JP7019442B2 - Image pickup device and its control method - Google Patents

Description

The present invention relates to an image pickup device that performs automatic focus detection and a control method thereof.

The focus detection method of the image pickup apparatus includes a phase difference detection method and a contrast detection method. Further, there is a technique in which the pixel portion of the image pickup element receives light from different pupil surfaces of the image pickup optical system, and the focus detection of the phase difference detection method is performed at the same time as the image pickup.

In the apparatus disclosed in Patent Document 1, the light flux focused by one microlens is photoelectrically converted by a divided photodiode (PD), so that each PD receives light from different pupil surfaces of the image pickup lens. .. Focus detection is performed by comparing the outputs of each PD. In Patent Document 2, two images having uniform brightness captured by the imaging unit are stored in a memory, and the two images are read out when the correction value is calculated, and the gain correction value and the offset correction value are calculated. The image pickup device is disclosed. By using the gain correction value and the offset correction value, it is possible to remove the fixed pattern noise for each pixel.

Japanese Unexamined Patent Publication No. 2001-083407 Japanese Unexamined Patent Publication No. 2011-44813

However, Patent Document 1 does not disclose focus detection in consideration of noise, and the correlation calculation result at the time of focus detection may vary due to noise for each pixel. Therefore, when there is a lot of noise, it is difficult to perform highly accurate focus detection. Further, in Patent Document 2, since it is necessary to record the correction value for the entire surface of the image pickup element, a memory capacity corresponding to the entire surface of the image pickup element is required. Further, since the correction is performed using the correction value stored in advance, the random noise cannot be corrected.
An object of the present invention is to provide an image pickup apparatus capable of more accurate focus detection with reduced influence of noise.

The image pickup apparatus of the present invention is either an image pickup element in which a pixel unit having a plurality of photoelectric conversion units for receiving light rays passing through different pupil region regions of the image pickup optical system is arranged, or any of the plurality of photoelectric conversion units. An acquisition means for acquiring a first signal acquired by the photoelectric conversion unit, a second signal obtained by subtracting the first signal from the signals obtained by the plurality of photoelectric conversion units, and a first signal with respect to the first signal. The reference signal and the second reference signal for the second signal are calculated, the first correction signal is calculated from the first signal and the first reference signal, and the second from the second signal and the second reference signal. The second signal is corrected by the first calculation means for calculating the correction signal and the first correction signal or the second reference signal, and the first signal is corrected by the second correction signal or the first reference signal. A second calculation means for calculating the image shift amount or the defocus amount by calculating the corrected correlation value between the first signal and the second signal is provided.

According to the present invention, it is possible to provide an image pickup apparatus capable of more accurate focus detection with reduced influence of noise.

It is a block diagram of the image pickup apparatus in embodiment of this invention. It is a pixel arrangement diagram of an image pickup element, and the block diagram of a readout circuit. It is explanatory drawing about the focal point detection of the phase difference detection method. It is explanatory drawing of the focal point detection signal of the phase difference detection method. It is explanatory drawing of the correlation operation using a focus detection signal. It is a flowchart of the focus detection in embodiment of this invention. It is explanatory drawing of the random noise and the correlation value of A image and B image. It is explanatory drawing of the inverse correlation noise and the correlation value of A image and B image. It is a flowchart of a signal correction process. It is explanatory drawing of the signal correction in 1st Embodiment. It is explanatory drawing of the signal correction following FIG. It is a detailed explanatory drawing which shows the example of a signal correction. It is a detailed explanatory drawing which shows another example of a signal correction. It is explanatory drawing of the signal correction in 2nd Embodiment.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In each embodiment, a camera system in which a lens device can be attached to a main body of an image pickup device will be described. However, the present invention is not limited to this, and the present invention can be applied to an image pickup device in which a lens unit (imaging optical system) and a camera body are integrally configured. Further, the present embodiment is not limited to the digital still camera, and can be applied to an image pickup device such as a video camera.

[First Embodiment]
First, the configuration of the image pickup apparatus of the present embodiment will be described with reference to FIG. FIG. 1 is a block diagram illustrating an interchangeable lens single-lens reflex type digital camera as an image pickup apparatus 10. The image pickup apparatus 10 includes a lens unit 100, which is an interchangeable lens, and a camera body 120. The lens unit 100 is detachably attached to the camera body 120 via the mount M shown by the dotted line in FIG.

The lens unit 100 includes a first lens group 101, an aperture 102, a second lens group 103, a focus lens group (hereinafter, simply referred to as “focus lens”) 104, and a drive and control system that constitute an imaging optical system. Have. The lens unit 100 is a photographing lens device that forms a subject image.

The first lens group 101 is arranged at the front end portion (end portion on the subject side) of the lens unit 100, and is held so as to be able to advance and retreat along the optical axis direction (OA). The aperture 102 adjusts the amount of light at the time of shooting by adjusting the aperture diameter thereof, and also functions as a shutter for adjusting the exposure seconds at the time of shooting a still image. The aperture 102 and the second lens group 103 can be integrally moved along the optical axis direction (OA), and the zoom function is realized by interlocking with the advancing / retreating operation of the first lens group 101. The focus lens 104 is movable along the optical axis direction (OA). Focus adjustment control, that is, focus control of the lens unit 100 is performed according to the position of the focus lens 104, and the focusing distance when the subject is in focus changes.

The zoom actuator 111 is driven by the zoom drive circuit 114 and moves the first lens group 101 and the second lens group 103 in the optical axis direction. As a result, a zoom operation for changing the angle of view of the lens unit 100 is performed. The aperture shutter actuator 112 is driven by the aperture shutter drive circuit 115 to change the aperture diameter of the aperture 102 and open / close the aperture 102. The focus actuator 113 is driven by the focus drive circuit 116 and moves the focus lens 104 in the optical axis direction. As a result, the focus of the lens unit 100 is controlled. The focus drive circuit 116 has a position detection function for detecting the current position (lens position) of the focus lens 104.

The lens MPU (Micro Processing Unit) 117 is a lens control unit that performs all calculations and controls related to the lens unit 100. The lens MPU 117 is connected to the camera MPU 125 through the mount M and communicates commands and data. For example, the lens MPU 117 acquires the position detection information of the focus lens 104 and notifies the lens position information in response to a request from the camera MPU 125. This lens position information is a lens that limits the position of the focus lens 104 in the optical axis direction, the position and diameter of the exit pupil in the optical axis direction when the movable optical member of the imaging optical system is not moving, and the light beam of the exit pupil. Includes information such as the position and diameter of the frame in the optical axis direction. Further, the lens MPU 117 controls the zoom drive circuit 114, the aperture shutter drive circuit 115, and the focus drive circuit 116 according to a control command from the camera MPU 125.

The lens memory 118 stores optical information necessary for automatic focus adjustment (AF) control. The camera MPU 125 controls the operation of the lens unit 100 by, for example, executing a program stored in the built-in non-volatile memory or the lens memory 118.

The camera body 120 includes an optical low-pass filter 121, an image sensor 122, and a drive and control system. The first lens group 101, the aperture 102, the second lens group 103, the focus lens 104, and the optical low-pass filter 121 constitute an imaging optical system. The optical low-pass filter 121 is an optical element that reduces false colors and moire in captured images.

The image pickup element 122 constitutes an image pickup unit, photoelectrically converts a subject image (optical image) formed via the lens unit 100, and outputs an image pickup signal and a focus detection signal. The image pickup device 122 is composed of, for example, a CMOS (complementary metal oxide semiconductor) image sensor and its peripheral circuit. The image pickup device 122 has a pupil division function, and includes pupil division pixels capable of focus detection (phase difference AF) of a phase difference detection method.

The image sensor drive circuit 123 controls the operation of the image sensor 122, converts the image signal output from the image sensor 122 into A / D (analog / digital), and transmits the image signal to the camera MPU 125 and the image processing circuit 124.

The image processing circuit 124 acquires an image signal output from the image pickup element 122 and performs general image processing performed by a digital camera, such as γ conversion, white balance adjustment processing, color interpolation processing, and compression coding processing. .. The image processing circuit 124 generates data for phase difference AF and image data for display, recording, and contrast AF (TVAF) based on the acquired image signal.

The camera MPU 125 is a control center that performs calculations and controls for the entire camera system. The camera MPU 125 controls an image sensor drive circuit 123, an image processing circuit 124, a display unit 126, an operation unit 127, a memory 128, a phase difference AF unit 129, and a TVAF unit 130. The camera MPU 125 is connected to the lens MPU 117 via the signal line of the mount M, and communicates commands and data with the lens MPU 117. The camera MPU 125 issues a lens position acquisition request and a lens drive request with a predetermined drive amount to the lens MPU 117, and also issues an optical information acquisition request specific to the lens unit 100 and the like.

The camera MPU 125 has a built-in ROM 125a, RAM 125b, and EEPROM 125c. The ROM (Read-Only Memory) 125a stores a program that controls the operation of the camera body 120. The RAM 125b (Random Access Memory) is a memory for storing data such as variable values. The EEPROM (Electrically Erasable Programmable Read-Only Memory) 125c stores various parameters. The camera MPU 125 executes the focus detection process based on the program stored in the ROM 125a. In the focus detection process, correlation calculation processing is executed using a pair of image signals obtained by photoelectric conversion of an optical image formed by light flux passing through different pupil region regions of an imaging optical system.

The display unit 126 includes an LCD (liquid crystal display) and the like, and displays information on the shooting mode of the image pickup device 10, a preview image before shooting and a confirmation image after shooting, a display image of the in-focus state at the time of focus detection, and the like. .. The operation unit 127 includes a power switch, a release (shooting trigger) switch, a zoom operation switch, a shooting mode selection switch, and the like. The memory 128 is, for example, a detachable flash memory for recording captured image data and the like.

The phase-difference AF unit 129 performs focus detection by the imaging surface phase-difference method. That is, the phase difference AF unit 129 performs the focus detection process by the phase difference method based on the image signal for focus detection obtained by the image sensor 122 and the image processing circuit 124. Specifically, the image processing circuit 124 generates a pair of image data formed from light passing through a pair of pupil portion regions of an imaging optical system as focus detection data. The phase difference AF unit 129 detects the amount of defocus based on the amount of misalignment of the pair of image data. In the present embodiment, the image pickup surface phase difference AF based on the output of the image pickup element 122 is performed without using a dedicated AF sensor. The phase difference AF unit 129 has an acquisition unit 129a and a calculation unit 129b. These operations will be described later. The acquisition unit 129a or the calculation unit 129b may be provided in the camera MPU 125.

The TVAF unit 130 performs the focus detection process of the contrast detection method based on the evaluation value for TVAF generated by the image processing circuit 124. The evaluation value for TVAF is an evaluation value based on the contrast information of the image data. In the focus detection process of the contrast detection method, the focus lens 104 moves, and the focus lens position where the evaluation value reaches the peak value is detected as the in-focus position.

As described above, the image pickup apparatus 10 can be executed by combining the image pickup surface phase difference AF and the TVAF. Depending on the situation, the imaging surface phase difference AF and TVAF can be selectively used, or both methods can be used in combination. The phase difference AF unit 129 and the TVAF unit 130 constitute a focus control means for controlling the position of the focus lens 104 by using the data of the respective focus detection results.

Next, the operation of the phase difference AF unit 129 will be described with reference to FIG. FIG. 2A is a pixel arrangement diagram of the image pickup device 122 according to the present embodiment. The horizontal direction is defined as the X direction, and the vertical direction is defined as the Y direction. FIG. 2A shows a state in which the range of 6 rows in the Y direction and 8 columns in the X direction of the two-dimensional CMOS area sensor is observed from the lens unit 100 side. The image pickup device 122 is provided with a Bayer array color filter. Green (G) and red (R) color filters are alternately arranged in the odd-numbered row pixel portions in order from the left, and blue (B) and green (G) in order from the left in the even-numbered row pixel portion. Color filters are arranged alternately. In the pixel unit 211, the on-chip microlens 211i is shown by a circular frame. A plurality of rectangular frames arranged inside the on-chip microlens 211i indicate a first photoelectric conversion unit 211a and a second photoelectric conversion unit 211b.

All the pixel units 211 in the image sensor 122 are divided into two in the X direction, and the photoelectric conversion signal by one of the photoelectric conversion units (211a or 211b) and the sum of the two photoelectric conversion signals can be read independently. be. Hereinafter, the image based on the output of the first photoelectric conversion unit 211a is referred to as an A image, and the image based on the output of the second photoelectric conversion unit 211b is referred to as a B image. An image based on the outputs of the photoelectric conversion units 211a and 211b is called an A + B image.

The camera MPU 125 subtracts the photoelectric conversion signal by the first photoelectric conversion unit 211a from the photoelectric conversion signal related to the A + B image, and acquires the photoelectric conversion signal related to the B image. That is, a B image signal based on the photoelectric conversion signal by the second photoelectric conversion unit 211b can be obtained. The camera MPU 125 performs the above series of processes, acquires an A image signal by the first photoelectric conversion unit, and acquires a B image signal by the second photoelectric conversion unit. Data for phase difference AF can be calculated from the A image signal and the B image signal. Further, the A image signal and the B image signal can be used to generate a 3D (3-Dimensional) image, that is, an image having a parallax for forming a stereoscopic image (parallax image). In this case, the image generation unit generates data of a parallax image composed of a plurality of images having different viewpoints. Further, the A + B image data obtained from the sum of the signals of the two photoelectric conversion units can be used as normal captured image data.

Here, a pixel signal when performing phase-difference AF will be described. In the present embodiment, the emission light flux of the imaging optical system is divided into pupils by the microlens 211i shown in FIG. 2 and the photoelectric conversion units 211a and 211b divided into two. The photoelectric conversion units 211a and 211b form pupil-divided pixels. An image formed by connecting the outputs of the first photoelectric conversion unit 211a in a plurality of pixel units 211 arranged in the same pixel row within a predetermined range is referred to as an AF image A. Similarly, an image formed by connecting the outputs of the second photoelectric conversion unit 211b is referred to as an AF B image.

As the output of the photoelectric conversion units 211a and 211b, a pseudo luminance (Y) signal calculated by adding the green, red, blue, and green outputs of the Bayer array included in the unit array of the color filter is used. .. However, the AF image A and the B image may be organized for each of the red, blue, and green colors. The relative image shift amount between the AF image A and the AF image B generated in this way is calculated by the correlation calculation. Based on this image shift amount, the focus shift amount in a predetermined region, that is, the defocus amount can be calculated.

In the present embodiment, a process of reading the output of one of the two photoelectric conversion units and the output of both photoelectric conversion units from the image pickup element 122 is performed. One of the AF image A and the AF image B is not output from the image sensor 122, but the sum of the A image output and the B image output is output as described above. Therefore, the focus detection can be performed by acquiring the other signal from the difference between the sum signal (A + B image signal) and one output signal. Since the structure of the image pickup device 122 is known, detailed description thereof will be omitted.

FIG. 2B is a configuration diagram of a readout circuit of the image pickup device 122. The readout circuit includes a horizontal scanning circuit 151 and a vertical scanning circuit 153. Horizontal scanning lines 152a and 152b and vertical scanning lines 154a and 154b are wired at the boundary of each pixel unit, respectively. Each signal from the photoelectric conversion units 211a and 211b is read out to the outside via these scanning lines.

The image pickup device 122 has the following two types of read-out modes in addition to the read-out in the pixel unit. The first readout mode is referred to as an all-pixel readout mode, which is a mode for acquiring a high-definition still image. In the first read mode, the signals of all pixels are read. On the other hand, the second read mode is referred to as a thinned read mode, which is a mode for recording a moving image or displaying a preview image. In the second read mode, the number of pixels required is less than the total number of pixels. That is, of all the pixels, only the pixels thinned out at a predetermined ratio in both the X direction and the Y direction are read out. Further, when high-speed reading is required, the thinning-out reading mode is used. At the time of thinning out reading in the X direction, signals can be added to improve the S / N (signal-to-noise) ratio. Further, when reading out thinned out in the Y direction, the signal output of the line to be thinned out is ignored and is not used. Normally, phase difference AF or TVAF processing (automatic focus detection and focus adjustment) is performed in the second readout mode.

Next, the focus detection process of the phase difference detection method will be described with reference to FIGS. 3 and 4. FIG. 3 is a schematic diagram illustrating the focus detection process of the phase difference detection method. FIG. 4 is a graph showing an A image signal and a B image signal.

With reference to FIG. 3, the relationship between the focus (focus position) and the phase difference in the image pickup device 122 will be described. FIG. 3A shows the positional relationship between the lens unit 100, the subject 300, the optical axis 301, and the image pickup device 122, and the luminous flux when the subject is in focus, that is, in the focused state. FIG. 3B shows the positional relationship and the luminous flux of each part in the out-of-focus state where the subject is out of focus. FIG. 3 shows a state in which the image pickup device 122 shown in FIG. 2 (A) is viewed from the cross-sectional direction.

A microlens 211i is provided in each pixel portion of the image pickup device 122. The photoelectric conversion units 211a and 211b composed of photodiodes receive light that has passed through the same microlens 211i. Lights having different phases and different images are incident on the photoelectric conversion units 211a and 211b. In FIG. 3, the first photoelectric conversion unit is represented by A, and the second photoelectric conversion unit is represented by B.

The image pickup element 122 has a first photoelectric conversion unit and a second photoelectric conversion unit for one microlens 211i, and a large number of microlenses 211i are arranged two-dimensionally. In this embodiment, an example in which two photoelectric conversion units 211a and 211b are arranged for one microlens 211i will be described, but the present invention is not limited to this. For example, a configuration having four photoelectric conversion units for one microlens, that is, a configuration in which two photodiodes are arranged in each of the vertical direction and the horizontal direction may be used. A plurality of photodiodes may be configured to share one microlens 211i and to be arranged in a plurality of positions in at least one of the vertical direction and the horizontal direction.

The lens unit 100 shown in FIG. 3 is an equivalent image pickup lens when the first lens group 101, the second lens group 103, and the focus lens 104 are considered as one lens. The light emitted from the subject 300 passes through each region of the lens unit 100 with the optical axis 301 as the center, and forms an image on the image pickup device 122. Here, the position of the exit pupil and the center position of the image pickup lens are the same.

Such a configuration is equivalent to a configuration in which the pupil region of the imaging optical system is symmetrically divided depending on whether the imaging optical system is viewed from the first photoelectric conversion unit or the second photoelectric conversion unit. In other words, it has a so-called pupil division configuration in which the luminous flux from the imaging optical system is divided into two light fluxes, and the divided light fluxes are incident on the first and second photoelectric conversion units, respectively. The first and second photoelectric conversion units receive light passing through different pupil region regions of the exit pupils of the imaging optical system, perform photoelectric conversion, and function as focus detection pixels. The first and second photoelectric conversion units also function as imaging pixels by adding (summing or synthesizing) signals from each other.

The luminous flux from a specific point on the subject 300 is divided into a luminous flux ΦLa and a luminous flux ΦLb. The luminous flux ΦLa passes through the first pupil portion region corresponding to the first photoelectric conversion unit 211a and is incident on the first photoelectric conversion unit. The luminous flux ΦLb passes through the pupil portion region corresponding to the second photoelectric conversion unit 211b and is incident on the second photoelectric conversion unit. The two luminous fluxes ΦLa and ΦLb are incident from the same point on the subject 300. Therefore, in the focused state of the image pickup optical system, that is, in the state where the subject 300 is in focus, as shown in FIG. 3A, the image sensor passes through the same microlens 211i and reaches one point on the image pickup element 122. do. Therefore, the image signals obtained from the first and second photoelectric conversion units match each other.

On the other hand, FIG. 3B shows an out-of-focus state in which the image pickup position and the image pickup device 122 are deviated by Y in the optical axis direction. In the out-of-focus state, the arrival positions of the luminous fluxes ΦLa and ΦLb on the imaging surface are deviated from each other in the direction perpendicular to the optical axis 301 by the change in the incident angle of the luminous fluxes ΦLa and ΦLb to the microlens 211i. Therefore, a phase difference occurs in the image signals obtained from the first and second photoelectric conversion units, respectively. The first photoelectric conversion unit 211a photoelectrically converts the subject image to generate a first signal, and the second photoelectric conversion unit 211b photoelectrically converts the subject image to generate a second signal. The first signal and the second signal having a phase difference generated from the subject image are output to the outside of the image pickup device 122 and used for focus detection.

As described above, the image sensor 122 independently reads the first signal as a focus detection signal, and adds (sums or combines) the respective signals of the first photoelectric conversion unit and the second photoelectric conversion unit to obtain an image pickup signal. read out. The first signal corresponds to the A image signal, and the second signal corresponds to the B image signal. The signal obtained by adding the signals of the first and second photoelectric conversion units corresponds to the signal of the IMG image which is the added image.

In the present embodiment, a plurality of photoelectric conversion units are arranged for one microlens, and the pupil-divided light flux is incident on each photoelectric conversion unit, but the present invention is not limited to this. For example, as a configuration of a focus detection pixel, there is a configuration in which one photodiode is provided for a microlens, and pupil division is performed by light-shielding in the left-right or up-down direction by a light-shielding layer. Further, there is a configuration in which a pair of focus detection pixels are discretely arranged in an array of a plurality of image pickup pixels, and an A image signal and a B image signal are acquired from the pair of focus detection pixels.

The phase difference AF unit 129 performs focus detection using the A image signal and the B image signal acquired by the acquisition unit 129a. FIG. 4A is a graph showing the intensity distribution of the A image signal and the B image signal in the focused state shown in FIG. 3A. In FIG. 4A, the horizontal axis represents the pixel position and the vertical axis represents the intensity of the output signal. In the focused state, the A image signal and the B image signal coincide with each other.

FIG. 4B is a graph showing the intensity distribution of the A image signal and the B image signal in the out-of-focus state shown in FIG. 3B. The settings of the horizontal axis and the vertical axis are the same as those in FIG. 4 (A). In the out-of-focus state, the A image and the B image have a phase difference for the reason described above, and the peak positions of the intensities are deviated from each other. The deviation amount X is shown in the figure. The calculation unit 129b of the phase difference AF unit 129 calculates the deviation amount X for each frame of the captured image, and performs a predetermined calculation process using the calculated deviation amount X, whereby the focus deviation amount, that is, FIG. 3 (B). ) Is calculated. The phase difference AF unit 129 outputs the calculated Y value to the camera MPU 125. The camera MPU 125 outputs a control signal based on the Y value to the focus drive circuit 116 via the lens MPU 117.

The focus drive circuit 116 calculates the amount of movement of the focus lens 104 based on the Y value indicated by the control signal, and outputs the drive signal to the focus actuator 113. The focus lens 104 is moved to a focusing position in which a predetermined subject is in focus by the focus actuator 113. In this way, the in-focus state is realized.

The correlation operation will be described with reference to FIG. FIG. 5 is an explanatory diagram of the correlation operation. FIG. 5A shows a plurality of graphs with the horizontal position (horizontal pixel position) of the pixel as the horizontal axis and the level (intensity) of the A image signal and the B image signal with respect to the horizontal pixel position as the vertical axis. FIG. 5A shows a plurality of graphs when the position of the A image signal is shifted within the range of the shift amount (−S to + S). Here, the sign of the shift amount when the A image signal is shifted to the left is negative, and the sign of the shift amount when the A image signal is shifted to the right is positive. The absolute value of the difference between the A image signal and the B image signal corresponding to each position is calculated, and the value obtained by adding the absolute values of each pixel position is calculated as the correlation value (correlation amount or correlation data) for one line. .. The correlation value calculated in each row may be added over a plurality of rows in each shift amount.

FIG. 5B is a graph showing the correlation value calculated for each shift amount in the example of FIG. 5A. In FIG. 5B, the horizontal axis represents the shift amount and the vertical axis represents the correlation data. In the example shown in FIG. 5A, "shift amount = X" is a shift amount corresponding to the focal position where the A image signal and the B image signal overlap each other. As shown in FIG. 5B, the correlation value becomes the minimum value when “shift amount = X”. That is, the shift amount at which the correlation value becomes the minimum value can be calculated. The method for calculating the correlation value applicable to the present embodiment is not limited to this. Any calculation method can be applied as long as it is a calculation method showing the correlation between the data of the A image and the B image.

Focus detection will be described with reference to FIG. FIG. 6 is a main flowchart of focus detection in this embodiment. The processes shown in each step of FIG. 6 are mainly executed by the phase difference AF unit 129, the focus actuator 113, the focus drive circuit 116, and the lens MPU 117 based on the command of the camera MPU 125.

In S601, the phase difference AF unit 129 sets a focus detection region to be targeted for focus detection from the effective pixel region of the image sensor 122. Subsequently, in S602, the phase difference AF unit 129 acquires a focus detection signal. That is, the first signal (A image signal) by the first photoelectric conversion unit 211a and the second signal (B image signal) by the second photoelectric conversion unit 211b are acquired from the image sensor 122. In S603, the phase difference AF unit 129 performs shading correction processing (optical correction processing) on each of the A image signal and the B image signal. In the focus detection of the phase difference detection method, the focal state is detected based on the correlation value (the degree of signal matching) between the A image and the B image. The occurrence of shading may reduce the correlation between the A image signal and the B image signal. A shading correction process is executed in order to improve the correlation between the A image and the B image and perform focus detection satisfactorily.

In S604, the phase difference AF unit 129 performs signal correction processing on each of the A image signal and the B image signal. The details of the signal correction process will be described later with reference to FIG. In S605, the phase difference AF unit 129 performs filter processing on each of the A image signal and the B image signal. Generally, the focus detection of the phase difference detection method is performed in a state where the defocus amount is relatively large (large defocus state), so that the filter is configured so that the pass band of the filter processing includes the low frequency band. However, if necessary, when the focus is adjusted from the large defocus state to the small defocus state, the pass band of the filter processing at the time of focus detection may be adjusted to the high frequency band side according to the defocus state. ..

In S606, the calculation unit 129b in the phase difference AF unit 129 performs a correlation calculation on the filtered A image signal and the B image signal, and performs a correlation calculation indicating the degree of coincidence of the signals (correlation between the A image and the B image). Calculation result) is calculated. In S607, the calculation unit 129b calculates the defocus amount based on the correlation value calculated in S606. Specifically, the phase difference AF unit 129 calculates the image shift amount X based on the shift amount at which the correlation value becomes the minimum value by the sub-pixel calculation. Then, the phase difference AF unit 129 calculates the defocus amount by multiplying the image shift amount X by the conversion coefficient. The conversion coefficient is determined according to the image height of the focal point detection region, the F value (aperture value) of the aperture 102, and the exit pupil distance of the lens unit 100. This completes the focus detection process.

With reference to FIG. 7, the concept of the random noise of the A image and the B image and the correlation value calculated from the A image and the B image including the random noise will be described. FIG. 7A is an explanatory diagram of random noise of the A image and the B image. The A image signal and the B image signal each include noise for each pixel. In FIG. 7A, the pixel position is represented in the horizontal axis direction, and the intensity of the output signal is represented in the vertical axis direction. In order to show each of the A image and the B image in an easy-to-understand manner, the reference positions of the A image and the B image are displayed by shifting each other in the vertical direction for convenience.

The noise for each pixel shown in FIG. 7A is random noise randomly generated in the A image and the B image. The intensities of the output signals of the A image and the B image are different for the same pixel. Therefore, there is no correlation between the A image and the B image.

FIG. 7B is a graph showing the correlation values calculated for the intensities of the output signals of the A image and the B image of FIG. 7A. In FIG. 7B, the horizontal axis represents the shift amount, and the vertical axis represents the correlation value (correlation data). It can be seen that the correlation value changes irregularly with respect to the change in the shift amount and varies.

As described above, the random noise generated in the A image and the B image causes the variation of the correlation value, has a plurality of minimum values and maximum values, and lowers the calculation accuracy of the image shift amount X. Therefore, it is necessary to reduce random noise in order to perform highly accurate focus detection.

Next, with reference to FIG. 8, the noise (inverse correlation noise) caused by subtracting the A image from the IMG image to obtain the signal corresponding to the B image will be described. FIG. 8 is an explanatory diagram of inversely correlated noise and the correlation value. FIG. 8A shows noise caused by subtracting the A image from the IMG image to obtain a signal corresponding to the B image. The pixel position is represented in the horizontal axis direction, and the strength of the output signal is represented in the vertical axis direction. In order to show the A image and the B image, respectively, the reference positions of the A image and the B image are shifted in the vertical direction for convenience.

When a B image is generated from an IMG image and an A image, which are additive images, and a correlation calculation is performed, the random noise amount in which the sign of the random noise amount superimposed on the A image is inverted is superimposed on the B image. At the position where the A image becomes the maximum value, the B image takes the minimum value, and the A image and the B image have opposite phase noise. Therefore, when the shift amount is zero, the degree of coincidence between the A image signal and the B image signal becomes low, and a large maximum value is generated in the correlation value at the position where the shift amount = 0.

FIG. 8B shows a graph of the correlation value, the horizontal axis represents the shift amount, and the vertical axis represents the correlation value (correlation data). When the shift amount is zero, the correlation amount is calculated to be larger than that of other shift amounts, and the maximum value is generated at the position where the shift amount = 0.

As described above, the inverse correlation noise generated in the A image and the B image causes a maximum value in the correlation value, and lowers the calculation accuracy of the image shift amount X. In order to perform high-precision focus detection, it is necessary to reduce inverse correlation noise. Therefore, in the present embodiment, the noise generated in the A image and the B image is reduced by correction, and the correlation value is calculated using the corrected signal. As a result, the influence of noise generated in the A image and the B image can be removed or reduced, and highly accurate focus detection can be performed without lowering the calculation accuracy of the image shift amount X.

The signal correction (S604 in FIG. 6) in the present embodiment will be described with reference to FIGS. 9 to 13. FIG. 9 is a flowchart illustrating signal correction, and the processing shown in each step is mainly executed by the camera MPU 125 and the phase difference AF unit 129. 10 and 11 are explanatory views of the signal state in each step of FIG. 9, where the horizontal axis represents the pixel position and the vertical axis represents the signal value. 12 and 13 are detailed explanatory views of signal correction regarding the sign and magnitude relationship between the A image and the IMG image.

In S901 of FIG. 9, the camera MPU 125 acquires the shading-corrected first signal in S603 of FIG. 6, that is, the focus detection signal corresponding to the A image. In S902, the camera MPU 125 acquires a shading-corrected second signal, that is, a focus detection signal corresponding to the B image. The focus detection signals acquired in S901 and S902 are in a state where noise is superimposed on the subject signal. Specific examples will be described with reference to FIGS. 10 and 11.

FIG. 10A shows a subject signal in a state where noise components are not superimposed. FIG. 10B shows an IMG image signal on which a noise component is superimposed, and FIG. 10C shows an A image signal and a B image signal on which a noise component is superimposed, respectively. These show a signal in which the random noise of FIG. 7 and the inverse correlation noise of FIG. 8 are superimposed on the subject signal. In this way, when the A image signal and the B image signal contain a large amount of random noise and inverse correlation noise, a plurality of minimum values (minimum points) and maximum values (maximum points) are generated in the correlation calculation result, and the focus is highly accurate. Cannot detect.

In S903 of FIG. 9, the camera MPU 125 calculates the first reference signal used for correction, and in S904, the second reference signal is calculated. The first reference signal and the second reference signal are as shown in FIG. 11 (A). The level of these reference signals is the black level of the output of the image sensor, and is a constant value stored in the memory 128. In the present embodiment, an example in which the black level is stored as a constant value will be described, but the present invention is not limited to this, and a different value is stored as a level value for each pixel in consideration of the fixed pattern noise for each pixel. May be good. Further, in the present embodiment, the level value stored in the memory is acquired and used as the reference signal, but the black level may be calculated from the image signal or the like acquired for black level correction and used as the reference signal. Further, a reference signal may be generated using a value obtained by adding a predetermined value (noise variation amount, etc.) to the minimum value of the acquired focus detection signal, or an average value or an intermediate value obtained from a plurality of lower values. .. In the present embodiment, a common reference signal is used for the A image and the B image, but individual reference signals may be calculated for each of the A image and the B image.

Subsequently, in S905, the camera MPU 125 calculates the first correction signal from the difference between the first signal acquired in S901 and the first reference signal acquired in S903. Similarly, in S906, the camera MPU 125 calculates the second correction signal from the difference between the second signal acquired in S902 and the second reference signal acquired in S904. FIG. 11B exemplifies the first correction signal and the second correction signal. These signals are signals obtained by subtracting the signal shown in FIG. 11 (A) from the signal shown in FIG. 10 (C).

The first correction is performed in S907 to S910, and the second correction is performed in S911 to S914. Both the first correction signal and the second correction signal are signals obtained by subtracting a reference signal (noise variation center value) from the focus detection signal. Therefore, when the value of the correction signal is a positive value, it may contain a component of the subject signal, but when it is a negative value, it can be said that it is only noise. That is, when the values of the first correction signal and the second correction signal are negative values, it can be determined that they are at least noise components, and negative noise can be detected. Further, the noise of the A image is inverted and superimposed on the B image. Therefore, it can be said that the negative noise detected in the A image is superimposed on the B image as positive noise, and the negative noise detected in the B image is superimposed on the A image as positive noise. Therefore, in the first correction, a process of removing or reducing the positive noise of the B image is performed by adding the negative noise component detected in the A image to the B image. Further, a process of removing or reducing the positive noise of the A image is performed by adding the negative noise component detected in the B image to the A image.

Specifically, in S907, the camera MPU 125 compares the value of the first correction signal with the first threshold value. The first threshold is, for example, zero. The camera MPU 125 determines whether or not the value of the first correction signal is a negative value, and if the value is negative, the process proceeds to S908, and if the value is not negative, the process proceeds to S909. In S908, the camera MPU 125 performs the first correction for the second signal by adding the first correction signal to the second signal. Then proceed to S909.

In S909, the camera MPU 125 compares the value of the second correction signal with the second threshold value. The second threshold is, for example, zero. The camera MPU 125 determines whether or not the value of the second correction signal is a negative value, and if the value is negative, the process proceeds to S910, and if the value is not negative, the process proceeds to S911. In S910, the camera MPU 125 performs the first correction for the first signal by adding the second correction signal to the first signal. FIG. 11C illustrates the first signal (A image) and the second signal (B image) after the first correction. It can be seen that the noise can be reduced as compared with the first signal (A image) and the second signal (B image) before the correction shown in FIG. 10 (C).

Further, when the first signal (A image) and the second signal (B image) after the first correction are smaller than the reference signal (noise variation center value), it can be determined that they are noise components. Therefore, in the second correction, when both the first signal (A image) and the second signal (B image) are smaller than the reference signal, noise is removed by using the first signal and the second signal as the reference signal. Is executed.

Specifically, in S911, the camera MPU 125 determines whether or not the first signal is smaller than the first reference signal (for example, the signal value is zero). If the first signal is smaller than the first reference signal, the process proceeds to S912, and if the first signal is not smaller than the first reference signal, the process proceeds to S913. In S912, the camera MPU 125 performs the second correction with respect to the first signal by using the first reference signal as the first signal. Then proceed to S913.

In S913, the camera MPU 125 determines whether the second signal is smaller than the second reference signal (for example, the signal value is zero). When the second signal is smaller than the second reference signal, the process proceeds to S914, and when the second signal is not smaller than the second reference signal, the signal correction is terminated. In S914, the camera MPU 125 performs the second correction with respect to the second signal by using the second reference signal as the second signal. FIG. 11D exemplifies the first signal (A image) and the second signal (B image) after the second correction. It can be seen that the negative noise can be removed as compared with the first signal (A image) and the second signal (B image) after the first correction shown in FIG. 11 (C).

With reference to FIGS. 12 and 13, signal correction will be described in detail with respect to the sign and magnitude relationship between the A image and the IMG image. Here, for the sake of simplicity, a case where the level of the reference signal is set to zero will be described as an example. That is, when the data value is a positive value, it may contain a component of the subject signal, but when the data value is a negative value, it can be said that it is a noise component.

The bar graphs shown in FIGS. 12 and 13 show the data values of the A image, the B image, and the IMG image. From the left side, the acquired A image, IMG image, A image and B image after B image generation, A image and B image after the first correction, and A image and B image after the second correction, respectively. Represents. Further, Sa, Sb (= Simg-Sa) and Simg shown in the figure represent subject signals of A image, B image and IMG image, respectively, and Na and Nimg represent noise of A image and IMG image, respectively. ..

FIG. 12A shows an example of signal correction when the sign and magnitude relationship between the A image and the IMG image are 0 <A <IMG. In FIG. 12A, since the data of the A image and the IMG image are both positive values, the acquired A image and the IMG image are represented as follows.
A = Sa + Na
IMG = Simg + Nimg

The B image data is generated by subtracting the A image data from the IMG image data. The generated B image is represented as follows.
B = IMG-A
= Simg + Nimg- (Sa + Na)
= Sb + Nimg-Na

In this case, regarding the first correction, since the data of both the A image and the B image are positive values, the correction is not performed. Similarly, regarding the second correction, since both the A image and the B image are positive values in the data, the correction is not performed. As described above, when 0 <A <IMG, the data of the A image and the B image may contain the component of the subject signal, so that the correction is not performed. That is, the component of the subject signal is not removed from each data.

FIG. 12B shows an example of signal correction when the sign and magnitude relationship between the A image and the IMG image are 0 <IMG <A. In FIG. 12B, since the data of the A image and the IMG image are both positive values, the acquired data of the A image and the IMG image are represented as follows.
A = Sa + Na
IMG = Simg + Nimg

The B image data is generated by subtracting the A image data from the IMG image data. The generated B image is represented as follows.
B = IMG-A
= Simg + Nimg- (Sa + Na)
= Sb + Nimg-Na

In this case, in the first correction, since the data of the B image is a negative value, the first correction is performed on the A image. Further, since the data of the B image is a negative value, the subject signal Sb of the B image is 0, so Sa = Simg from the above equation. The corrected A image is represented as follows.
A = Sa + Na + B
= Sa + Na + Nimg-Na
= Sa + Nimg

In the second correction, since the data of the B image is a negative value, that is, smaller than the reference signal, the second correction is performed on the B image. The corrected B image is represented as follows.
B = 0

As described above, when 0 <IMG <A, the component of the A image that may contain the component of the subject signal is not removed, and the noise Na of the A image is removed. .. That is, noise reduction processing is performed.

FIG. 12C shows an example of signal correction when the sign and magnitude relationship between the A image and the IMG image are A <0 <IMG. In FIG. 12C, the data of the A image is a negative value, and the data of the IMG image is a positive value. Therefore, the acquired A image and IMG image are represented as follows.
A = Na (Sa = 0, Sb = Simg)
IMG = Simg + Nimg

The B image generated by subtracting the A image data from the IMG image data is represented as follows.
B = IMG-A
= Simg + Nimg-Na
= Sb + Nimg-Na

In this case, in the first correction, since the data of the A image is a negative value, the first correction is performed on the B image. The corrected B image is represented as follows.
B = Sb + Nimg-Na + A
= Sb + Nimg-Na + Na
= Sb + Nimg

In the second correction, since the data of the A image is a negative value, that is, smaller than the first reference signal, the second correction is performed on the A image. The corrected A image is represented as follows.
A = 0

As described above, when A <0 <IMG, the component of the B image that may contain the component of the subject signal is not removed, and the noise Na of the A image is removed.

FIG. 13A shows an example of signal correction when the sign and magnitude relationship between the A image and the IMG image are IMG <0 <A. In FIG. 13A, since the data of the A image is a positive value and the data of the IMG image is a negative value, Simg = 0. Since the data of the IMG image is calculated by adding the data of the A image and the B image, when Simg = 0, Sa = Sb = 0. Therefore, the acquired A image and IMG image are represented as follows.
A = Na (Sa = 0)
IMG = Nimg (Simg = 0)

The B image generated by subtracting the A image data from the IMG image data is represented as follows.
B = IMG-A
= Nimg-Na

In this case, in the first correction, since the data of the B image is a negative value, the first correction is performed on the A image. The corrected A image is represented as follows.
A = Na + B
= Na + Nimg-Na
= Nimg

In the second correction, since each data of the A image and the B image is a negative value, that is, smaller than the reference signal, the second correction is performed on the A image and the B image. The corrected A image and B image are represented as follows.
A = 0
B = 0

As described above, when IMG <0 <A, neither the A image nor the B image may contain the component of the subject signal, and both have only the noise component. Therefore, both noise Na and Nimg are removed.

FIG. 13B shows an example of signal correction when the sign and magnitude relationship between the A image and the IMG image are A <IMG <0. In FIG. 13B, since the data of the A image and the IMG image are both negative values, Sa = Simg = 0. Therefore, the acquired A image and IMG image are represented as follows.
A = Na (Sa = 0)
IMG = Nimg (Simg = 0)

The B image generated by subtracting the A image data from the IMG image data is represented as follows.
B = IMG-A
= Nimg-Na

In this case, in the first correction, since the data of the A image is negative, the first correction is performed on the B image. The corrected B image is represented as follows.
B = Nimg-Na + A
= Nimg-Na + Na
= Nimg

In the second correction, since each data of the A image and the B image is a negative value, that is, smaller than the reference signal, the second correction is performed on the A image and the B image. The corrected A image and B image are represented as follows.
A = 0
B = 0

As described above, when A <IMG <0, neither the A image nor the B image may contain the component of the subject signal, and they have only the noise component. Therefore, both noise Na and Nimg are removed.

FIG. 13C shows an example of signal correction when the sign and magnitude relationship between the A image and the IMG image are IMG <A <0. In FIG. 13C, since the data of the A image and the IMG image are both negative values, Sa = Simg = 0. Therefore, the acquired A image and IMG image are represented as follows.
A = Na (Sa = 0)
IMG = Nimg (Simg = 0)

The B image generated by subtracting the A image data from the IMG image data is represented as follows.
B = IMG-A
= Nimg-Na

In this case, in the first correction, since the data of the A image and the B image are both negative values, the first correction is performed on the A image and the B image. The corrected A image and B image are represented as follows.
A = Na + B
= Na + (Nimg-Na)
= Nimg
B = Nimg-Na + A
= Nimg-Na + Na
= Nimg

In the second correction, since each data of the A image and the B image is a negative value, that is, smaller than the reference signal, the second correction is performed on the A image and the B image. The corrected A image and B image are represented as follows.
A = 0
B = 0

As described above, when IMG <A <0, neither the A image nor the B image may contain the component of the subject signal, and they have only the noise component. Therefore, both noise Na and Nimg are removed.

As described with reference to FIG. 10C, even when random noise and inverse correlation noise are superimposed on the focus detection signal, by executing the process shown in FIG. 9, the subject signals in the A image and the B image are displayed. Noise can be reduced without removing the components. Then, it is possible to calculate the correlation calculation result in which the influences of the minimum value and the maximum value shown in FIGS. 7 (B) and 8 (B) are reduced.

In the present embodiment, when the random noise and the inverse correlation noise generated in the A image and the B image, which are not related to the signal obtained from the subject image, are superimposed on the focus detection signal, the focus detection signal is corrected. As a result, it is possible to calculate a correlation value that is less affected by noise and detect an appropriate focus position (focus state). As a result, highly accurate focus detection with reduced influence of noise becomes possible.

[Second Embodiment]
Next, a second embodiment of the present invention will be described. This embodiment is different from the first embodiment in terms of the method of calculating the reference signal. Therefore, the differences from the first embodiment will be described in this embodiment, and detailed description thereof will be omitted for the same items as in the first embodiment.

The signal correction (S604 in FIG. 6) in the present embodiment will be described with reference to FIGS. 9 and 14. FIG. 14 is an explanatory diagram of signal correction in the present embodiment, showing pixel positions in the horizontal axis direction and signal strength in the vertical axis direction.

In the present embodiment, in S903 and S904 of FIG. 9, the first reference signal and the second reference signal used by the camera MPU 125 for correction are calculated, respectively. The first reference signal and the second reference signal shown in FIG. 14A are calculated as intermediate values of the pixel values of a plurality of pixels (peripheral pixels) located around the pixel of interest related to the first signal and the second signal. To. This intermediate value is an example, and the signal value of each reference signal may be calculated as the average value or the weighted average value of the pixel values of the peripheral pixels. Further, in the present embodiment, individual reference signals are used for the A image signal and the B image signal, but a common reference signal for each of the A image and the B image may be calculated and corrected. This embodiment is different from the first embodiment in which the reference signal is calculated from the black level in that the reference signal is calculated using the peripheral pixels of the pixel of interest regarding the A image signal and the B image signal which are the focus detection signals. different.

The camera MPU 125 in S905 calculates the first correction signal from the difference between the first signal acquired in S901 and the first reference signal acquired in S903. The camera MPU 125 in S906 calculates the second correction signal from the difference between the second signal acquired in S902 and the second reference signal acquired in S904. The first correction signal and the second correction signal shown in FIG. 14 (B) are the first reference signal and the first reference signal shown in FIG. 14 (A) from the focus detection signal (A image, B image) shown in FIG. 10 (C). It is a signal after subtracting each of the two reference signals.

In S908, the first correction is performed on the second signal by adding the first correction signal to the second signal, and in S910, the first correction is performed on the first signal by adding the second correction signal to the first signal. Correction is made. FIG. 14C shows the first signal (A image) and the second signal (B image) after the first correction. It can be seen that the positive noise component can be reduced as compared with the first signal (A image) and the second signal (B image) before correction shown in FIG. 10 (C).

In S912, the second correction is performed with respect to the first signal by using the first reference signal as the first signal. Further, in S914, the second correction is performed with respect to the second signal by using the second reference signal as the second signal. FIG. 14D shows the first signal (A image) and the second signal (B image) after the second correction. It can be seen that the negative noise component can be removed as compared with the first signal (A image) and the second signal (B image) after the first correction shown in FIG. 14 (C).

This embodiment is different from the first embodiment in that noise can be reduced with respect to a pixel signal containing a large amount of subject signal components. When random noise and inverse correlation noise are superimposed on the focus detection signal as shown in FIG. 10C, the first correction and the first correction are based on the negative value of the correction signal calculated from the focus detection signal and the reference signal. Noise can be reduced by performing the second correction. By performing the first correction and the second correction on the focus detection signal, it is possible to obtain a focus detection signal with reduced noise. It is possible to calculate the correlation calculation result in which the influences of the minimum value and the maximum value shown in FIGS. 7 (B) and 8 (B) are reduced.

104 Focus Lens 117 Lens MPU
122 Image sensor 125 Camera MPU
129 Phase-difference AF unit 211i On-chip microlens 211a, 211b Photoelectric conversion unit

Claims (13)

An image sensor in which a pixel unit having a plurality of photoelectric conversion units for receiving light flux passing through different pupil region regions of the image pickup optical system is arranged, and an image pickup device.
Acquisition to acquire the first signal acquired by any one of the plurality of photoelectric conversion units and the second signal obtained by subtracting the first signal from the signals obtained by the plurality of photoelectric conversion units. Means and
The first reference signal for the first signal and the second reference signal for the second signal are calculated, the first correction signal is calculated from the first signal and the first reference signal, and the second signal and the said. The first calculation means for calculating the second correction signal from the second reference signal, and
The second signal is corrected by the first correction signal or the second reference signal, the first signal is corrected by the second correction signal or the first reference signal, and the corrected first signal and the first signal are used. An image pickup apparatus comprising: a second calculation means for calculating an image shift amount or a defocus amount by calculating a correlation value with two signals. The first calculation means calculates the first correction signal from the difference between the first signal and the first reference signal, and the second correction signal is calculated from the difference between the second signal and the second reference signal. The image pickup apparatus according to claim 1, wherein the image is calculated. When the value of the first correction signal is smaller than the first threshold value, the second calculation means corrects the second signal by adding the first correction signal to the second signal, and the second calculation means. The first or second aspect of claim 1, wherein when the value of the correction signal is smaller than the second threshold value, the first signal is corrected by adding the second correction signal to the first signal. Imaging device. The second calculation means performs a first correction for correcting the second signal and the first signal by using the first correction signal and the second correction signal, and further performs the first correction signal and the first reference signal. 2. The image pickup apparatus according to any one of claims 1 to 3, wherein the second correction for correcting the first signal and the second signal is performed using the reference signal. In the second correction, when the first signal is smaller than the first reference signal, the second calculation means corrects the first signal by the first reference signal, and the second signal is the second. 2. The image pickup apparatus according to claim 4, wherein when the signal is smaller than the reference signal, the second signal is corrected by the second reference signal. The image pickup apparatus according to any one of claims 1 to 5, wherein the first reference signal and the second reference signal are signals of the same level. The image pickup apparatus according to claim 6, wherein the first reference signal and the second reference signal are black level signals. The first calculation means according to any one of claims 1 to 5, wherein the first reference signal and the second reference signal are calculated from the first signal and the second signal, respectively. The imaging device described. The signal value of the first reference signal or the second reference signal is an average value, a weighted average value, or an intermediate value calculated from the pixel values of the peripheral pixels of the pixel of interest related to the first signal or the second signal. The imaging device according to claim 8, wherein the image pickup device is characterized by the above. The image pickup apparatus according to any one of claims 1 to 9, further comprising a control means for adjusting the focus of the image pickup optical system using the defocus amount. The image pickup apparatus according to any one of claims 1 to 9, further comprising a generation means for generating image data having parallax from the corrected first signal and the second signal. The image pickup apparatus according to any one of claims 1 to 11, wherein the pixel unit of the image pickup element includes a microlens and a plurality of photoelectric conversion units corresponding to the microlens. It is a control method executed by an image pickup device including an image pickup device in which a pixel portion having a plurality of photoelectric conversion portions for receiving light flux passing through different pupil region regions of an image pickup optical system is arranged.
A step of acquiring a first signal acquired by one of the plurality of photoelectric conversion units and a second signal obtained by subtracting the first signal from the signals obtained by the plurality of photoelectric conversion units. When,
The first reference signal for the first signal and the second reference signal for the second signal are calculated, the first correction signal is calculated from the first signal and the first reference signal, and the second signal and the said. The process of calculating the second correction signal from the second reference signal and
The second signal is corrected by the first correction signal or the second reference signal, the first signal is corrected by the second correction signal or the first reference signal, and the corrected first signal and the first signal are used. (2) A control method for an image pickup apparatus, which comprises a step of calculating a correlation value with a signal and calculating an image shift amount or a defocus amount.

Images