JP6794185B2 - Focus detector - Google Patents

Description

The present invention relates to a focus detection device, and more particularly to a focus detection device that performs focus detection by a phase difference detection method.

In the autofocus (AF) control that automatically detects the focus of the imaging lens on the subject, one of the weak subjects is a case where the subject includes a bright region (saturation region). This is because it is difficult to acquire an accurate signal from the saturated region in the case of a method of acquiring a focus detection signal from an image sensor for photographing a subject (imaging surface phase difference detection method). For this weak subject, in Patent Document 1, when the phase difference detection results of a plurality of scanning lines are added and focus detection control is performed, it is determined whether or not each scanning line is saturated, and the saturated scanning line is determined. The influence of the saturation region is reduced by controlling so that it is not added.

Japanese Unexamined Patent Publication No. 2014-215606

However, as in the method disclosed in Patent Document 1, when the saturation of each scanning line is detected, if the saturation is detected in many scanning lines, it is possible to obtain a phase difference detection result with sufficient accuracy. There was a problem that I couldn't do.

The present invention has been made in view of such problems of the prior art, and provides a focus detection device capable of suitable focus detection for a saturated region in focus detection control using an imaging surface phase difference detection method. The purpose.

The focus detection device of the present invention for achieving the above object is an imaging lens for imaging a subject by using a plurality of image signals acquired from an imaging element having a plurality of photoelectric conversion units for the microlens. A focus detection device that performs focus detection, and a generation means that generates a plurality of focus detection signals using the plurality of image signals based on the amount of incident light from the imaging lens, and a correlation amount of the plurality of focus detection signals. It is determined whether or not at least one of the plurality of image signals has reached the saturation level, the correlation amount detecting means for detecting, the focus detecting means for detecting the focus of the imaging lens based on the correlation amount. The image signal based on the difference in the amount of incident light for each of the plurality of photoelectric conversion units, and the determination means for clipping the signal level of the image signal to a predetermined level based on the determination result of the determination means. The adjusting means has an adjusting means for adjusting the difference between the above and the above using a predetermined adjusting parameter, and the adjusting means adjusts the difference of the image signal based on the determination result of the image height determining means. It is characterized in that the clipping means is configured to control the predetermined level for each photoelectric conversion unit according to the adjustment parameter held by the holding means.

According to the present invention, it is possible to provide a focus detection device capable of suitable focus detection for a saturated region in focus detection control using an imaging surface phase difference detection method.

It is a block diagram which shows the structure of the image pickup system which concerns on this invention. It is a figure which shows the pixel arrangement of an image sensor. It is a flowchart which shows the imaging process which concerns on this invention. It is a flowchart which shows the still image imaging process which concerns on this invention. It is a flowchart which shows the saturation level setting process which concerns on this invention. It is a flowchart which shows the focus state detection process which concerns on this invention. It is a figure which shows the AF area used in the focus detection process. It is a figure which shows the pair of image signals obtained from the AF area. It is a figure which shows the relationship between the shift amount and the correlation amount of a pair of image signals. It is a figure which shows the relationship between the shift amount of a pair of image signals, and the correlation change amount. It is a flowchart which shows the focus detection process which concerns on this invention. It is a figure which shows the process of generating an image signal. It is a figure which shows the incident light amount difference correction. It is a figure which shows the case where the incident light amount difference correction is applied to a pair of image signals. It is a figure which shows the case where the incident light amount difference correction is applied to a pair of image signals. It is a flowchart which shows the saturation level setting process which concerns on 2nd Embodiment. It is a figure which shows the case where the incident light amount difference correction is not applied. It is a block diagram which shows the structure of the image pickup system which concerns on 3rd Embodiment. It is a flowchart which shows the still image imaging process which concerns on 3rd Embodiment. It is a flowchart which shows the saturation region determination value setting process which concerns on 3rd Embodiment. It is a flowchart which shows the focus state detection process which concerns on 3rd Embodiment. It is a flowchart which shows the pixel thinning-out amount / bandpass filter setting process which concerns on 3rd Embodiment. It is a flowchart which shows the correlation amount addition determination process which concerns on 3rd Embodiment. It is a figure which shows the case where the incident light amount difference correction is applied to a pair of image signals.

Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the accompanying drawings. However, the dimensions, materials, shapes, relative arrangements, and the like of the components exemplified in the present invention can be appropriately changed depending on the configuration of the apparatus to which the present invention is applied and various conditions, and the present invention exemplifies them. It is not limited to.

(First Embodiment)
FIG. 1 is a block diagram showing a functional configuration example of an interchangeable lens type camera system as an example of the image pickup apparatus according to the first embodiment of the present invention.

The imaging device of this embodiment is composed of an interchangeable lens unit 10 and a camera body 20. The lens control unit 106 that controls the operation of the entire lens unit 10 and the camera control unit 214 that controls the operation of the entire camera system including the lens unit 10 mutually control each other through terminals provided on a lens mount (not shown). Communication is possible. In the lens mount of the present embodiment, communication between the lens control unit 106 and the camera control unit 214 is performed via terminals (electrical contacts), but the present invention is not limited to this. For example, at least a part of the communication may be performed by wireless communication, or a control command or the like may be communicated with an external device by providing a terminal other than the lens mount.

First, the configuration of the lens unit 10 will be described. The fixed lens 101, the aperture 102, the focus lens 103, and the like constitute an imaging optical system. The diaphragm 102 is driven by the diaphragm drive unit 104 and controls the amount of light incident on the image sensor 201, which will be described later. The focus lens 103 is driven by the focus lens driving unit 105, and the focusing distance of the imaging optical system changes according to the position of the focus lens 103. The aperture drive unit 104 and the focus lens drive unit 105 are controlled by the lens control unit 106 to determine the aperture amount of the aperture 102 and the position of the focus lens 103.

The lens operation unit 107 is a group of input devices for the user to make settings related to the operation of the lens unit 10, such as switching between AF (autofocus) / MF (manual focus) modes and adjusting the position of the focus lens 103 in the MF mode. Is. When the lens operation unit 107 is operated, the lens control unit 106 performs control according to the operation. The lens operation unit 107 may include not only an input device such as a toggle switch or a touch panel, but also a display unit that displays a lens state such as the current position of the focus lens 103 and various setting information.

The lens control unit 106 controls the aperture drive unit 104 and the focus lens drive unit 105 according to control commands and control information received from the camera control unit 214, which will be described later. Further, the lens control information is transmitted to the camera control unit 214.

Next, the configuration of the camera body 20 will be described. The camera body 20 is configured to be able to acquire an imaging signal from a luminous flux that has passed through the imaging optical system of the lens unit 10.

The image sensor 201 is composed of a CCD or CMOS sensor. The luminous flux incident from the image pickup optical system of the lens unit 10 is imaged on the light receiving surface of the image pickup element 201, and the amount of incident light is generated by the photodiodes provided in the pixels two-dimensionally arranged horizontally and vertically on the image pickup element 201. It is converted into a signal charge according to. The signal charge accumulated in each photodiode is sequentially read from the image sensor 201 as a voltage signal corresponding to the signal charge from the drive pulse output by the timing generator 216 according to the command of the camera control unit 214. In this embodiment, the photodiode corresponds to the photoelectric conversion unit. Therefore, the same effect can be obtained even when an organic or inorganic photoelectric conversion film or the like is used as the photoelectric conversion film.

Each pixel of the image sensor 201 used in the present embodiment is composed of two (pair) photodiodes A and B and one microlens provided for the pair of photodiodes A and B. .. In each pixel, the light incident from the lens unit 10 is divided by a microlens to form a pair of optical images in the pair of photodiodes A and B. Then, a pair of pixel signals (A signal and B signal) used for the focus detection signal described later are generated from the photodiodes A and B. Further, by adding the outputs of the pair of photodiodes A and B, an imaging signal (A + B signal) can also be obtained. In the embodiment of the present invention, the image sensor 201 is configured to output two types of output signals, an image pickup signal (A + B signal) and a focus detection signal (A signal). The reason why all three types of the image pickup signal (A + B signal) and the focus detection signal (A signal and B signal) are not output from the image sensor 201 is to reduce the output load and read out as soon as possible. The focus detection signal (B signal) is generated by subtracting the focus detection signal (A signal) from the imaging signal (A + B signal). The focus detection signal (B signal) is generated by the saturation level clip unit / B signal generation unit 204, which will be described later. Further, in the present embodiment, the image sensor 201 has a configuration having two photodiodes, but may have a configuration having two or more photodiodes.

Here, the pixel configuration of the image pickup device 201 will be described in detail with reference to FIG. FIG. 2A shows a pixel configuration that does not support conventional imaging surface phase difference detection, and FIG. 2B shows a pixel configuration of the image sensor 201 that supports imaging surface phase difference detection. In each of the figures, Bayer color filters are used, where R indicates a red color filter, B indicates a blue color filter, and Gr and Gb indicate a green color filter. The pixel configuration shown in FIG. 2 (b) includes two pixels that are horizontally divided into two in a pixel corresponding to one pixel (shown by a solid line) in the conventional pixel configuration shown in FIG. 2 (a). Photodiodes A and B are provided. The pixel division method shown in FIG. 2B is merely an example, and the pixel may be divided in the vertical direction of the figure, or may be divided into two or more. Further, a plurality of types of pixels divided by different division methods in the same image sensor may be included.

The CDS / AGC / AD converter 202 performs correlated double sampling, gain adjustment, and AD conversion on the focus detection signal and the image pickup signal read from the image sensor 201 to remove reset noise. The AD converter 202 outputs the image pickup signal and the focus detection signal that have undergone these processes to the image input controller 203 and the AF signal processing unit 206, respectively. A part or all of the CDS / AGC / AD converter 202, the timing generator 216, and the like may be included in the image sensor 201.

The image input controller 203 stores the imaging signal output from the CDS / AGC / AD converter 202 in the SDRAM 211 as an image signal via the bus 21. The image signal stored in the SDRAM 211 is read by the display control unit 207 via the bus 21 and displayed on the display unit 208. Further, in the recording mode in which the image signal is recorded, the image signal stored in the SDRAM 211 is recorded in the recording medium 210 such as a semiconductor memory by the recording medium control unit 209.

The ROM 212 stores control programs and processing programs executed by the camera control unit 214, various data necessary for their execution, and the like. The flash ROM 213 stores various setting information and the like related to the operation of the camera body 20 set by the user.

The saturation level clip unit / B signal generation unit 204 compares the A signal, which is the focus detection signal output from the CDS / AGC / AD converter 202, with a set constant signal level (saturation threshold value). This is because when the level of the focus detection signal has reached saturation, the accuracy of the correlation calculation described later is lowered. Therefore, in the focus detection signal, if the signal exceeds the saturation threshold value, a process of clipping at the saturation level (clip process) is performed. Further, the B signal is generated by taking the difference between the imaging signal (A + B signal) output from the CDS / AGC / AD converter 202 and the A signal clipped at the saturation level.

The reason why the B signal is generated in consideration of saturation from the difference between the imaging signal (A + B signal) and the A signal will be described in detail with reference to FIG. In FIG. 12, the horizontal axis is the horizontal pixel position, and the vertical axis is the plot of the output value from each pixel. Then, the case where the B signal is generated without performing the clip processing is shown in FIG. 12 (a). Here, it is assumed that the A + B signal hits the sensor saturation level and the A signal also hits the sensor saturation level (FIG. 12 (a) left). As shown in the left figure of FIG. 12A, when both the A + B signal and the A signal have reached the saturation level, the difference between the two signals becomes very small. Therefore, when the B signal is calculated, the shape of the B signal is dented and collapsed in the region where the A signal reaches the saturation level, as shown in the right figure of FIG. 12A. As a result, the correlation between the A signal and the B signal cannot be obtained by the correlation calculation described later, and the image shift amount and the defocus amount cannot be detected.

On the other hand, the case where the clip processing is performed is shown in FIG. 12 (b). When the A + B signal and the A signal have the same signal level as in the left figure of FIG. 12 (a), the saturation threshold value is such that the A signal is clipped at half the saturation level as shown in the left figure of FIG. 12 (b). To set. Therefore, half of the saturation level becomes the maximum value of the A signal as shown in the A signal in the left figure of FIG. 12B. By calculating the B signal in this state, it is possible to suppress the phenomenon that the shape of the B signal collapses like a dent. Further, the A signal and the B signal can be correlated by the correlation calculation described later, and the image shift amount and the defocus amount can be appropriately detected.

The A signal and the B signal, which are the focus detection signals, are further processed by the light amount difference correction unit 205 and the AF signal processing unit 206, which will be described later. Further, as a characteristic operation of the present invention in the saturation level clip unit / B signal generation unit 204, the saturation level of a pair of image signals for each pixel is determined from the camera control unit 214, which will be described later, based on the incident light amount difference information for each pixel. Set and change the saturation level for each pixel. This feature will be described in detail later.

The light amount difference correction unit 205 multiplies the A signal and B signal input from the saturation level clip unit / B signal generation unit 204 by the signal level correction coefficient calculated based on the incident light amount difference information to obtain the signal level. Make corrections. As shown in FIG. 13, the signal level is corrected so as to make the difference in the amount of incident light generated depending on the position (pixel position) of the pixel to which each signal is output uniform. The left figure of FIG. 13A shows the A signal and the B signal near the center of the screen, with the signal level on the vertical axis and the horizontal pixel position on the horizontal axis. The middle figure of FIG. 13A shows the incident light amount difference correction coefficient for correcting the incident light amount difference near the center of the screen. The right figure of FIG. 13 (a) shows A after each of the A signal and the B signal shown in the left figure of FIG. 13 (a) is corrected by using the incident light amount difference correction coefficient shown in the middle figure of FIG. 13 (a). It shows the signal levels of the signal and the B signal. Further, each figure of FIG. 13B shows an A signal, a B signal, and the like corresponding to the area on the right side (periphery of the screen) of the screen with respect to each figure of FIG. 13A.

As shown in the left figure of FIG. 13 (a) and the left figure of FIG. 13 (b), when a difference in signal level or a difference in slope occurs due to a difference in the amount of incident light, the correlation calculation accuracy described later is improved. It will drop. As a result, if the error of the correlation calculation for calculating the correlation amount becomes large, the error of the image shift amount becomes large, that is, the accuracy of the focus detection is lowered. The light amount difference correction unit 205 corrects the signal level difference and the inclination difference based on the incident light amount difference information in order to reduce the influence of the large error of the correlation amount due to the signal level difference and the inclination difference. The incident light amount difference information, that is, the incident light amount difference correction coefficient, is different in the aperture information of the lens, the exit pupil distance information, and FIGS. 13 (a) and 13 (b), and the AF frame position in the screen (for each image height). The data calculated in advance based on the information or the like is stored in the ROM 212. Then, the camera control unit 214, which will be described later, makes settings suitable for the operating state according to the operating state of the camera body 20. It should be noted that a part of the parameters for calculating the incident light amount difference correction coefficient or the correction coefficient may be stored in a memory or the like on the lens unit 10 side. In this case, the camera body 20 acquires the parameters and the like stored in the memory when the lens is attached.

The AF signal processing unit 206 performs a correlation calculation on each image signal and calculates the amount of image shift and the reliability of each image signal. The reliability is calculated by using the two-image coincidence and the steepness of the correlation change amount, which will be described later. The AF signal processing unit 206 outputs the image shift amount (detection amount) and reliability information calculated in the AF region to the camera control unit 214. In the present embodiment, the AF signal processing unit 206 corresponds to the correlation amount detection unit for detecting the correlation amount of the focus detection signal.

The camera control unit 214 controls each unit in the camera body 20 and the lens control unit 106 while exchanging information with each other. The camera control unit 214 changes the settings of the AF signal processing unit 206 as necessary based on the image shift amount, reliability, and information indicating the states of the lens unit 10 and the camera body 20 obtained by the AF signal processing unit 206. To do. For example, when the amount of image shift for the AF signal processing unit 206 is equal to or greater than a predetermined amount, a wide area for performing correlation calculation can be set, or the type of bandpass filter can be changed according to the contrast of a pair of image signals. Or something. Further, as described above, the saturation level of each pixel is set in the saturation level clip unit / B signal generation unit 204, and the light amount difference correction parameter of the image signal is set in the light amount difference correction unit 205. .. Further, the camera control unit 214 performs user operations such as power ON / OFF, change of various settings, imaging processing, focus detection processing, and recording image reproduction processing in response to input from the camera operation unit 215 based on user operation. Performs various processes corresponding to. Further, the camera control unit 214 transmits a control command to the lens unit 10 (lens control unit 106) and information on the camera body 20 to the lens control unit 106, and acquires information on the lens unit 10 from the lens control unit 106. To do. The camera control unit 214 is composed of a microcomputer, and controls the entire camera system including the interchangeable lens 10 by executing a computer program recorded in the ROM 212.

The camera control unit 214 calculates the defocus amount using the image shift amount in the AF region calculated by the AF signal processing unit 206, and drives the focus lens 103 through the lens control unit 106 based on the defocus amount. To control.

The saturation level clipping unit / B signal generation unit 204, which is a feature of the present invention, describes changing the saturation level for clipping the image signal for each pixel. As described above with reference to FIG. 12, when the difference between the A + B signal and the A signal is taken to generate the B signal, the B signal can be generated without breaking by clipping the A signal at the set saturation level. Here, a case where the processing of the light amount difference correction unit 205 described with reference to FIG. 13 is performed on the A signal and the generated B signal will be described with reference to FIG. Each signal shown in FIG. 14 is the same as in FIG. 13, but the application of the incident light amount difference correction coefficient to the image signal after the clip processing is taken into consideration. If the incident light amount difference correction is performed in a state where clips of saturated pixels are generated as shown in the right figure of FIG. 13, the shape of the pair of image signals is distorted by the correction coefficient of the incident light amount difference. In this way, when the correlation calculation described later is performed on a pair of image signals that are distorted by correcting the incident light amount difference after clipping the saturated pixels, the calculation accuracy of the image shift amount decreases due to the image shape difference, and the image The error of the defocus amount calculated based on the deviation amount also becomes large.

To solve such a problem, the present invention changes the saturation level at which clipping is performed for each pixel for a pair of image signals. As shown in FIG. 15, even when the pair of image signals contains saturated pixels, in order to make the signal levels uniform after clipping the saturated pixels and applying the incident light amount difference correction. The signal level to be clipped is changed for each pixel. As a result, the phenomenon that the pair of image signals are distorted and a shape difference is generated after the incident light amount difference correction shown in FIG. 14 does not occur, the error of the calculated defocus amount is reduced, and the saturated subject is treated. AF accuracy can also be improved.

Hereinafter, each process performed by the camera body 20 will be described. The camera control unit 214 performs the following processing according to an imaging processing program which is a computer program.

FIG. 3 is a flowchart showing a procedure of imaging processing of the camera body 20. S means a step.

In S301, the camera control unit 214 performs initialization processing such as various settings set in the camera body 20, and proceeds to S302.

In S302, the camera control unit 214 determines whether the image pickup mode of the camera body 20 is the moving image imaging mode or the still image capturing mode, and processes the camera body 20 to S303 if it is the moving image imaging mode or S304 if it is the still image capturing mode. Proceed.

In S303, the camera control unit 214 performs moving image imaging processing and proceeds to S305.

In S304, the camera control unit 214 performs a still image imaging process and proceeds to S305. In the present embodiment, the details of the moving image imaging process are omitted, but the same can be applied as in the application example of the invention in the still image imaging process described later.

In S305, the camera control unit 214 determines whether or not the imaging process has been stopped, proceeds to S306 if it has not been stopped, and ends the imaging process if it is stopped. When the imaging process is stopped, for example, when the power of the camera body 20 is turned off through the camera operation unit 215, the user setting process of the camera body 20, the reproduction process for confirming the captured image / moving image, etc. When an operation other than is performed.

In S306, the camera control unit 214 determines whether or not the imaging mode has been changed. When the change is made, the camera control unit 214 returns to S301, performs initialization processing, and then performs imaging processing in the changed imaging mode. On the other hand, if the imaging mode has not been changed, the camera control unit 214 returns to S302 and continues the processing in the current imaging mode.

Next, the still image imaging process performed in S304 in FIG. 3 will be described with reference to the flowchart of FIG.

In S401, the camera control unit 214 sets the saturation level for the saturation level clip unit / B signal generation unit 204, and proceeds to S402. The saturation level setting process of S401 is a characteristic part of the present invention, and details will be described later using the flowchart of FIG.

In S402, the camera control unit 214 performs the light amount difference correction value setting process on the light amount difference correction unit 205, and proceeds to the process to S403. In the light amount difference correction value setting process of S402, as described with reference to FIG. 13, a process of adjusting the level of the image signal so as to eliminate the difference in the incident light amount for each pixel and make it uniform is performed. In the imaging surface phase difference detection method, the image signal is based on information such as the position (image height) on the image sensor 201 that forms an image signal, the state of the aperture 102, and the distance information between the image sensor 201 and the aperture 102. The amount of light changes. Focus detection accuracy can be improved by correcting the difference in the amount of incident light of each image signal and eliminating the level difference and tilt difference of the pair of image signals.

In S403, the camera control unit 214 causes the AF signal processing unit 206 to perform the focus detection process. The focus detection process is a process for acquiring information on the amount of defocus and reliability. Details will be described later.

In S404, the camera control unit 214 determines whether or not a focus detection process start instruction (hereinafter, referred to as an AF instruction) has been input (on) from the camera operation unit 215. The AF instruction is input from the camera operation unit 215 when the shutter button provided on the camera body 20 is half-pressed, or when the AFON button for executing AF is pressed. The camera control unit 214 proceeds to S405 when the AF instruction is input, and proceeds to S406 when the AF instruction is not input.

In S405, the camera control unit 214 performs the focus detection process. This focus detection process will be described later.

In S406, the camera control unit 214 determines whether or not an imaging process start instruction (hereinafter referred to as an imaging instruction) has been input (on) from the camera operating unit 215. The imaging instruction is output from the camera operation unit 215 when the shutter button is fully pressed. The camera control unit 214 proceeds to S407 when an imaging instruction is input, and proceeds to S409 when an imaging instruction is not input.

In S407, the camera control unit 214 determines whether or not the focus is currently stopped by the focus detection process in S405. The focusing stop state is a state in which the imaging optical system is in the focusing state and the driving of the focus lens 103 is stopped. If it is not in the in-focus stop state, the camera control unit 214 proceeds to S405 and obtains the in-focus state by starting or continuing the focus detection process. When the focus is stopped, the camera control unit 214 proceeds to S408, performs imaging processing, stores the captured image (recorded image) on the recording medium 210 via the recording medium control unit 209, and proceeds to S409.

In S409, the camera control unit 214 releases the focusing stop state and ends the still image imaging process.

Next, the saturation level setting process performed by the camera control unit 214 in S401 of FIG. 4 will be described with reference to the flowchart of FIG.

First, in S501, the camera control unit 214 sets the AF region (subject) to be AF on the image pickup screen (that is, the image sensor 201). The AF area setting method is performed based on an instruction by the user through the camera operation unit 215 or, in the case of a configuration of a camera system capable of detecting a subject, a subject detection state.

Next, in S502, the camera control unit 214 acquires or calculates the light amount difference correction value for each pixel of the image signal in the AF region set in S501.

Next, in S503, the camera control unit 214 calculates the saturation level for each pixel from the light amount difference correction value for each pixel of the image signal acquired in S502. The saturation level calculation method is based on a constant saturation level value SATLVL, the saturation level value of A signal is the light amount difference correction value of SATLVL × B signal, and the saturation level of B signal is the light amount difference of SATLVL × A signal. It is calculated as a correction value. The constant saturation level SATLVL corresponds to the saturation clip level described in the graph of the pair of image signals before correction of the incident light amount difference in FIG. 14. The constant saturation level SATLVL multiplied by the light intensity difference correction value is the saturation clip described in the graph of the pair of image signals before the incident light intensity difference correction when the saturation clip level change for each pixel in FIG. 15 is applied. Corresponds to the level.

Next, in S504, the camera control unit 214 sets the saturation level calculated in S503 in the saturation level clip unit / B signal generation unit 204, and ends the saturation level setting process.

The effect of changing the saturation clip level based on the incident light amount difference correction information for each pixel of the pair of image signals is as described above with reference to FIG.

Next, the focus state detection process performed by the AF signal processing unit 206 in S403 of FIG. 4 will be described with reference to the flowchart of FIG. The AF signal processing unit 206 is configured by a microcomputer and performs focus detection processing according to a focus detection program (which may be a part of an imaging processing program) as a computer program.

In S601, the AF signal processing unit 206 performs clipping processing of saturated pixels on the AF region set in S501 of FIG. 5 described above by the saturation level clip unit / B signal generation unit 204, and the light amount difference correction unit 205 performs the clipping processing. The image signal corrected for the difference in the amount of incident light is acquired. In addition, based on the saturation level and the light amount difference correction value set in S401 and S402 of FIG. 4, the result of each operation of the saturation level clip unit / B signal generation unit 204 and the light amount difference correction unit 205 can be acquired in S601. However, each operation is omitted in the flowchart.

Next, in S602, the AF signal processing unit 206 calculates the amount of correlation between these image signals while relatively shifting the acquired image signals by one pixel (1 bit). The amount of correlation is calculated as described later in each of the plurality of pixel lines (hereinafter referred to as scanning lines) provided in the AF region. After calculating the correlation amount of each scanning line, each correlation amount is added and averaged to calculate as one correlation amount. Further, in the present embodiment, the pair of image signals are relatively shifted one pixel at a time in the correlation calculation, but a configuration in which the pair of image signals are shifted in units of more pixels may be used. For example, a configuration in which two pixels are relatively shifted may be used. Further, in the present embodiment, one correlation amount is calculated by adding and averaging the correlation amounts of each scanning line. For example, the pair of image signals of each scanning line are added and averaged, and then added and averaged. The configuration may be such that the correlation amount is calculated for a pair of image signals.

Next, in S603, the AF signal processing unit 206 obtains the correlation change amount from the correlation amount calculated in S602. The method of calculating the amount of correlation change will be described later.

Then, in S604, the AF signal processing unit 206 calculates the image shift amount using the correlation change amount calculated in S603. The method of calculating the amount of image shift will be described later.

Further, in S605, the AF signal processing unit 206 calculates the reliability indicating the high reliability of the image shift amount calculated in S604. The method of calculating the reliability will be described later.

Next, in S606, the AF signal processing unit 206 calculates the defocus amount of the AF region using the image shift amount of the AF region calculated in S604, and ends the focus detection process.

Next, the focus detection process will be described in detail. FIG. 7 shows an example of the AF region 702 on the pixel array 701 of the image sensor 201 in the focus detection process. The shift regions 703 on both sides of the AF region 702 are regions necessary for the correlation calculation. Therefore, the pixel area 704, which is the sum of the AF area 702 and the shift area 703, is the pixel area required for the correlation calculation. In the figure, p, q, s, and t represent the coordinates in the horizontal direction (x-axis direction), p and q represent the x-coordinates of the start point and the end point of the pixel area 704, respectively, and s and t represent the AF area, respectively. The x-coordinates of the start point and the end point of 702 are shown.

FIG. 8 shows an example of a pair of image signals for AF acquired from a plurality of pixels included in the AF region 702 shown in FIG. The solid line 801 is one A signal, and the broken line 802 is the other B signal. 8 (a) shows the A signal and the B signal before the shift, and FIGS. 8 (b) and 8 (c) shift the A signal and the B signal from the state of FIG. 8 (a) in the positive direction and the negative direction, respectively. It shows the state of. When calculating the correlation amount of the pair of A signal 801 and B signal 802, both the A signal 801 and the B signal 802 are shifted by 1 bit in the direction of the arrow.

Next, a method of calculating the correlation amount will be described. First, as shown in FIGS. 8B and 8C, the A signal 801 and the B signal 802 are shifted by 1 bit each, and the sum of the absolute values of the differences between the A signal 801 and the B signal 802 is calculated. Let i be the shift amount, ps be the maximum shift amount in the minus direction, qt be the maximum shift amount in the plus direction, x be the start coordinate of the AF region 702, and y be the end coordinate of the AF region 702. Then, the correlation amount COR can be calculated by the following equation (1).

FIG. 9A shows an example of the relationship between the shift amount and the correlation amount COR. The horizontal axis shows the shift amount, and the vertical axis shows the correlation amount COR. Of the extreme values 902 and 903 in the correlation amount 901 that changes with the shift amount, the degree of coincidence between the A signal and the B signal is the highest in the shift amount corresponding to the smaller correlation amount.

Next, a method of calculating the amount of correlation change will be described. The difference in the correlation amount every other shift in the waveform of the correlation amount 901 shown in FIG. 9A is calculated as the correlation change amount. Assuming that the shift amount is i, the maximum shift amount in the minus direction is ps, and the maximum shift amount in the plus direction is qt, the correlation change amount ΔCOR can be calculated by the following equation (2).
ΔCOR [i] = COR [i-1] -COR [i + 1]
{(P-s + 1) <i <(q-t-1)} (2)

FIG. 10A shows an example of the relationship between the shift amount and the correlation change amount ΔCOR. The horizontal axis shows the shift amount, and the vertical axis shows the correlation change amount ΔCOR. The correlation change amount 1001 that changes with the shift amount changes from plus to minus in the portion of 1002 and 1003. The state in which the amount of correlation change is 0 is called zero cross, and the degree of coincidence between the pair of image signals A and B is the highest. Therefore, the shift amount that gives zero cross is the image shift amount PRD.

FIG. 10 (b) shows an enlarged portion of the portion shown by 1002 in FIG. 10 (a). 1004 is a part of the correlation change amount 1001. A method of calculating the image shift amount PRD will be described with reference to FIG. 10B.

The shift amount (k-1 + α) that gives a zero cross is divided into an integer part β (= k-1) and a decimal part α. The fractional part α can be calculated by the following equation (3) from the similarity relationship between the triangle ABC and the triangle ADE in the figure.

The integer part β can be calculated from FIG. 10 (b) by the following equation (4).
β = k-1 (4)

Then, the image shift amount PRD can be calculated from the sum of α and β.

When there are a plurality of zero crosses of the correlation change amount ΔCOR as shown in FIG. 10A, the one having a larger change in the correlation change amount ΔCOR in the vicinity thereof is defined as the first zero cross. This steepness is an index showing the ease of performing AF, and the larger the value, the easier it is to perform accurate AF. The steepness maxder can be calculated by the following equation (5).
maxder = │ ΔCOR [k-1] │ + │ ΔCOR [k] │ (5)

As described above, in the present embodiment, when there are a plurality of zero crosses of the correlation change amount, the first zero cross is determined by the steepness thereof, and the shift amount giving the first zero cross is defined as the image shift amount PRD.

Next, a method of calculating the reliability of the image shift amount PRD will be described. The reliability of the image shift amount can be defined by the degree of coincidence of the pair of image signals A and B (hereinafter referred to as the degree of coincidence between two images) fnclvl and the steepness of the above-mentioned amount of correlation change. The two-image coincidence degree is an index showing the accuracy of the amount of image shift, and in the correlation calculation in the present embodiment, the smaller the value, the better the accuracy. The reliability of the image shift amount PRD can also be used as the reliability of the defocus amount calculated based on the image shift amount PRD.

In FIG. 9B, the portion shown by 902 in FIG. 9A is enlarged, and 904 is a part of the correlation amount 901. The two-image agreement fnclvl can be calculated by the following equation (6).
(I) When │ΔCOR [k-1] │ × 2 ≦ maxder
fnclvl = COR [k-1] + ΔCOR [k-1] / 4
(Ii) │ ΔCOR [k-1] │ × 2> maxder
fnclvl = COR [k] -ΔCOR [k] / 4 (6)

Next, the focus detection process performed by the camera control unit 214 in S405 of FIG. 4 will be described with reference to the flowchart of FIG.

In S1101, the camera control unit 214 determines whether or not the AF is currently completed and is in the focusing stop state, proceeds to S1102 if the focus is not stopped, and proceeds to S1109 if the focus is stopped.

In S1102, the camera control unit 214 determines whether or not the reliability of the defocus amount calculated by the focus state detection process in S403 is higher than the predetermined reliability. The reliability referred to here is determined by the steepness of the two-image coincidence and the image shift amount described above, and not only the calculated defocus amount but also the defocus direction determines the maximum value of the unreliable reliability range. It is desirable to set it as reliability. The reliability may be obtained by using both the degree of coincidence between the two images and the steepness of the amount of image shift, or may be obtained by using only one of them. Further, other indexes such as the signal level of two images may be used. If the defocus amount is higher than the predetermined reliability, the process proceeds to S1103, and if not, the process proceeds to S1107.

In S1103, since the camera control unit 214 performs AF using a highly reliable defocus amount, it first determines whether or not the defocus amount is within the depth of focus, and if it is within the depth of focus, proceeds to S1104. If it is not within the depth of focus, the process proceeds to S1105.

In S1104, the camera control unit 214 determines that the defocus amount is in the in-focus state within the depth of focus, shifts to the in-focus stop state, and ends the focus detection process.

On the other hand, in S1105, the camera control unit 214 considers that the in-focus state has not been obtained yet, sets the lens drive for driving the focus lens 103 based on the defocus amount, and proceeds to S1106. The lens drive setting is a setting such as a drive speed of the focus lens 103 and a gain applied to the defocus amount in consideration of an error in the defocus amount.

In S1106, the camera control unit 214 transmits a control command for the focus lens 103 to the lens control unit 106 based on the defocus amount and the lens drive setting information set in S1105. That is, the drive control of the focal position of the focus lens 103 is performed. As a result, the focus detection process ends.

On the other hand, in S1107, the defocus amount calculated in S606 cannot be used. Therefore, the camera control unit 214 is a search drive that calculates the defocus amount while moving the focus lens 103 toward the movable end in order to detect the position of the focus lens 103 that can obtain a highly reliable defocus amount. I do. Therefore, the camera control unit 214 first sets the lens drive for search drive. The lens drive setting for search drive is a setting such as a drive speed of the focus lens 103 and a direction for starting drive.

Subsequently, in S1108, the camera control unit 214 transmits a control command for the focus lens 103 to the lens control unit 106 based on the lens drive setting for search drive set in S1107. Then, the focus detection process is terminated. In this embodiment, the camera body 20 that can perform only the imaging surface phase difference detection method using a pair of image signals as the output signal of the image sensor 201 has been described. However, when the contrast detection method AF can be performed using the output signal of the image sensor 201, the contrast detection method AF may be performed when it is determined in S1102 that the reliability of the defocus amount is low.

In S1109, the camera control unit 214 keeps the focusing stop state and ends the focus detection process.

As described above, in the present embodiment, when clipping the signal level of the saturated pixel, the saturation clip level of the saturated pixel is changed based on the information of the incident light amount difference of the pair of image signals. After applying the clipping processing of the saturated pixels, the signal level of the saturated pixel portion of the pair of image signals after the correction of the incident light amount difference is made constant. This reduces the occurrence of shape differences such as signal level and tilt deviation of the pair of image signals, and improves the calculation accuracy of the image deviation amount by correlation calculation, that is, the calculation accuracy of the defocus amount, thereby achieving AF accuracy. Can be improved.

The scope of the present invention is not limited to the configuration illustrated in the present embodiment. For example, in the present embodiment, as shown in FIG. 2B, the configuration of the image sensor that can perform the correlation calculation in the horizontal direction is used, but the configuration that can perform the correlation calculation in the vertical direction or the configuration that can perform the correlation calculation in the horizontal direction or the horizontal / vertical direction The image sensor may be configured so that the correlation calculation can be performed.

(Second Embodiment)
Hereinafter, a second embodiment of the present invention will be described. In the first embodiment, when clipping the signal level of the saturated pixel, the signal level to be clipped is changed based on the light amount difference information of the pair of image signals. On the other hand, in the second embodiment, the signal level to be clipped is constant, and when the light amount difference correction based on the light amount difference information is performed, the light amount difference correction is performed between the unsaturated pixels and the saturated pixels. To change.

The configuration of the interchangeable lens camera including the lens and the camera body 20 in the present embodiment is the same as the configuration shown in FIG. 1 in the first embodiment. Therefore, the description of the same contents as that of the first embodiment will be omitted. However, for the saturation level clip unit / B signal generation unit 204, it is sufficient that at least one clipping level can be set, and it is not necessary to set each pixel. Further, in the light amount difference correction unit 205 in the present embodiment, the signal level is clipped by the saturation level clip unit / B signal generation unit 204, that is, it is detected as a saturated pixel for each pixel. It is possible to switch whether or not to apply the light intensity difference correction.

The operation of the light amount difference correction unit 205, which is a characteristic part of the present invention, will be described with reference to FIG. FIG. 17 corresponds to the content of FIG. 15 in the first embodiment. In the image signal before the incident light amount difference correction, if the signal level reaches the set saturation clip level SATLVL, the signal level becomes constant, otherwise there is a signal level difference corresponding to the incident light amount difference. Become in a state. When performing the incident light amount difference correction in that state, as shown in the graph (middle figure) of the incident light amount difference correction coefficient in FIG. 17, the pixel portion that has reached the saturation clip level should not be corrected for the incident light amount difference. (Apply the solid line part of the saturation detection pixel part). By performing the incident light amount difference correction on the pixel part where the incident light amount difference remains and not correcting the pixel part where the incident light amount difference does not remain due to the clip of the saturated pixel, the signal after the incident light amount difference correction is performed. Avoid level differences and tilt differences. As a result, the accuracy of obtaining the image shift amount from the pair of image signals, that is, the accuracy of obtaining the defocus amount can be improved, and the accuracy of AF can be improved.

Next, the process performed by the camera body 20 in the present embodiment will be described. The imaging process, the still image imaging process, the focus state detection process, and the focus detection process in the present embodiment are the same as those shown in FIGS. 3, 4, 6, and 11 in the first embodiment. Omit.

The saturation level setting process performed by the camera control unit 214 in S401 of FIG. 4 will be described with reference to the flowchart of FIG.

Since the process of S1601 in FIG. 16 is the same as the process of S501 in FIG. 5, the description thereof will be omitted.

In S1602, the camera control unit 214 sets the saturation level for the saturation level clip unit / B signal generation unit 204, and ends the process. The saturation clip level set at this time may be a constant value.

The light amount difference correction unit 205 in the present embodiment is used for each pixel depending on whether the signal level is clipped by the saturation level clip unit / B signal generation unit 204, that is, it is detected as a saturated pixel. Switches whether or not to apply the light intensity difference correction. In this embodiment, a bit for a flag may be added to each signal to determine whether or not the signal level has been clipped by the saturation level clipping unit / B signal generation unit 204. Appropriate correction is possible by detecting the bit for this flag in the light amount difference correction unit 205. Further, the threshold value used for detecting saturated pixels is preferably set to a value smaller than the saturated clip level. This is because the light amount difference is corrected after the clip processing, so that a level difference of each image signal may occur depending on the amount of gain to be transpired. Therefore, it can be said that it is preferable not to apply the light amount difference correction even to a pixel having a signal level slightly smaller than the saturation clip level.

As described above, in the present embodiment, when correcting the incident light amount difference, the incident light amount difference is corrected for the non-saturated pixels based on the information on whether or not each pixel of the image signal is saturated. , Saturated pixels are controlled so that the difference in the amount of incident light is not corrected. After applying the clipping process of saturated pixels, the signal level difference and inclination difference will occur in the part where the signal level is already constant, that is, the pixel part where saturation is detected, if the incident light amount difference correction is performed. , The incident light amount difference is not corrected. On the other hand, the pixel portion where the saturated pixel is not detected corrects the difference in the amount of incident light. This reduces the occurrence of shape differences such as signal level and tilt deviation of the pair of image signals, and improves the calculation accuracy of the image deviation amount by correlation calculation, that is, the calculation accuracy of the defocus amount, thereby achieving AF accuracy. Can be improved. Although the system is different between the first embodiment and the present embodiment, the same effect can be expected. Therefore, it is desirable to use a method that is easy to apply on the system.

(Third Embodiment)
Hereinafter, a third embodiment of the present invention will be described.

In the first embodiment, when clipping the signal level of the saturated pixel, the signal level to be clipped is changed based on the light amount difference information of the image signal. Further, in the second embodiment, the signal level to be clipped is constant, and when performing the light amount difference correction based on the light amount difference information, it is determined whether or not the light amount difference correction is performed between the unsaturated pixels and the saturated pixels. changed. On the other hand, in the third embodiment, it is assumed that the saturation clip level can be set only to a certain value, and it is not possible to switch whether or not the light amount difference correction is applied to each pixel. In the system, a suitable AF can be executed by changing the method of detecting the saturation of the AF region and the method of calculating the defocus amount at the time of detection, or changing the selection method.

The configuration of the interchangeable lens type camera system including the lens and the camera body 30 in the present embodiment will be described with reference to FIG.

Since the lens unit 10 and its components in FIG. 18 are the same as the lens unit 10 and its components in FIG. 1, the description thereof will be omitted.

Since the components from the image sensor 301 to the timing generator 316 of the camera body 30 in FIG. 18 are the same as the components from the image sensor 201 to the timing generator 216 of the camera body 20 in FIG. 1, the description thereof will be omitted. However, for the saturation level clip unit / B signal generation unit 304, as in the operation of the second embodiment, it is sufficient that at least one clipping level can be set, and it is not necessary to be able to set for each pixel. Further, the light amount difference correction unit 305 may be able to correct the light amount difference of the image signal in the same manner as the operation of the first embodiment, and the light amount difference for each pixel depending on the presence or absence of saturated pixels as in the second embodiment. It is not necessary that the correction of is possible.

The saturation region determination unit 317 of FIG. 18 determines whether the AF region or the row subjected to the correlation calculation is saturated based on the number of pixels clipped as saturated pixels by the saturation level clipping unit / B signal generation unit 304. To do. The saturation determination value is set by the camera control unit 314, and the saturation determination value set at this time is changed based on the parameters of the lens unit 10 and the camera body 30 to change the ease of saturation determination. Control to do.

Further, the AF signal processing unit 306 changes the signal processing method based on the saturation information determined by the saturation region determination unit 317. Details will be described later based on the flowchart. Further, in the present embodiment, the AF signal processing unit 306 can change the type of bandpass filter applied to the image signal and the amount of pixel thinning out when calculating the correlation amount from the image signal. In calculating the correlation amount, the bandpass filter and the thinning amount of the pixels are changed, for example, in order to calculate the correlation amount having higher frequency characteristics and high focusing accuracy. Then, it is possible to calculate the correlation amount that can detect the amount of image shift even if the subject having the characteristic of lower frequency is blurred. In the present embodiment, the set bandpass filter and pixel thinning amount will be described with a configuration in which one correlation amount is calculated, but a configuration in which a plurality of correlation amounts in which the bandpass filter and pixel thinning amount are changed can be calculated at the same time is also possible. Good. Further, the camera control unit 314 changes the AF control method based on the saturation information of the AF region determined by the saturation region determination unit 317. Details will be described later based on the flowchart.

Next, the process performed by the camera body 30 in the present embodiment will be described.

The imaging process, saturation level setting process, and focus detection process in the present embodiment are shown in FIG. 3 in the first embodiment, shown in FIG. 16 in the second embodiment, and in the first embodiment. Since it is the same as that described with reference to FIG. 11, the description will be omitted.

The still image imaging process performed by the camera control unit 314 in S304 of FIG. 3 will be described with reference to the flowchart of FIG.

The processing of S1901, S1902, and S1904 to S1910 in FIG. 19 is the same as the processing of S401, S402, and S403 to S409 in FIG.

In S1903, the camera control unit 314 sets the saturation area determination threshold value set by the saturation area determination unit 317. The saturation region determination threshold setting process will be described later using the flowchart of FIG.

Next, the saturation region determination threshold value setting process performed by the camera control unit 314 in S1903 of FIG. 19 will be described with reference to the flowchart of FIG.

First, in S2001, the camera control unit 314 acquires the exit pupil distance information. The exit pupil distance information corresponds to the distance from the aperture 102 in the lens unit 10 to the image sensor 301 in the camera body 30, and the camera control unit 314 acquires it from the lens unit 10 via the lens control unit 106.

Next, in S2002, the camera control unit 314 determines whether or not the exit pupil distance acquired in S2001 is equal to or less than the predetermined value, proceeds to S2003 if it is equal to or less than the predetermined value, and proceeds to S2005 if it is larger than the predetermined value. Proceed.

Next, in S2003, which proceeds when the exit pupil distance is equal to or less than a predetermined value, the camera control unit 314 sets α as the saturation row determination threshold value.

Next, in S2004, the camera control unit 314 sets X as the saturation area determination threshold value and ends the saturation area determination threshold value setting process.

In S2005, which proceeds when the exit pupil distance is larger than a predetermined value in S2002, the camera control unit 314 sets β as the saturation row determination threshold value.

Next, in S2006, the camera control unit 314 sets Y as the saturation area determination threshold value and ends the saturation area determination threshold value setting process.

The saturated row determination threshold value is used to determine whether or not the row for which the correlation amount in the AF region is calculated is saturated. Counts how many pixels of the pair of image signals that make up the row for calculating the correlation amount have saturated clips, and if the number is greater than the saturated row judgment threshold, it is judged as a saturated row and the saturated row is determined. If it is less than or equal to the judgment threshold, it is judged that the row is not saturated. The determination result of the saturated row is used in the determination of the saturation region and the correlation amount addition determination process described later in the flowchart of FIG.

The saturation region determination threshold is used to determine whether or not the AF region is saturated. Count how many saturated rows are in the row that calculates the correlation amount that constitutes the AF region, and if the number of rows is greater than the saturation region determination threshold value, it is determined as a saturated region, and if it is less than or equal to the saturation region determination threshold value Judged as an unsaturated region. The determination result of the saturation region is used in the pixel thinning amount / bandpass filter setting process described later in the flowchart of FIG.

The relationship between the saturated row determination threshold value and the saturated region determination threshold value is preferably set to be “α <β” and “X <Y”. The smaller this value is, the easier it is to detect as saturation. That is, the shorter the exit pupil distance, the easier it is to detect as saturation. When the exit pupil distance is short, the difference in the amount of incident light of the image signal is generally large. When the incident light amount difference of the image signal becomes large, as shown in the pair of image signals after the incident light amount difference correction in FIG. 14, the level difference and the inclination difference between the pair of image signals are large and the image shift due to the correlation calculation. The accuracy of calculating the amount is reduced. However, as shown in FIG. 24, in the case where the incident light amount difference of the image signal is small, even if the phenomenon that the shape difference of the image signal after the incident light amount difference correction occurs occurs, the degree of occurrence of the shape difference is slight. Therefore, AF accuracy can be maintained. Therefore, the shorter the exit pupil distance, which tends to increase the difference in the amount of incident light, the easier it is to detect saturation, depending on the degree of influence on the AF accuracy, that is, the size of the difference in the amount of incident light. Take measures. On the contrary, on the side where the exit pupil distance, which tends to have a small difference in the amount of incident light, is long, if the AF accuracy can be maintained within an allowable range, it is controlled as usual without detecting saturation.

For example, when the difference in the amount of incident light becomes large in the range where the AF accuracy is unacceptable on the side where the exit pupil distance is long, the saturation row determination threshold value and the saturation region determination threshold value are made smaller so that saturation can be easily detected. You may control it.

Further, instead of the exit pupil distance information, the incident light amount difference information itself may be calculated, and the saturation row determination threshold value and the saturation region determination threshold value may be controlled to be switched based on the information of the incident light amount level difference and the inclination difference.

Next, the focus state detection process performed by the AF signal processing unit 306 in S1904 of FIG. 19 will be described with reference to the flowchart of FIG.

The processing of S2101, S2103, and S2105 to S2108 in FIG. 21 is the same as the processing of S601, S602, and S603 to S606 in FIG.

In S2102, the AF signal processing unit 306 changes the type of bandpass filter applied to the pair of image signals and the amount of pixel thinning. Details will be described later using the flowchart of FIG.

In S2104, the AF signal processing unit 306 determines whether or not to add the correlation amount to the correlation amount for each row in the AF region calculated in S2103. Details will be described later using the flowchart of FIG.

Next, using the flowchart of FIG. 22, the pixel thinning amount / bandpass filter setting process performed by the AF signal processing unit 306 in S2102 of FIG. 21 will be described.

First, in S2201, the AF signal processing unit 306 receives whether or not the saturation region determination unit 317 determines that the AF region is saturated, and if it is determined to be the saturation region, proceeds to processing to S2202 and proceeds to the saturation region. If it is determined that it is not, the process proceeds to S2204.

Next, in S2202, which proceeds when the saturation region is determined in S2201, the AF signal processing unit 306 sets m as the pixel thinning amount.

Next, in S2203, the AF signal processing unit 306 sets the frequency characteristic of the bandpass filter to f1 and ends the pixel thinning amount / bandpass filter setting process. The pixel thinning amount and bandpass filter set here are processed for the image signal acquired in S2101 of FIG. 21, and in the correlation calculation of S2103, the image signal after pixel thinning and bandpass filter setting processing is processed. Is done.

Next, in S2204, the AF signal processing unit 306 determines whether or not the reliability of the defocus amount calculated last time is equal to or higher than a predetermined value, and if the reliability is equal to or higher than a predetermined value, proceeds to processing to S2205 to improve the reliability. If it is worse than the predetermined value, the process proceeds to S2206. It is desirable that the reliability threshold value here is determined at a level at which the absolute value of the defocus amount calculated in the previous S2108 can be trusted to some extent. Further, if the process of S2204 is the first execution and the previously calculated defocus amount does not exist, the process is advanced to S2206.

Next, in S2205, the AF signal processing unit 306 determines whether or not the previously calculated defocus amount is within the predetermined range, proceeds to S2202 if it is within the predetermined range, and processes to S2206 if it is larger than the predetermined value. Proceed. If the previously calculated defocus amount does not exist as in S2204, the process proceeds to S2206.

Next, in S2206, the AF signal processing unit 306 sets the pixel thinning amount to n and proceeds to S2207.

Next, in S2207, the AF signal processing unit 306 sets the frequency characteristic of the bandpass filter to f2 and ends the pixel thinning amount / bandpass filter setting process.

The pixel thinning amounts m and n set in FIG. 22 are set so that m ≦ n. Further, the frequency characteristics f1 and f2 of the bandpass filter are set so that f1 has a higher frequency than f2.

When the saturated region is not determined in S2201, that is, in the basic control in which the saturated subject is not captured, it is determined in S2204 and S2205 whether or not the subject is close to the in-focus position to some extent. If it is close to the focusing position of the subject to some extent, in S2202 and S2203, the amount of pixel thinning is smaller, the bandpass filter is also a high frequency filter, and the defocus amount with higher focusing accuracy is calculated as described above. It can be so. This makes it possible to improve the AF accuracy when focusing on the subject. If the subject is not close to the in-focus position or the defocus amount cannot be detected in the first place, the pixel thinning amount is larger and the bandpass filter is lower frequency in S2206 and S2207. use. As a result, although the AF accuracy is lowered as described above, the defocus amount can be detected even in a more blurred state, and the in-focus position direction and the defocus amount of the subject can be easily detected even in the blurred area.

When the saturation region is determined in S2201 with respect to the above-mentioned basic control, S2202 and S2203 have higher AF accuracy without determining whether the subject is close to the in-focus position to some extent in S2204 and S2205. Control to make high settings. When it is determined to be a saturated region, that is, it is assumed that the shape difference of the pair of image signals as described in the graph of the pair of image signals after the incident light amount difference correction in FIG. 14 is remarkably generated. In such a case, if an effective defocus amount is used when the subject such as S2206 or S2207 is out of focus, the reliability values of the two-image coincidence and the steepness are not sufficiently improved. Then, in the determination of S2204 and S2205, it may not be possible to determine that the subject is close to the in-focus position to some extent. Therefore, problems such as the out-of-focus state and the continued blurring occur. When the saturation region is determined in S2201, the problem of out-of-focus as described above can be achieved by controlling S2202 and S2203 so that the pixel thinning amount and the bandpass filter with high AF accuracy can be set. Can be suppressed. In the present embodiment, the determination is made based on the saturation determination result of the AF region in S2201, but the determination may be made using the saturation determination result of the row in the AF region.

Next, the correlation amount addition process performed by the AF signal processing unit 306 in S2104 of FIG. 21 will be described with reference to the flowchart of FIG. 23.

First, in S2301, the AF signal processing unit 306 determines whether or not the saturation row determination threshold value set in S2003 and S2005 in FIG. 20 is equal to or higher than a predetermined value, and if it is equal to or higher than a predetermined value, proceeds to S2302 and less than a predetermined value. If, the process proceeds to S2304.

Next, in S2302, the AF signal processing unit 306 determines whether or not there is a row determined to be a saturated row by the saturation region determination unit 317 in each row for which the correlation amount in the AF region is calculated. Then, if there is a saturated row, the process proceeds to S2303, and if there is no saturated row, the process proceeds to S2304.

Next, in S2303, the AF signal processing unit 306 adds the correlation amounts other than the rows determined to be saturated rows among the rows for which the correlation amount in the AF region has been calculated, and ends the correlation amount addition determination processing.

Further, in S2304, the correlation amounts of all the rows calculated in the AF region are added to end the correlation amount addition determination process.

When all the rows are saturated in S2303, an initial value or the like that can determine that the correlation cannot be detected is set. The initial value may be, for example, a position where the focus is at infinity or close to infinity, or may be a position where the shooting frequency is high, such as 2 m.

For example, when a saturated subject exists in some rows in the AF region, the AF accuracy is improved by adding an erroneous correlation amount signal when the correlation amount for each row is added to obtain the correlation amount in the AF region. I want to avoid being affected by saturated subjects so that it does not decrease. Therefore, when there is a row determined to be a saturated row in S2302, the correlation amount of the rows other than the saturated row is controlled in S2303 to be added. However, in S2002, S2003, and S2005 of FIG. 20, the ease of detecting the saturated row is changed based on the magnitude of the incident light amount difference such as the exit pupil distance. In the process of adding the correlation amount of only the rows other than the saturated row of S2303, the saturated row determination threshold value set in S2301 is set to a predetermined value or more, that is, the saturation determination threshold value β of S2005 of FIG. 20 is set in the present embodiment. Is shown. On the other hand, when the saturation determination threshold value α is set in S2003 of FIG. 20, that is, in the case where the difference in the amount of incident light of the pair of image signals is large, S2304 is controlled so as to always add all the correlation amounts. ..

As shown in FIG. 25, when the difference in the amount of incident light between the pair of image signals is small, the AF accuracy can be maintained to some extent even if the saturated pixels are clipped. At this time, since the saturated row determination threshold value to be detected as a saturated row is set to more β, it is not determined as a saturated row unless many ranges in the row are detected as saturated pixels.

On the other hand, when the difference in the amount of incident light between the pair of image signals is large, the AF accuracy after performing the saturated pixel clipping process is significantly lowered. At this time, since the saturated row determination threshold value to be detected as a saturated row is set to less α, even if a saturated pixel is detected even in a small range in the row, it is immediately detected as a saturated row. Therefore, for example, if a small amount of saturated pixels are mixed in all the rows in the AF region and the processing of S2303 is applied, the correlation amount cannot be detected and the defocus amount cannot be known in many cases. It ends up. Then, the phenomenon of not being able to focus on the saturated subject described above occurs.

In order to solve this problem, when the saturated row determination threshold is small, that is, when a saturated row is detected even when there is a small saturated subject in the AF region, all the correlation amounts calculated in S2304 even if it is a saturated row. Is controlled to be added. Although the amount of correlation of saturated rows is added and used, by setting the characteristic that blurring does not easily occur in the saturated region by the pixel thinning amount / bandpass filter setting process in FIG. 22 described above, the saturated subject is used. On the other hand, the opportunity to focus can be improved. Further, the fact that the saturated subject in the line is small means that there is a high possibility that a subject other than the saturated subject exists in the AF region. Therefore, it is possible to increase the chances of focusing by controlling S2304 so that all the correlation amounts are added, rather than controlling S2303 so that the correlation amounts of the saturated rows are not added.

As described above, in the present embodiment, the saturation row determination is performed based on the number of pixels for which the saturation level clip processing is performed for each row in the AF region. In addition, the saturation region is determined based on the number of saturated rows in the AF region. Further, in performing these saturation determinations, the decrease in AF accuracy due to saturation increases as the difference in the amount of incident light between the pair of image signals increases. Therefore, the saturation line is based on the size of the difference in the amount of incident light and the exit pupil distance information in this embodiment. The judgment threshold of the above and the judgment threshold of the saturation region are changed. When the difference in the amount of incident light of the pair of image signals is small, it is easy to maintain the AF accuracy, so it is difficult to detect saturation. When the difference in the amount of incident light is large, there is a concern that the AF accuracy will decrease, so it will be easy to detect saturation. Furthermore, when it is detected as a saturated region, it is necessary to prevent the focus from stopping by using the characteristics of the pixel thinning number / bandpass filter that prioritizes the detection ability when blurring over the AF accuracy. , Always select the characteristic that gives priority to AF accuracy. Depending on the magnitude of the set determination threshold value of the saturated row, the calculated method of adding the correlation amount for each row is switched to a method suitable for each. When the saturated row determination threshold value is large, control is performed so that the saturated rows are not added in order to eliminate the influence of the saturated rows when the correlation amount is added for each row. In addition, when the saturated row judgment threshold is small, by controlling so as to add the correlation amounts of all rows, it is possible to suppress the occurrence of the phenomenon that AF cannot be performed without adding saturated rows and the focus cannot be adjusted. .. Furthermore, by using a non-saturated subject in the AF area or selecting a characteristic with higher AF accuracy in the pixel thinning number / bandpass filter setting process described above, the focus can be adjusted even in the AF area including the saturated subject. Improve the chances of being saturated.

As described above, it is possible to set the signal level threshold value of the saturation level clip processing for each pixel, and to set the correction presence / absence of the incident light amount difference correction for each pixel. Therefore, even in a system in which a high-load system cannot be realized, it is possible to improve the AF accuracy for a saturated subject and the opportunity for AF.

10 Lens unit 20 Camera body 101 Fixed lens 102 Aperture 103 Focus lens 104 Aperture drive unit 105 Focus lens drive unit 106 Lens control unit 107 Lens operation unit 21 Bus 201 Imaging element 202 CDS / AGC / AD
203 Image input controller 204 Saturation level clip unit / B signal generation unit 205 Light amount difference correction unit 206 AF signal processing unit 207 Display control unit 208 Display unit 209 Recording medium control unit 210 Recording medium 211 SDRAM
212 ROM
213 flash ROM
214 Camera control unit 215 Camera operation unit 216 Timing generator 317 Saturation region determination unit

Claims (9)

A focus detection device that detects the focus of an image pickup lens for photographing a subject based on a plurality of image signals acquired from an image sensor having a plurality of photoelectric conversion units for the micro lens.
A generation means for generating a plurality of focus detection signals using the plurality of image signals based on the amount of incident light from the image pickup lens, and
Correlation amount detecting means for detecting the correlation amount of the plurality of focus detection signals, and
A focus detection means that detects the focus of the imaging lens based on the correlation amount, and
A determination means for determining whether or not at least one of the plurality of image signals has reached a saturation level, and
A clipping means that clips the signal level of the image signal to a predetermined level based on the determination result of the determination means, and
An adjusting means for adjusting the difference in the image signal based on the difference in the amount of incident light for each of the plurality of photoelectric conversion units using predetermined adjustment parameters, and
A holding means for holding the predetermined adjustment parameter for adjusting the difference in the amount of incident light for each photoelectric conversion unit, and
The adjusting means controls the predetermined adjustment parameter for adjusting the difference between the image signals based on the determination result of the image height determining means.
The focus detection device is characterized in that the clip means controls the predetermined level for each photoelectric conversion unit according to the adjustment parameter held by the holding means . The adjustment parameter held by the holding means includes a gain amount for adjusting the difference in the amount of incident light.
The focus detection device according to claim 1, wherein the clip means controls the predetermined level based on the reciprocal of the gain amount. A holding means for holding the predetermined adjustment parameter for adjusting the difference in the amount of incident light for each photoelectric conversion unit is further provided.
The focus detection device according to claim 1, wherein the adjusting means adjusts the difference between the image signal according to the determination result of the determining means and the adjusting parameter held by the holding means. It said adjusting means focus detecting apparatus according to claim 3, characterized in that no correction of the incident light amount difference when it is determined to be saturated by the previous SL-size constant means. The focus detection device according to any one of claims 1 to 4, wherein the adjusting means adjusts a difference between the image signals after the clip means processes the image signal. The plurality of image signals are based on a first image signal based on a charge obtained by adding charges generated by the plurality of photoelectric conversion units and a charge generated by one of the plurality of photoelectric conversion units. Including the second image signal
Claims 1 to 1, wherein the plurality of focus detection signals generated by the generation means include a focus detection signal generated by subtracting the second image signal from the first image signal. 5. The focus detection device according to any one of 5. The focus detection device according to any one of claims 1 to 6, wherein the adjustment parameter includes the exit pupil distance of the image pickup lens or the incident light amount difference information for each photoelectric conversion unit. The correlation amount detecting means adds at least one or more correlation amounts or image signals and adds them.
The focus detection device according to any one of claims 1 to 7, wherein the focus detection means performs focus detection based on an added correlation amount or an image signal. An imaging device including the focus detection device according to any one of claims 1 to 8.

Images