JP2017223741A - Imaging device and control method thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an imaging device and its control method that can improve accuracy of a reliability determination in focus detection.SOLUTION: An imaging device is configured to acquire a pair of image signals varying in accordance with a focus state of an imaging optical system to implement focus detection. In phase difference AF (autofocus) processing, a correlation computation is implemented that uses the pair of image signals, and an amount of correlation and amount of image deviation are calculated. In the correlation computation, a first evaluation value in a first area (window area) where a correlation between the image signals is checked is calculated, and a second evaluation in a second area (viewing field area) broader than the first area is calculated. In focus control, a reliability determination with respect to an amount of image deviation between the pair of image signals is made by selecting the first evaluation value or increasing a ratio of the first evaluation value when the focus state is a focusing state or a state in the vicinity of focusing, or by selecting the second evaluation value or increasing a ratio of the second evaluation value when the focus state is not the focusing state.SELECTED DRAWING: Figure 9

Description

  The present invention relates to focus detection of an imaging apparatus such as a digital camera or a video camera.

  As an autofocus (AF) system of the imaging apparatus, there is a phase difference detection type focus adjustment (hereinafter referred to as phase difference AF). In the phase difference AF, the light beam that has passed through the exit pupil of the photographing lens is divided and each of the plurality of focus detection sensors receives light. By detecting the shift amount of each signal output by the plurality of focus detection sensors, that is, the relative positional shift amount in the beam splitting direction (hereinafter referred to as image shift amount), the focus shift amount ( Hereinafter, the defocus amount is calculated). The focus lens is driven and controlled based on the defocus amount. In addition, an image pickup apparatus equipped with an image pickup surface phase difference detection type image sensor does not require a dedicated focus detection sensor, and can realize phase difference AF at high speed. A pixel signal obtained by pupil division by a plurality of photoelectric conversion units constituting pixels of the image sensor is obtained. Focus detection is performed by individually processing a plurality of image signals based on the pixel signals, and the combined calculation power of the pixel signals is used as a captured image signal.

  Patent Document 1 discloses a technique for reducing the calculation time required for phase difference AF. Whether or not to calculate the image shift amount is determined from the contrast of the calculated image signal, and the final determination is performed by calculating the contrast again after the image shift amount is calculated. Patent Document 2 discloses processing for determining the reliability of the imaging surface phase difference AF based on the result of contrast calculated from the image signal in the imaging surface phase difference AF. When the defocus amount is large and the contrast value is large, it is determined that the reliability is low.

JP 2003-172870 A JP 2013-171257 A

In order to increase the number of AF frames (detection frames) in the image corresponding to the focus detection area, it is necessary to speed up the calculation of the phase difference AF. For this purpose, it is necessary to simultaneously calculate the image shift amount and the contrast by parallel calculation. However, in the configurations disclosed in Patent Document 1 and Patent Document 2, the contrast calculation range cannot be changed so as to be an appropriate range according to the image shift amount. That is, in parallel processing, the contrast calculation range cannot be set to an appropriate range according to the image shift amount after calculation, and therefore the accuracy of reliability determination using contrast may not be stable.
An object of the present invention is to provide an imaging apparatus capable of improving the accuracy of reliability determination in focus detection and a control method thereof.

  An imaging apparatus according to an embodiment of the present invention is an imaging apparatus that performs focus detection by acquiring a pair of image signals that change depending on a focus state of an imaging optical system, and performs a correlation calculation using the pair of image signals. Calculating a correlation amount and an image shift amount, calculating an evaluation value from each of a plurality of areas set for focus detection, and calculating an index representing the focus state, and using the evaluation value and the index Control means for determining the reliability of the image shift amount and controlling the focus adjustment using the image shift amount determined to be reliable. When the focus state is the first state, the calculation unit increases the ratio of the first evaluation value calculated in the first region among the plurality of regions, and the focus state is the second state. In the case of the state, the ratio of the second evaluation value calculated in the second region wider than the first region among the plurality of regions is increased.

  According to the imaging apparatus of the present invention, the accuracy of reliability determination in focus detection can be improved.

It is a block diagram which shows the structural example of the imaging device in the Example of this invention. It is a flowchart which shows the imaging operation of the imaging device in the Example of this invention. It is a figure explaining the image pick-up element in the Example of this invention. It is a figure explaining the pupil division of the photographic lens in the Example of the present invention. It is a figure which shows the focus detection area | region in the Example of this invention. It is a figure which shows typically the pixel arrangement | sequence of the focus detection area | region in the Example of this invention. It is a figure which illustrates the image signal and correlation amount waveform in the Example of this invention. It is a block diagram of a phase difference AF processing unit in an embodiment of the present invention. It is a figure which shows the detail of the focus detection area | region in the Example of this invention. It is a figure explaining the imaging optical system in the Example of this invention. It is a flowchart which shows the focus detection operation | movement in Example 1 of this invention. It is a figure which shows the ratio of the synthesis | combination according to the blur evaluation value in the Example of this invention. It is a flowchart which shows the focus adjustment operation | movement in the Example of this invention. It is a flowchart which shows the focus detection operation | movement in Example 2 of this invention. It is a flowchart which shows the reliability determination operation | movement in Example 3 of this invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the drawings, the same components are denoted by the same reference numerals, and redundant description is omitted.

[Example 1]
First, with reference to FIG. 1, an imaging apparatus according to a first embodiment of the present invention will be described. FIG. 1 is a block diagram illustrating a configuration example of the imaging apparatus 100. The imaging apparatus 100 is a video camera, a digital still camera, or the like, and can capture a subject and record moving image and still image data on various media such as a tape, a solid-state memory, an optical disk, and a magnetic disk. Each unit in the imaging apparatus 100 is connected via a bus 160 and is controlled by a main CPU (central processing unit) 151.

  The photographic lens unit 101 includes a first fixed lens group 102, a zoom lens 111, a diaphragm 103, a second fixed lens group 121, and a focus lens 131. The taking lens unit 101 is attached to the camera body or is provided on the camera body. The diaphragm control unit 105 adjusts the light amount at the time of shooting by adjusting the aperture diameter of the diaphragm 103 by driving the diaphragm 103 via the diaphragm motor 104 in accordance with a command from the main CPU 151. The zoom control unit 113 changes the focal length by driving the zoom lens 111 via the zoom motor 112. The focus control unit 133 controls the focus state (focus state) by driving the focus lens 131 via the focus motor 132. In FIG. 1, the focus lens 131 that is a focus adjustment lens is simply shown as a single lens, but it is usually composed of a plurality of lenses.

  The image sensor 141 is a pupil-division type image sensor including a plurality of photoelectric conversion units sharing one microlens. That is, phase difference focus detection is performed using the image signal output from the image sensor 141. A subject image formed on the image sensor 141 by the photographing lens unit 101 is photoelectrically converted into an electric signal by the image sensor 141. The image sensor 141 has a configuration in which m and n are variables of natural numbers and m pixels in the horizontal (horizontal) direction and n pixels in the vertical (vertical) direction are arranged in a two-dimensional array. As will be described later, the pixel portion corresponding to each microlens includes a plurality of photoelectric conversion elements. The electrical signal output by the imaging element 141 is processed by the imaging signal processing unit 142 and arranged as an image signal.

  The phase difference AF processing unit 135 acquires a plurality of image signals from the imaging signal processing unit 142. The plurality of image signals are, for example, image signals (signal values) output individually from the first and second photoelectric conversion elements constituting the pixel unit. Since the first and second photoelectric conversion elements each independently output a signal, it is possible to calculate the image shift amount in the division direction of the image obtained by dividing the light from the subject. That is, phase difference detection is performed based on signals obtained by the first and second photoelectric conversion elements respectively receiving and photoelectrically converting light beams that have passed through different pupil partial regions of the photographing lens unit 101. Further, the phase difference AF processing unit 135 calculates an evaluation value related to the phase difference AF when detecting the image shift amount. This evaluation value is calculated from the contrast values of a plurality of image signals. The phase difference AF processing unit 135 outputs the calculated image shift amount and evaluation value to the focus control unit 133.

  The focus control unit 133 calculates the defocus amount of the photographing lens unit 101 based on the image shift amount acquired from the phase difference AF processing unit 135. The defocus amount is calculated by multiplying the image shift amount by a predetermined coefficient (conversion coefficient). In addition, the focus control unit 133 performs reliability determination using the evaluation value acquired from the phase difference AF processing unit 135. The reliability determination is to determine whether the detected image shift amount or defocus amount is reliable. The focus control unit 133 determines the drive amount of the focus motor 132 based on the defocus amount and the reliability determination result. The focus control unit 133 performs drive control of the focus motor 132 according to the drive amount. AF control is realized by movement control of the focus lens 131. Note that the focus control unit 133 performs each process of calculating the defocus amount and determining the reliability based on a command from the main CPU 151. The main CPU 151 may be configured to execute at least a part of these processes.

  Image data output from the imaging signal processing unit 142 is sent to the imaging control unit 143 and is temporarily stored in a RAM (random access memory) 154. The imaging control unit 143 controls the imaging element 141 according to a command from the main CPU 151. The accumulation time of the image sensor 141, the setting of a gain value when outputting from the image sensor 141 to the image signal processor 142, and the like are performed, and the drive of the image sensor 141 is controlled.

  The image compression / decompression unit 153 performs a process of reading and compressing the image data stored in the RAM 154 and then recording it on the image recording medium 157. In parallel with this, the image data read from the RAM 154 is sent to the image processing unit 152. The image processing unit 152 performs processing of captured image data. The captured image data is image data based on an image signal obtained from an addition signal of the first and second photoelectric conversion elements that constitute each pixel portion of the image sensor 141. In this case, there is an embodiment in which captured image data is acquired based on a signal obtained by adding signals output from the first photoelectric conversion element and the second photoelectric conversion element. There is an embodiment in which captured image data is acquired based on a signal acquired by photoelectric conversion from a light beam received in a region including the first and second photoelectric conversion elements. The image processing unit 152 performs, for example, reduction or enlargement processing on the image data to an optimum size. The image data processed to the optimum size is sent to the monitor display 150 and displayed on the screen. Thereby, the operator can observe a captured image in real time. Further, immediately after shooting, the monitor display 150 displays the shot image for a predetermined time, so that the operator can check the shot image.

  The operation unit 156 includes input devices such as operation switches and a touch panel. The operator instructs the imaging apparatus 100 using the operation unit 156, and the main CPU 151 receives an operation instruction signal input from the operation unit 156. The power management unit 158 appropriately manages the battery 159 and stably supplies power to the entire imaging apparatus 100. The flash memory 155 stores a control program necessary for the operation of the imaging apparatus 100. When the imaging apparatus 100 is activated by shifting from the power-off state to the power-on state according to a user operation instruction, the control program stored in the flash memory 155 is loaded and stored in a part of the RAM 154. The main CPU 151 controls the operation of the imaging apparatus 100 according to a control program loaded in the RAM 154.

  Next, a shooting operation including the focus adjustment control of the imaging apparatus 100 will be described with reference to the flowchart of FIG. Processing of each step shown in FIG. 2 is performed based on a command according to a control program executed by the main CPU 151.

  First, in step S201, when the power of the imaging apparatus 100 is turned on, the main CPU 151 starts control. Subsequently, in step S202, the flags and control variables of the imaging apparatus 100 are initialized, and in step S203, the main CPU 151 performs control to move the imaging optical member such as the focus lens 131 to the initial position.

  Next, in step S204, the main CPU 151 determines whether or not a power-off operation has been performed by the operator. If it is determined whether or not the power OFF operation has been performed and it is determined that the power OFF operation has been performed, the process proceeds to step S205. If the power OFF operation has not been performed, the process proceeds to step S207. In step S205, the main CPU 151 executes post-processing for turning off the power of the imaging apparatus 100. In the post-processing, after the imaging optical member is moved to the initial position and processing such as clearing of various flags and control variables is performed, the process proceeds to step S206, and the control of the imaging operation of the imaging device 100 is terminated.

  In step S207, the main CPU 151 performs focus detection processing. In the next step S208, the focus control unit 133 controls the movement of the focus lens 131 according to the driving direction, speed, and position of the focus lens 131 determined in step S207. Focus adjustment control is performed by moving the focus lens 131 to the target position.

  Subsequently, in step S209, the image sensor 141 is exposed, and the image sensor 141 photoelectrically converts the subject image to generate an imaging signal (imaging processing). The imaging signal processing unit 142 performs predetermined image processing on the imaging signal generated by the photoelectric conversion and outputs an image signal. In step S210, the main CPU 151 determines whether or not the operator has pressed a recording button included in the operation unit 156. As a result of the determination, if a recording instruction is issued, the process proceeds to step S211. If there is no recording instruction, the process returns to step S204 to continue the process.

  In step S211, an image recording process is performed. The image data is recorded on the image recording medium 157 after the image compression / decompression unit 153 compresses the image signal. And it returns to step S204 and repeats each above-mentioned step.

  Next, the imaging surface phase difference detection method in the present embodiment will be described. First, a configuration example of the image sensor 141 will be described with reference to FIG. FIG. 3A is a schematic cross-sectional view illustrating the pixel configuration of the imaging element 141 having a pupil division function. The photoelectric conversion element 30 in each pixel unit is divided into a first photoelectric conversion element 30-1 and a second photoelectric conversion element 30-2.

  The microlens 31 is an on-chip microlens and has a function of efficiently condensing the photoelectric conversion element 30. The two photoelectric conversion elements 30-1 and 30-2 are arranged such that the optical axis is aligned with the boundary between them. Each pixel portion includes a planarizing film 32, a color filter 33, a wiring 34, and an interlayer insulating film 35.

  FIG. 3B is a plan view partially showing an array of pixel portions included in the image sensor 141. The imaging element 141 is formed as a structure in which a plurality of pixel portions illustrated in FIG. In order to capture an image, R (red), G (green), and B (blue) color filters 33 are alternately arranged in each pixel portion, and a set of pixel blocks 40, 41, and 42 is formed by four pixels. . This is a so-called Bayer array. In FIG. 3B, “1” added below each symbol of R, G, and B corresponds to the first photoelectric conversion element 30-1, and “2” represents the second photoelectric conversion element 30. -2.

  FIG. 3C is an optical principle diagram of the image sensor 141. FIG. 3C schematically shows part of a cross-sectional view obtained by cutting along line AA in FIG. The image sensor 141 is disposed on the planned image plane of the photographic lens unit 101. Due to the action of the microlens 31, the photoelectric conversion elements 30-1 and 30-2 each receive a pair of light beams that have passed through different positions (pupil partial areas) of the exit pupil of the photographing lens unit 101. The first photoelectric conversion element 30-1 mainly receives a light beam that passes through the right region (pupil partial region) in FIG. 3C among the pupils of the photographing lens unit 101. On the other hand, the second photoelectric conversion element 30-2 mainly receives a light beam that passes through the left region (pupil partial region) in FIG.

  The pupil of the photographic lens unit 101 will be described with reference to FIG. FIG. 4 is a schematic diagram showing the pupil 50 of the photographic lens unit 101 when viewed from the image sensor 141. The right region 51-1 is a sensitivity region (hereinafter referred to as “A image pupil”) of the first photoelectric conversion element 30-1. The left region 51-2 is a sensitivity region (hereinafter referred to as “B image pupil”) of the second photoelectric conversion element 30-2. The centroids 52-1 and 52-2 indicate the centroid positions of the A image pupil and the B image pupil, respectively.

  In the imaging process of this embodiment, it is possible to generate a captured image signal by adding the outputs of two photoelectric conversion elements in which color filters of the same color are arranged in the same pixel. On the other hand, in the focus detection process of this embodiment, the focus detection signal of one pixel is acquired by integrating the outputs of the first photoelectric conversion elements 30-1 in one pixel block. By acquiring this focus detection signal continuously in the horizontal (horizontal) direction as in the pixel blocks 40, 41, and 42, an A image signal can be generated. Similarly, the focus detection signal of one pixel is acquired by integrating the outputs of the second photoelectric conversion elements 30-2 in one pixel block. A B image signal can be generated by continuously acquiring this focus detection signal for pixel blocks in the horizontal direction. A pair of phase difference detection signals is generated from the A image signal and the B image signal.

  An area (focus detection area) set for focus detection will be described with reference to FIG. FIG. 5 illustrates the focus detection area used in this embodiment. A focus detection area 61 is set at an appropriate position with respect to the imaging angle of view 60 indicated by a rectangular frame. For example, in the focus detection region 61, the phase difference AF processing unit 135 generates a pair of phase difference detection signals and performs focus detection. Note that it is possible to set a plurality of focus detection areas at the imaging angle of view 60. In the present embodiment, the method of generating the phase difference detection signal corresponding to the focus detection area has been described in which all the pixel units constituting the imaging element 141 have two photoelectric conversion elements, but the present invention is not limited thereto. It is not something. For example, the image sensor 141 having the structure (divided pixel structure) shown in FIG. 3A may be used only in the focus detection region.

  With reference to FIG. 6, the pixel blocks 40, 41, and 42 in the Bayer array will be described. In this embodiment, a signal obtained by adding luminance lines in an appropriate range in the vertical direction of FIG. 6 is used. In the pixel blocks 40, 41, and 42 shown in FIG. 6, an image signal for focus detection is generated by averaging the signals using the two vertical pixel blocks in the Bayer array. However, the number of luminance lines added can be arbitrarily set. The greater the number of luminance lines added, the larger the focus detection area 61 in FIG.

  Next, a method of calculating the image shift amount from the A image signal and the B image signal (hereinafter collectively referred to as “image signal”) will be described with reference to FIG. FIG. 7A is a diagram for explaining an image signal. The vertical axis indicates the level of the image signal, and the horizontal axis indicates the pixel position. The A image signal is indicated by a solid line graph, and the B image signal is indicated by a broken line graph. When the image shift amount of the generated pair of phase difference detection signals is expressed as X, the image shift amount X changes according to the imaging state in the photographing lens unit 101. The imaging state is a focused state, a front pin state, or a rear pin state.

  When the photographing lens unit 101 is in focus, the image shift amount between the two image signals is eliminated. On the other hand, in the front pin state or the rear pin state, an image shift amount occurs in different directions between the two image signals. The image shift amount has a certain relationship with the distance between the position where the light from the subject is focused by the photographing lens unit 101 and the upper surface of the microlens, so-called defocus amount. Therefore, the phase difference AF processing unit 135 performs a correlation operation on the two image signals in order to calculate the image shift amount X. In this correlation calculation, the correlation amount between the two image signals is calculated while shifting the pixels, and the difference between the positions having the highest correlation (or the positions having the lowest difference) is calculated as the image shift amount.

  With reference to FIG. 7B, the relationship between the pixel shift amount and the correlation amount will be described. FIG. 7B is a diagram illustrating a waveform (hereinafter referred to as a correlation amount waveform) in which the correlation amounts of two image signals are arranged at the shift position when the pixels of the image signal are shifted. In FIG. 7B, the horizontal axis indicates the pixel shift position (denoted i). The vertical axis represents the correlation amount (denoted as COR (i)) between the A image signal and the B image signal at the shift position i. When calculating the correlation amount, two image signals are overlapped, the corresponding signals are compared with each other, and a cumulative value of the difference is acquired. There are also embodiments in which the cumulative value of the larger value is acquired, or in which the cumulative value of the smaller value is acquired. The accumulated value is an index indicating the correlation. In the embodiment for acquiring the cumulative value of the difference or the embodiment for acquiring the cumulative value of the larger value, the time when the cumulative value is the smallest is when the correlation is high. In the embodiment in which the cumulative value of the smaller value is acquired, the time when the cumulative value is the largest is when the correlation is high.

  The phase difference AF processing unit 135 calculates an image shift amount X and calculates an evaluation value used in reliability determination described later from the image signal and the correlation amount waveform. Hereinafter, a process of calculating an evaluation value for reliability determination when acquiring a cumulative value of differences between two image signals will be described.

  The reliability can be defined by the size of the contrast of the subject, the degree of coincidence between the two image signals, or the amount of change in the correlation amount when the two images most closely match. For example, a peak bottom value that is a difference between the maximum value (peak value) and the minimum value (bottom value) can be calculated from the acquired image signal as an index representing the contrast level of the subject. The larger the peak bottom value, the higher the reliability.

  As indicated by a range 70 in FIG. 7B, when the correlation between the two image signals is the highest, the correlation amount becomes a minimum value. The magnitude of the correlation amount at this time is a difference value between the two image signals when the correlation between the two image signals is the highest. When the degree of coincidence of two image signals is expressed as F, the smaller the value of the degree of coincidence F, the higher the reliability.

FIG. 7C is a diagram showing a waveform in which correlated displacement amounts are arranged. The horizontal axis represents the pixel shift position i, and the vertical axis represents the correlation displacement amount (hereinafter referred to as ΔCOR (i)). In the correlation amount waveform shown in FIG. 7B, ΔCOR is calculated as shown in the following equation (1) from the difference in the correlation amount of skipping one shift, for example.
ΔCOR = COR (i−1) −COR (i + 1) Formula (1)

A range 71 illustrated in FIG. 7C corresponds to a range 70 illustrated in FIG. That is, when the correlation between the two image signals is the highest, the correlation amount COR (i) is minimal. Corresponding correlation displacement amount ΔCOR (i) has a portion that changes from plus to minus in range 71. A shift position where the value of the correlation displacement amount ΔCOR (i) is 0 is denoted as “j”. The shift position j corresponds to a position where the correlation amount becomes a minimum value, and the change amount of the correlation displacement amount in the vicinity of the shift position is defined as a correlation change amount (denoted as M) when the two images most closely match. That is, the M value is calculated from the following formula (2).
M = | ΔCOR (j−1) | + | ΔCOR (j) | Equation (2)

  The correlation change amount M is the sum of absolute values of the respective correlation displacement amounts at two adjacent shift positions, and the larger the M value, the higher the reliability. In this embodiment, the M value is exemplified as the evaluation value for reliability determination, but the evaluation value for reliability determination may be calculated by another known method.

  A process performed by the phase difference AF processing unit 135 will be described with reference to FIG. FIG. 8 is a block diagram illustrating a configuration example of the phase difference AF processing unit 135. The luminance line addition unit 81 acquires a signal from the imaging signal processing unit 142 and performs luminance line addition. Thereafter, the correction processing unit 82 corrects the image signal. The filter processing unit 83 performs filter processing on the corrected signal and outputs the result to the correlation calculation processing unit 84. The correlation calculation processing unit 84 performs a correlation calculation on the signal after the filter processing, and calculates an image shift amount. The correlation amount waveform and the image shift amount are output to the focus control unit 133.

  Saturation detectors 85 and 86 obtain signals from the luminance line adder 81 and detect pixel saturation. That is, the saturation detection units 85 and 86 count the number of saturated pixel signals among the signals input from the imaging signal processing unit 142 and output the number of saturated pixels. The saturation detection unit 85 is a window region saturation detection unit, and the saturation detection unit 86 is a visual field region saturation detection unit. Details of each detection unit will be described later.

  Contrast value calculation units 87 and 88 acquire signals output from the correction processing unit 82, and calculate and output an evaluation value related to contrast according to an evaluation value calculation method. The contrast value calculator 87 is a window area contrast value calculator, and the contrast value calculator 88 is a visual field area contrast value calculator. Details of each calculation unit will be described later.

  FIG. 9 is a diagram showing in detail the focus detection area 61 shown in FIG. As regions in the correlation calculation, a first region (window region) and a second region (viewing region) are shown. The window area is an area having a size for confirming the coincidence and correlation between the A image signal and the B image signal. The visual field area is a shift area of the A image signal and the B image signal in the correlation calculation, and is an area wider than the window area. FIG. 9 illustrates a case where the window area is located at the center of the visual field area.

  The phase difference AF processing unit 135 is configured by hardware in order to shorten the calculation time. Therefore, the image shift amount calculation process and the contrast value calculation process are executed at the same time, and therefore the contrast value calculation range cannot be changed according to the image shift amount. That is, saturation detection and contrast value calculation are performed in the window region and the visual field region. The window area saturation detection unit 85 performs pixel saturation detection for the window area. The window area contrast value calculation unit 87 calculates a contrast value for the window area. On the other hand, the visual field region saturation detection unit 86 performs pixel saturation detection for the visual field region. The visual field region contrast value calculation unit 88 calculates a contrast value for the visual field region.

  With reference to FIG. 10, the conversion from the image shift amount calculated by the correlation calculation to the defocus amount will be described. The defocus amount is calculated by the focus control unit 133. FIG. 10 is a schematic diagram showing an image pickup optical system including the taking lens unit 101 and the image pickup element 141. Light from the subject 90 passes through the taking lens unit 101 and forms an image. The position p1 of the focus detection surface is shown on the optical axis at the position p0 of the planned imaging plane. The relationship between the image shift amount and the defocus amount is determined according to the imaging optical system. The defocus amount can be calculated by multiplying the image shift amount X by a predetermined conversion coefficient K. The conversion coefficient K is calculated based on the gravity center positions of the A image pupil and the B image pupil.

Assume that the position p1 of the focus detection surface is moved to the position p2. Positions p1, q2, and q3 are on the focus detection surface before movement, and positions p2, q2 ^* , and q3 ^* are on the focus detection surface after movement. The amount of image shift changes according to the similarity between the triangle having the positions p0, q2, and q3 as vertex positions and the triangle having the positions p0, q2 ^* , and q3 ^* as vertex positions. Therefore, the defocus amount at the position p2 on the focus detection surface can be calculated. Based on the defocus amount, the main CPU 151 calculates the position of the focus lens 131 for obtaining a focused state with respect to the subject.

  The focus detection operation in the present embodiment will be described with reference to the flowchart of FIG. The processing of each step in FIG. 11 is performed by the main CPU 151 executing a control program to control each unit, and corresponds to step S207 in FIG.

  First, in step S1101, focus detection starts. In step S1102, in order to acquire an image signal used for focus detection, the imaging element 141 images a subject and outputs an imaging signal. The main CPU 151 and the imaging control unit 143 control the imaging element 141 to perform charge accumulation (exposure) according to a predetermined accumulation time, and read out the pixel value of the image signal in the focus detection area.

  In step S1103, the phase difference AF processing unit 135 performs a correlation operation on the acquired image signal. The phase difference AF processing unit 135 calculates a shift amount that maximizes the correlation from the calculated correlation amount. The phase difference AF processing unit 135 calculates an image shift amount and an evaluation value such as a contrast value. In step S1104, the main CPU 151 and the focus control unit 133 acquire the image shift amount, the contrast value in the window area, and the contrast value in the viewing area output from the phase difference AF processing unit 135.

  In step S1105, a blur evaluation value is calculated. The blur evaluation value is an index representing the focus state of the imaging optical system. For example, the blur evaluation value is obtained by dividing the correlation change amount M when two images are the best match by the peak bottom value (difference value between the maximum value and the minimum value) obtained from the luminance waveform of the image signal. . Alternatively, instead of the correlation change amount M when the two images most closely match, the absolute sum of the adjacent difference of the correlation amount waveform or the square sum of the adjacent difference of the correlation amount waveform may be used. Each of these is an index representing the steepness of the correlation amount waveform, and the value is larger and the blur is larger as the focus state is near the in-focus state (the state where the defocus amount is equal to or less than a predetermined threshold). The smaller the value. Since the correlation amount waveform is affected by the level of the image signal, the influence of the level of the image signal can be reduced by normalizing the correlation change amount M by dividing it by the peak bottom value. The smaller the blur evaluation value, the larger the shooting scene, and the larger the blur evaluation value, the higher the degree of focus.

  Also, the degree of coincidence F of the two image signals may be used as the blur evaluation value. The degree of coincidence F is an index representing the degree of correlation between image signals. The value of the degree of coincidence F is smaller as the focus state is closer to the in-focus state, and the value is calculated larger as the blur increases. Therefore, the degree of blur can be determined by the value of the matching degree F. However, since the degree of coincidence F is also affected by the level of the image signal, normalization is performed by dividing by the peak bottom value. In this case, the larger the blur evaluation value is, the larger the shooting scene is, and the smaller the blur evaluation value is, the higher the degree of focus is.

  In addition, as the blur evaluation value, an image shift amount or a defocus amount that is a current or previous focus detection result may be used. These values are smaller as the focus state is closer to the in-focus state, and the larger the blur, the larger the value is calculated. Therefore, the image shift amount and the defocus amount itself can be used as the blur evaluation value. In this case, the larger the blur evaluation value, the larger the shooting scene is, and the smaller the blur evaluation value, the higher the degree of focus. Note that a blur evaluation value obtained by combining a plurality of indices can be defined.

  In step S1106, the main CPU 151 calculates a coefficient (denoted as α) used for reliability determination. The coefficient α indicates a ratio when performing weighted addition of contrast values in the window area and the visual field area from the blur evaluation value calculated in step S1105. A specific example will be described with reference to FIG. The horizontal axis in FIG. 12 represents the blur evaluation value, and the vertical axis represents the coefficient α. For example, the coefficient α is calculated based on the blur evaluation value. When the coefficient α exceeds the threshold value, the value increases to reach 1. Alternatively, the coefficient value may not be a continuous value. In this case, the coefficient α is set to 1 or 0 without taking the intermediate value shown in FIG. 12, and is switched according to the blur evaluation value.

In step S1107, a contrast determination (denoted as Cont) calculation for reliability determination is performed using the following equation (3).
Cont = α × contrast value (window region) + (1−α) × contrast value (viewing region) Equation (3)

  Cont is calculated by weighted addition using a composition ratio corresponding to the value of the coefficient α with respect to the contrast values in the window region and the visual field region. As a result, when the focus state is regarded as the in-focus state or near-in-focus state, the ratio of the contrast value in the window area increases, and when the blur is large, the contrast value including the visual field area is included. Is used. Alternatively, when the focus state is regarded as a focused state or a state near the focused state, the value of the coefficient α is 1, and the contrast value in the window region is calculated. When the subject is not in focus, the value of the coefficient α is 0, and the contrast value in the visual field region is calculated. In this case, one of the contrast value in the window area and the contrast value in the visual field area is selected as the evaluation value.

  In step S1108, the focus control unit 133 performs reliability determination using the contrast value Cont calculated in step S1107. In the reliability determination in step S1109, it is determined whether or not the calculated image shift amount X is reliable based on the contrast value and the matching degree F between the two images. The contrast value Cont is compared with a predetermined threshold (denoted as CTh), and it is determined whether or not Cont> CTh. If Cont> CTh, it is determined that the calculated image shift amount X is reliable, and the process proceeds to step S1110. If the contrast value Cont is equal to or less than the threshold value, the process proceeds to S1111.

  The reliability determination in step S1109 includes a reliability determination regarding the degree of coincidence F of the two image signals in addition to whether or not the contrast value of the image signal is sufficiently large. The main CPU 151 and the focus control unit 133 determine whether or not the value of the coincidence F between the two image signals is smaller than a predetermined threshold value (denoted as FTh). When F <FTh, it is determined that the calculated image shift amount X is reliable. On the other hand, if the value of the matching degree F is equal to or greater than the threshold value FTh, it is determined that the reliability of the image shift amount X is low, and the process proceeds to S1111. In addition, the reliability determination may include a determination as to whether or not the subject has a possibility that sufficient accuracy cannot be obtained by the correlation calculation from the image signal or the correlation amount waveform.

  If a reliable image shift amount X is obtained based on the result of the reliability determination in step S1109, the main CPU 151 and the focus control unit 133 multiply the calculated image shift amount X by the conversion coefficient K in step S1110. Then, the defocus amount Def is calculated. That is, the defocus amount Def is calculated by the relational expression Def = K × X. In calculating the defocus amount, the photographic lens unit 101 may perform post-correction processing to obtain a state where the subject is more focused. Then, the process proceeds to step S1112 to end the focus detection, shift to return processing, and return to the main routine that is the caller. On the other hand, in step S1111, since it is determined that a reliable image shift amount could not be detected, phase difference AF is not performed (focus detection NG). Then, control goes to a step S1112.

  Next, the focus control operation (focus adjustment operation) will be described with reference to FIG. FIG. 13 is a flowchart showing the focus control operation of this embodiment. The processing of each step shown in FIG. 13 corresponds to step S208 in FIG. 2, and is executed by the main CPU 151 and the focus control unit 133. The main CPU 151 performs a predetermined calculation according to the control program, and the focus control unit 133 controls the focus motor 132 based on a command from the main CPU 151.

  In step S1301, the focus control operation starts. In step S1302, the focus control unit 133 acquires a reliable defocus amount calculated by the focus detection in FIG. In step S1303, the focus control unit 133 calculates a drive amount (lens drive amount) of the focus lens 131 based on the acquired defocus amount. The lens driving amount calculation processing includes calculation of the lens driving direction and speed. Step S1304 is processing for determining whether or not the absolute value of the defocus amount is equal to or less than a predetermined value (threshold value). If the absolute value of the defocus amount is not less than the predetermined value in step S1304, the process proceeds to step S1305. If the absolute value of the defocus amount is less than the predetermined value, the process proceeds to step S1306.

  In step S1305, since the current position of the focus lens 131 is not considered to be the in-focus position (focus point), the focus lens 131 is driven according to the lens driving amount calculated in step S1303, and the process proceeds to step S1307. On the other hand, in step S1306, since the current position of the focus lens is considered to be in focus, the drive of the focus lens is stopped, and the process proceeds to step S1307. By the return processing in step S1307, the process returns to the main routine that is the caller. Thereafter, focus detection is performed according to the flowchart shown in FIG. 2, and when the defocus amount again exceeds a predetermined value, the focus lens 131 is driven. As illustrated in FIG. 2, the imaging apparatus 100 repeats the processes illustrated in FIGS. 11 and 13 until the power is turned off. Multiple focus detections are performed until the subject is focused.

  In the present embodiment, a contrast value for reliability determination is calculated using a composition ratio corresponding to the blur evaluation value. When the focus state is regarded as the first state, the contrast value in the window region is mainly used. The first state is a focused state where the subject is focused or a state in the vicinity of the focused state. When the focus state is regarded as the second state, a contrast value including the visual field region is used. The second state is a state where the subject is not in focus and the subject image is blurred. When the blur is large, the A image signal and the B image signal are greatly shifted, and the shift amount is large. When the focus state is an in-focus state or a state near the in-focus state, both the A image signal and the B image signal match in the vicinity of the window region. For this reason, the contrast value can be calculated within a range where the coincidence of the A image signal and the B image signal is expected. Since the range for calculating the contrast value is appropriately set and the reliability determination is performed, it is possible to improve the accuracy of the focus detection reliability determination. For example, when the blur is large, the contrast value in the window region becomes small, so that there is a high possibility that the focus adjustment operation is not performed in the conventional apparatus. On the other hand, in this embodiment, determination using a contrast value for reliability determination calculated from a contrast value not only in the window region but also in a visual field region wider than the region can be performed.

[Example 2]
Next, a second embodiment of the present invention will be described. In the first embodiment, the method of calculating the contrast value for reliability determination from the contrast values in the window region and the visual field region using the combination ratio according to the blur evaluation value has been described. In this embodiment, a method for calculating the number of saturation detection pixels for reliability determination from the number of saturation detection pixels in the window region and the visual field region using a combination ratio according to the blur evaluation value will be described. Since the image shift amount calculated using the saturated image signal is not reliable, it is necessary to perform reliability determination for saturation. In the present embodiment, the same components as those in the first embodiment will not be described in detail by using the reference numerals already used, and differences will be mainly described. Such a description is omitted in the embodiments described later.

  The focus detection operation in the present embodiment will be described with reference to the flowchart shown in FIG. Since the processing of steps S1401 to S1403 is the same as the processing of steps S1101 to S1103 in FIG. 11, the description thereof is omitted.

In step S1404, the main CPU 151 and the focus control unit 133 obtain the image shift amount X output from the phase difference AF processing unit 135, the saturation detection pixel number in the window region, and the saturation detection pixel number in the visual field region. Subsequent processes in steps S1405 and S1406 are the same as steps S1105 and S1106 in FIG.
In step S1407, the saturation detection pixel number (denoted as Sat) used for reliability determination is calculated from the following equation (4).
Sat = α × saturation detection pixel number (window region) + (1−α) × saturation detection pixel number (field of view region) Equation (4)

  The saturation detection pixel number Sat is calculated by weighted addition using a composition ratio corresponding to the value of the coefficient α with respect to the saturation detection pixel numbers in the window region and the visual field region. As a result, when the focus state is regarded as the in-focus state or the state near the in-focus state, the ratio of the saturation detection pixel number in the window region increases, and when the blur is large, the visual field region is included. The saturation detection pixel number is used.

  In step S1408, the focus control unit 133 performs reliability determination using the saturation detection pixel number Sat calculated in step S1407. Based on the number of saturation detection pixels, a process for determining whether or not the image shift amount X is reliable is performed. In that case, you may combine with the reliability determination by the contrast value demonstrated in Example 1. FIG. The saturation detection pixel number Sat is compared with a predetermined threshold (denoted as STh). If Sat <STh in step S1409, it is determined that the calculated image shift amount X is reliable, and the process proceeds to S1410, where the defocus amount is calculated. On the other hand, if Sat is greater than or equal to the threshold value STh, the process proceeds to S1411 and the phase difference AF is not performed. Since the processing from step S1410 to S1412 is the same as the processing from step S1110 to step S1112 in FIG. 11, the description thereof is omitted.

  In this embodiment, when the focus state is regarded as the in-focus state or the state near the in-focus state, the saturation detection pixel is mainly used in the window region, and when the blur is large, the saturation detection pixel including the visual field region is used. The reliability is determined using the number. Therefore, the saturation detection pixel number can be calculated within a range where the coincidence of the A image signal and the B image signal is expected. Since the range for calculating the saturation detection pixel number is appropriately set and the reliability determination is performed, it is possible to improve the accuracy of the focus detection reliability determination.

[Example 3]
Next, Embodiment 3 of the present invention will be described. In the present embodiment, a method for performing reliability determination regarding saturation from the number of saturation detection pixels in the window region and the contrast values in the window region and the visual field region will be described.

  With reference to the flowchart shown in FIG. 15, the reliability determination of the focus detection in a present Example is demonstrated. The processing of each step shown in FIG. 15 is executed by the main CPU 151 and the focus control unit 133.

  In step S1501, reliability determination starts. In step S1502, the main CPU 151 and the focus control unit 133 acquire the image shift amount, the number of saturation detection pixels in the window region, and the contrast values in the window region and the visual field region from the phase difference AF processing unit 135. The contrast value here uses a peak value. In step S1503, the main CPU 151 compares the number of saturation detection pixels in the window area with a predetermined value. The predetermined value (threshold value) may be a fixed value such as 20 pixels. If the number of saturation detection pixels in the window area is less than the predetermined value, the process proceeds to step S1504, and if it is greater than the predetermined value, the process proceeds to step S1508.

  Step S1504 is a determination process performed by comparing the absolute value of the image shift amount X with a predetermined value (threshold value). If the absolute value of the image shift amount X is less than the predetermined value, the process proceeds to step S1507, and the defocus amount is calculated. If the absolute value of the image shift amount X is greater than or equal to a predetermined value, the process proceeds to step S1505.

  In step S1505, the peak values in the visual field area and the window area are compared. In the present embodiment, a determination process is performed as to whether the evaluation value in the visual field area is greater than the evaluation value in the window area or whether the evaluation value in the window area is greater than or equal to the evaluation value in the visual field area. All of the evaluation values to be compared are contrast peak values. If the peak value in the visual field area is larger than the peak value in the window area, the process proceeds to step S1506. If the peak value in the visual field area is equal to or less than the peak value in the window area, the process proceeds to step S1507. In step S1506, the peak value in the visual field region is compared with a predetermined value. This predetermined value is a threshold value set according to the saturation luminance. If the peak value in the visual field area is greater than or equal to the threshold value, the process proceeds to step S1508, and if the peak value in the visual field area is less than the threshold value, the process proceeds to step S1507. Since the processes in steps S1507, S1508, and S1509 are the same as the processes in steps S1110, S1111, and S1112 in FIG. 11, their descriptions are omitted.

  In the present embodiment, when the focus state is regarded as a focused state or a state in the vicinity of the focused state, reliability determination is performed based on the number of saturation detection pixels in the window region. When it is considered that the focus state is largely blurred from the amount of image shift, not only the saturation detection pixel count in the window area but also the saturation detection reliability judgment using the contrast values in the window area and the visual field area Is done. Since the saturation detection range is appropriately set and the reliability determination is performed, it is possible to improve the accuracy of the focus detection reliability determination.

  According to each embodiment, the evaluation value of the window area is used when the focus state is regarded as the in-focus state or the near-in-focus state, and the evaluation value including the visual field area is used when the blur is large. The reliability judgment that was possible is possible. When the focus state is expected to be largely blurred, the accuracy of reliability determination in the phase difference AF can be improved by changing the calculation range of the evaluation value related to the phase difference AF. In each embodiment, the method for calculating the evaluation value for reliability determination from the respective evaluation values in the window region and the visual field region has been described. However, the number of divisions of the region is not limited to two. The number of divisions can be set to 3 or more. For example, the visual field region is divided into a window region, a right region and a left region excluding the window region, and an evaluation value for reliability determination is calculated by a combination thereof. When the blur is large, the evaluation value of the A image signal can be calculated from the evaluation values of the right region and the window region, and the evaluation value of the B image signal can be calculated from the evaluation values of the left region and the window region.

  As mentioned above, although preferable embodiment of this invention was described, this invention is not limited to these embodiment, A various deformation | transformation and change are possible within the range of the summary.

DESCRIPTION OF SYMBOLS 30 Photoelectric conversion element 31 Micro lens 84 Correlation calculation process part 85 Window area | region saturation detection part 86 Field of view area | region saturation detection part 87 Window area | region contrast value calculation part 88 Field of view area contrast value calculation part 131 Focus lens 133 Focus control part 135 Phase difference AF process Part 141 Image sensor 151 CPU

Claims (17)

An imaging apparatus that performs focus detection by acquiring a pair of image signals that change depending on a focus state of an imaging optical system,
The correlation calculation is performed using the pair of image signals, the correlation amount and the image shift amount are calculated, and evaluation values are calculated from a plurality of areas set for focus detection, and an index representing the focus state is calculated. Calculating means for
Control means for determining the reliability of the image shift amount based on the evaluation value and the index, and performing focus adjustment control using the image shift amount determined to be reliable,
When the focus state is the first state, the calculation unit increases the ratio of the first evaluation value calculated in the first region among the plurality of regions, and the focus state is the second state. In the state, the ratio of the second evaluation value calculated in a second area wider than the first area among the plurality of areas is increased. An imaging apparatus that performs focus detection by acquiring a pair of image signals that change depending on a focus state of an imaging optical system,
Calculation means for performing correlation calculation using the pair of image signals, calculating a correlation amount and an image shift amount, and calculating an evaluation value from each of a plurality of regions set for focus detection;
Control means for determining the reliability of the image shift amount based on the evaluation value, and performing focus adjustment control using the image shift amount determined to be reliable,
When the focus state is the first state, the calculation unit selects a first evaluation value calculated in the first region from the plurality of regions, and the focus state is in the second state. In some cases, the imaging apparatus is configured to select a second evaluation value calculated in a second region wider than the first region among the plurality of regions.   2. The first state is a focused state in which the subject is focused or a state in the vicinity of the focused state, and the second state is a state in which the subject is not focused. Or the imaging device of Claim 2.   The calculating means performs weighted addition of the first and second evaluation values using a composite ratio determined from the index to calculate an evaluation value, and the control means uses the calculated evaluation value as a threshold value. The imaging apparatus according to claim 1, wherein reliability of the image shift amount is determined by comparison.   The image pickup apparatus according to claim 2, wherein the control unit determines the reliability of the image shift amount by comparing the first and second evaluation values with a threshold value. The calculating means calculates a correlation amount of the pair of image signals by shifting pixels in the correlation calculation,
The first region is a window region for confirming a correlation between the pair of image signals, and the second region is a visual field region including a shift amount of the pair of image signals in the correlation calculation. The imaging device according to any one of claims 1 to 5. The calculating means includes
Calculate the correlation amount of the pair of image signals by shifting the pixels in the correlation calculation,
The index is obtained by dividing a change amount that changes at a shift position where a correlation displacement amount, which is a difference between the correlation amounts, becomes a minimum value of the correlation amount by a difference value between a maximum value and a minimum value of the pair of image signals. The imaging apparatus according to claim 1, wherein the imaging device is calculated. The calculating means includes
Calculate the correlation amount of the pair of image signals by shifting the pixels in the correlation calculation,
The index is calculated by dividing a correlation degree of an image signal at a shift position where the correlation value is a minimum value by a difference value between a maximum value and a minimum value of the pair of image signals. The imaging device according to claim 1 or 4.   The imaging apparatus according to claim 1, wherein the calculation unit calculates the index from the image shift amount or the defocus amount. The calculation means calculates the evaluation value from a contrast value of the pair of image signals,
The image pickup apparatus according to claim 1, wherein the control unit determines that the image shift amount is reliable when the evaluation value is greater than a threshold value. The calculation means calculates the evaluation value from the number of pixels detected by saturation detection for the pair of image signals,
The image pickup apparatus according to claim 1, wherein the control unit determines that the image shift amount is reliable when the evaluation value is smaller than a threshold value.   10. The imaging apparatus according to claim 1, wherein the calculation unit calculates the evaluation value from a contrast value of the pair of image signals and the number of pixels detected by saturation detection. 11.   The control means determines that the image shift amount is reliable when the focus state is the first state and the number of detected pixels in the first region is smaller than a threshold, The imaging apparatus according to claim 12, wherein when the number of detected pixels in one region is equal to or greater than a threshold value, the image shift amount is determined not to be reliable. The control means calculates the first evaluation value and the second evaluation value when the focus state is the second state and the number of detected pixels in the first region is smaller than a threshold value. Compare and
The imaging apparatus according to claim 13, wherein when the first evaluation value is greater than or equal to the second evaluation value, it is determined that the image shift amount is reliable.   When the second evaluation value is larger than the first evaluation value, the control means determines that the image shift amount is reliable when the second evaluation value is smaller than a threshold, and the second evaluation value The image pickup apparatus according to claim 14, wherein the image shift amount is determined not to be reliable when the evaluation value is equal to or greater than the threshold value. A control method executed by an imaging apparatus that acquires a pair of image signals that change depending on a focus state of an imaging optical system and performs focus detection,
The calculating means performs a correlation operation using the pair of image signals, calculates a correlation amount and an image shift amount, calculates an evaluation value from each of a plurality of areas set for focus detection, and represents the focus state Calculating a metric,
The control means determines the reliability of the image shift amount based on the evaluation value and the index, and performs a focus adjustment control using the image shift amount determined to be reliable,
When the focus state is the first state, the calculation unit increases the ratio of the first evaluation value calculated in the first region among the plurality of regions, and the focus state is the second state. When the state is a state, the ratio of the second evaluation value calculated in a second region wider than the first region among the plurality of regions is increased. A control method executed by an imaging apparatus that acquires a pair of image signals that change depending on a focus state of an imaging optical system and performs focus detection,
Calculating a correlation using the pair of image signals, calculating a correlation amount and an image shift amount, and calculating an evaluation value from each of a plurality of regions set for focus detection;
The control means determines the reliability of the image shift amount based on the evaluation value, and performs focus adjustment control using the image shift amount determined to be reliable,
When the focus state is the first state, the calculation unit selects a first evaluation value calculated in the first region from the plurality of regions, and the focus state is in the second state. In some cases, the control method of the imaging apparatus, wherein a second evaluation value calculated in a second region wider than the first region is selected from the plurality of regions.

Images