<Ninth Embodiment>
In the above embodiment, in order to obtain a normal image, the focus detection pixel is corrected as a defective pixel, and the pupil separation image is separated by calculation to obtain the defocus amount. However, the camera has means for obtaining the defocus amount. There is no need. The signal from the image sensor may be stored in the memory as it is. In this case, the pupil separation image cannot be used for focus detection, but a distance image or a color image can be obtained by performing defective pixel correction or phase difference calculation by a PC or the like.

Claims (12)

Rutotomoni covered by a filter that first light receiving region which is a part of the area on one side than the center of the light receiving region transmits light of a first wavelength band, the light receiving region other than the first light receiving region is the first A plurality of first focus detection pixels, which are covered with a filter that transmits light in a second wavelength band, which is a partial wavelength band of the first wavelength band, and are discretely arranged; covered by a filter in which the second light receiving region different from the first light receiving region for transmitting light of said first wavelength band Rutotomoni, the second light of the light receiving region other than the light receiving region and the second wavelength band A plurality of second focus detection pixels that are covered by a filter that transmits light and a light receiving region includes a filter that transmits light in the second wavelength band. A filter that is different from the filter that transmits light An imaging device including a plurality of normal pixels which are covered by Ri,
The first focus detection pixel, and a signal value outputted from each of the second focus detection pixels, the first focus detection pixels and disposed around each of the second focus detection pixel Based on signal values output from the plurality of normal pixels , a phase difference between the signal values obtained by performing defective pixel correction processing on the first focus detection pixels and the second focus detection pixels is obtained. An imaging apparatus comprising: an arithmetic means. Before Symbol normal pixel is claim 1, characterized in that are covered respectively by the includes a second filter that transmits light in the wavelength band of, any of a plurality of filters that transmit light of different wavelength bands The imaging device described in 1. The filter that transmits light in the first wavelength band has a characteristic of transmitting light in a wider wavelength band than the filter that covers the normal pixels arranged around the pixels covered by the filter. and, wherein the first wavelength band, imaging apparatus according to claim 1 or 2 filter covering the normal pixel that is disposed on the periphery is characterized in that it comprises a wavelength band of light transmitted. The first light receiving region, which is a partial region on one side from the center of the light receiving region, is not covered with a filter , and the light receiving regions other than the first light receiving region transmit light in a predetermined wavelength band. A plurality of first focus detection pixels covered by a filter and arranged discretely and a second light receiving region different from the first light receiving region with respect to the center are not covered by the filter , and the first The light receiving regions other than the two light receiving regions are covered with a filter that transmits light in the wavelength band, and a plurality of discretely arranged second focus detection pixels and a plurality of light receiving regions covered with the filter An image sensor including normal pixels;
The first focus detection pixel, and a signal value outputted from each of the second focus detection pixels, the first focus detection pixels and disposed around each of the second focus detection pixel Based on signal values output from the plurality of normal pixels , a phase difference between the signal values obtained by performing defective pixel correction processing on the first focus detection pixels and the second focus detection pixels is obtained. An imaging apparatus comprising: an arithmetic means. The normal pixels, the image pickup apparatus according to any one of claims 1 to 4, characterized in that they are covered by primary color filters. The normal pixels, the image pickup apparatus according to any one of claims 1 to 4, characterized in that they are covered by a complementary color filter. The first light receiving region is one of regions obtained by dividing the light receiving region into two in the horizontal direction, the vertical direction, or the oblique direction, and the second light receiving region is the other region. the imaging apparatus according to any one of claims 1 to 6, wherein. It said first light receiving region and the second light receiving region, an imaging apparatus according to any one of claims 1 to 6, characterized in that said a central slit-shaped region offset from the light-receiving region. The computing means calculates the first focus detection pixel and the second focus detection pixel from the signal values output from the first focus detection pixel and the second focus detection pixel, respectively. according to any one of claims 1 to 8, characterized in that determining the phase difference between the obtained signal values by subtracting the average of the signal values output from the plurality of normal pixels arranged around Imaging device.   The first light receiving region, which is a partial region on one side of the center of the light receiving region, is covered with a filter that transmits light in the first wavelength band, and the light receiving regions other than the first light receiving region are the first light receiving region. A plurality of first focus detection pixels discretely arranged and covered with a filter that transmits light in the second wavelength band, which is a part of the wavelength band of the first wavelength band, and the first with respect to the center A second light receiving region different from the light receiving region is covered with a filter that transmits light in the first wavelength band, and the light receiving regions other than the second light receiving region transmit light in the second wavelength band. A plurality of second focus detection pixels that are covered and discretely arranged, and a light-receiving region that includes a filter that transmits light in the second wavelength band. A filter that is different from the filter that transmits An imaging device including a plurality of normal pixels which are covered by Ri,
  Storage means for storing the signal value output from each of the first focus detection pixel and the second focus detection pixel and the signal value output from the normal pixel in a storage medium;
  An imaging device comprising:   The first light receiving region, which is a partial region on one side from the center of the light receiving region, is not covered with a filter, and the light receiving regions other than the first light receiving region transmit light in a predetermined wavelength band. A plurality of first focus detection pixels covered by a filter and arranged discretely and a second light receiving region different from the first light receiving region with respect to the center are not covered by the filter, and the first The light receiving regions other than the two light receiving regions are covered with a filter that transmits light in the wavelength band, and a plurality of discretely arranged second focus detection pixels and a plurality of light receiving regions are covered with the filter. An image sensor including normal pixels of
  Storage means for storing the signal value output from each of the first focus detection pixel and the second focus detection pixel and the signal value output from the normal pixel in a storage medium;
  An imaging device comprising: 12. The signal value output from each of the first focus detection pixel and the second focus detection pixel stored in the storage medium by the imaging device according to claim 10 or 11, and output from the normal pixel Means for reading out the measured signal value;
Using the signal values output from the first focus detection pixel and the normal pixel from the signal values output from the first focus detection pixel and the second focus detection pixel, a defect is obtained. An image processing apparatus comprising: processing means for performing at least one of pixel correction processing and processing for obtaining a phase difference .

Images