Focus detection device and imaging device

Abstract

To provide a focus detection device capable of performing highly accurate focus detection while performing the subject blur correction.SOLUTION: A focus detection device (100) includes: first determination means (112) for determining a blur correction area based on the subject information; and second determination means (109) for determining a focus detection area based on the subject information and determining a calculation area for the focus detection from the focus detection area. The second determination means determines the calculation area based on a relation between the blur correction area and the focus detection area.SELECTED DRAWING: Figure 8

Description

Technical Field

[0004] , , , , , , , , , ,

[0005]

[0001] The present invention relates to a focus detection device and an imaging device.

Background Art

[0002] Conventionally, a focus detection device that detects a subject and performs focus detection has been known. Patent Document 1 discloses a focus detection method for setting a position and a field length for reading an image signal used for focus detection based on the position and size of a detected subject. Patent Document 2 discloses a method for avoiding a phenomenon (perspective competition) in which, when the focus detection area deviates from the subject, the focus is set on the background instead of the subject. Patent Document 3 discloses an image blur correction device that controls so that a subject is always within an angle of view with respect to blur (subject blur) occurring in a subject in a captured image.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Patent Document 3

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, if different controls are performed on the subject to be corrected for subject blur as disclosed in Patent Document 3 and the subject to be detected for focus detection as disclosed in Patent Documents 1 and 2, the subject is likely to deviate from the focus detection area. This is because subject blur correction negatively reflects the movement of the subject so as not to degrade the shooting quality, while focus detection performs control to continuously focus in a manner sensitive to the movement of the subject. <00×00030>

[0005] Therefore, the present invention aims to provide a focus detection device that can perform highly accurate focus detection while correcting subject blur. [Means for solving the problem]

[0006] A focus detection device as one aspect of the present invention includes a first determination means for determining a blur correction area based on subject information, and a second determination means for determining a focus detection area based on the subject information and determining a calculation area for focus detection from the focus detection area. , a calculation means for calculating the amount of image shift of the image signal in the calculation domain, The second determination means determines the calculation area based on the relationship between the blur correction area and the focus detection area.

[0007] Other objects and features of the present invention are described in the following examples. [Effects of the Invention]

[0008] According to the present invention, it is possible to provide a focus detection device that can perform highly accurate focus detection while correcting subject blur. [Brief explanation of the drawing]

[0009] [Figure 1] This is a block diagram of the imaging device in each embodiment. [Figure 2] The images show a front view of the unit pixel cell of the image sensor in each embodiment, and a diagram illustrating the two-dimensional arrangement of the unit pixel cells. [Figure 3] This diagram illustrates the principle of focus detection using an image sensor in each embodiment. [Figure 4] This is a timing chart showing the operation related to distance measurement of the imaging device in each embodiment. [Figure 5] This diagram illustrates the subjects detected by the subject detection unit and the subject detection information in each embodiment. [Figure 6] This is a flowchart showing the operation of the imaging device in each embodiment. [Figure 7] This flowchart shows the process for selecting the image stabilization target in each embodiment. [Figure 8] This flowchart shows the process for changing the distance measurement area of ​​the subject in each embodiment. [Figure 9] This is an explanatory diagram of the distance measurement area based on the relationship between the motion correction area and the distance measurement subject area in Example 1. [Figure 10] This is an explanatory diagram of the distance measurement area based on the relationship between the motion correction area and the distance measurement subject area in Example 2. [Modes for carrying out the invention]

[0010] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. [Examples]

[0011] First, with reference to Figure 1, the imaging device (focus detection device) 100 in Embodiment 1 of the present invention will be described. Figure 1 is a block diagram of the imaging device 100 in this embodiment. Note that in Figure 1, blocks that are not directly related to the features of this embodiment are not shown.

[0012] 101 is an optical system unit consisting of an optical lens group (imaging optical system) including a focusing lens for adjusting focus, a shutter, an aperture, a lens control unit, etc., and is driven and controlled by the lens drive control unit 111 described later. The lens drive control unit 111 outputs a drive signal for the focusing lens included in the optical system unit 101 according to the defocus amount calculated by the image shift amount calculation unit 110 described later. In this embodiment, the imaging device 100 is integrated with the optical system unit 101, but is not limited to this, and the optical system unit may be configured to be detachable from the imaging device.

[0013] The image sensor 102 is a photoelectric conversion element having unit pixel cells arranged in a two-dimensional matrix, and exposure control is performed by the shutter and aperture included in the optical system unit 101. The unit pixel cells of the image sensor 102 are driven to any position within the image circle projected by the optical system unit 101 by the image sensor drive control unit 114, which will be described later.

[0014] Here, referring to FIGS. 2(a) and (b), the configuration of the image pickup device 102 will be described. FIG. 2(a) is a front view of the unit pixel cell 200 of the image pickup device 102. FIG. 2(b) is an array configuration diagram of the two-dimensional unit pixel cells 200. As shown in FIG. 2(b), the image pickup device 102 has unit pixel cells 200 repeatedly arranged in a Bayer array.

[0015] As shown in FIG. 2(a), the unit pixel cell 200 has a microlens 2 and pixels 1a and 1b including a photodiode. The pixels 1a and 1b are split PDs that take in incident light through the same microlens 2. The output of each split PD can obtain an image that is pupil-split according to its arrangement. Hereinafter, the image (image signal) obtained by the pixel 1a is referred to as the A image (A image signal), and the image (image signal) obtained by the pixel 1b is referred to as the B image (B image signal).

[0016] The A / D conversion unit 103 converts the analog electrical signal output from the image pickup device 102 into a digital electrical signal (pixel signal) after performing analog signal processing in an analog signal processing unit (not shown), and outputs the pixel signal to the capture unit 104. The analog signal processing unit is a CDS circuit, a non-linear amplification circuit, etc. that remove noise on the transmission line. The capture unit 104 determines the valid period and type of the pixel signal, and outputs the pixel signals of the A image and the B image to the digital signal processing unit 105, the subject detection unit 108, the image shift amount calculation unit 110, and the motion correction amount calculation unit 113.

[0017] The digital signal processing unit 105 performs addition processing between split pixels on the A image and the B image input in a Bayer array, and performs digital signal processing represented by synchronization processing, gamma processing, and noise reduction processing. The image data to which the digital signal processing is applied is converted into compressed image data represented by JPEG and output to the external recording device 106 for recording. Note that since the technologies of synchronization processing, gamma processing, noise reduction processing, and JPEG compression are not directly related to this embodiment, their descriptions are omitted.

[0018] The subject detection unit 108 detects a subject in the captured image data (pixel signal), adds information about the subject, and outputs it to the distance measurement area selection unit 109 and the motion correction area selection unit 112.

[0019] Here, with reference to Figures 5(a) and 5(b), the subjects detected by the subject detection unit 108 and the subject detection information will be explained. Figure 5(a) is an explanatory diagram of the subjects detected by the subject detection unit 108. Figure 5(b) is an explanatory diagram of the subject detection information.

[0020] Figure 5(a) shows a case where there are two subjects in the field of view. The subject detection unit 108 assigns subject detection information to the field of view as shown in Figure 5(b). Here, the index of each subject, an index linked if the same subject was present in the previous frame, and size information, which is information about the detected coordinates (position), width, and height, are assigned. The subject detection unit 108 also assigns the detection confidence level and subject type for each subject. For example, the detection confidence level is lowered for frames where the subject is facing away or to the side, compared to the detection level for a frontal face. The subject type information includes information such as the detected person, vehicle, or animal, and also assigns part information such as the pupil area, entire face, profile, and head.

[0021] Furthermore, the distance measurement area selection unit 109 and the motion correction area selection unit 112 control the selection of the calculation area (distance measurement calculation area) for the next frame based on the subject detection result for the captured image of the previous frame. This makes it possible to shorten the feedback time for the lens drive control described later. For this reason, it is not necessary to refer to the subject detection result of the same frame.

[0022] The motion correction area selection unit 112 selects (determines) the area (blur correction area) for correcting the angle of view motion based on the subject information output from the subject detection unit 108, and outputs information about the selected area to the motion correction amount calculation unit 113 and the distance measurement area selection unit 109. If the motion correction target is continuously switched to the detected subject with the highest reliability, the angle of view observed by the user may become unstable and the quality of the image may decrease significantly. For this reason, the motion correction area selection unit 112 controls the system so that the subject to be corrected for motion does not change frequently from the subject index of the previous frame to the subject with high reliability. In addition, if the reliability of all detected subjects is low, motion correction can be performed without lowering the quality of the user's observation by performing only camera shake correction (correction of the entire angle of view).

[0023] The motion correction amount calculation unit 113 calculates the amount of motion (motion vector) between the previous frame and the current frame of the region selected by the motion correction area selection unit 112 based on a correlation calculation such as SAD, and outputs the calculated motion vector to the image sensor drive control unit 114. The image sensor drive control unit 114 drives the image sensor 102 based on the motion vector output from the motion correction amount calculation unit 113 and restricts (controls) the readout range so that the region selected by the motion correction area selection unit 112 remains stationary in the center of the field of view.

[0024] In this embodiment, the image sensor 102 is controlled to read out any optical image within the image circle by moving it, but for example, the position of the image read out from the image sensor 102 may be specified. Alternatively, the capture unit 104 may be controlled to crop and output only a portion of the range read out from the image sensor 102 across the entire field of view.

[0025] The distance measurement area selection unit 109 determines (selects) the coordinates (position) of the distance measurement area and the area (calculation area) in which the correlation calculation will be performed by the image shift amount calculation unit 110, described later, based on the subject information output from the subject detection unit 108 and the information of the area selected by the motion correction area selection unit 112. The distance measurement area selection unit 109 then outputs the information of the distance measurement area (calculation area) to the image shift amount calculation unit 110.

[0026] Here, with reference to Figures 9(a) to (c), the distance measurement area (calculation area) determined by the distance measurement area selection unit 109 based on subject information and motion correction area (blur correction area) will be explained. Figures 9(a) to (c) are explanatory diagrams of the distance measurement area (calculation area) based on the relationship between the motion correction area and the distance measurement subject area. In Figures 9(a) to (c), the dashed line indicates the motion correction area, the dotted line indicates the distance measurement subject area (focus detection area), and the distance measurement area (calculation area) is shown as a gray area enlarged to the right of each figure.

[0027] Figure 9(a) shows the case where the motion correction area 901a represents the entire field of view and the distance measurement subject area 902a is within the field of view (when the distance measurement subject area 902a is part of the motion correction area 901a). In this case, subject blur makes it easier to lose track of the area in which the subject is being measured, so the calculation area 903a is controlled to be smaller than the distance measurement subject area 902a. Note that if the motion correction function is turned off by a user instruction member (not shown), image stabilization is also disabled, making it even easier to lose track of the area in which the subject is being measured. In such cases, the calculation area 903a may be controlled to be even smaller.

[0028] Figure 9(b) shows the case where the motion correction area 901b and the distance measurement subject area 902b are approximately the same. In this case, the displacement of the subject position within the field of view tends to stabilize due to camera shake and subject blur. Therefore, control is performed to widen the calculation area 903b.

[0029] Figure 9(c) shows the case where the motion correction area 901c and the distance measurement subject area 902c are different. In this case, the displacement of the subject position within the field of view tends to become unstable due to subject blur. Therefore, the calculation area 903c is controlled to be smaller than the distance measurement subject area 902c.

[0030] Next, I will explain the reason for this control. The subject selected by the distance measurement area selection unit 109 for focusing is selected independently of motion correction. For example, in sports where multiple people compete together, it is desirable that the subject's face is exposed to the imaging device 100, and if only the back of the head is photographed, the value of the captured image tends to be low even if focus is achieved. Therefore, by setting a high sensitivity to the subject for which motion correction is performed, it becomes possible to obtain a high-value captured image.

[0031] Next, we will explain the difference in effects between making the calculation area smaller and making it larger. When the calculation area is made smaller, and the XY coordinates (position) of the focused subject in the next frame are predicted based on the subject detection result of the previous frame, and the calculation area is selected and performed, there is an effect of not mixing subjects at different distances within the calculation area. Therefore, it is possible to reduce the possibility of mistakenly focusing on the background and foreground included in the calculation area. On the other hand, when the calculation area is made larger, more calculation areas (distance measurement calculation areas) can be allocated to the same subject. Therefore, it is possible to reduce the influence of elements that are difficult for distance measurement calculations, such as high ISO, low light, and low contrast, and improve focusing accuracy.

[0032] The image shift amount calculation unit 110 calculates the image shift amount between image A (image A signal) and image B (image B signal) output from the capture unit 104, and outputs the calculated image shift amount to the lens drive control unit 111. The lens drive control unit 111 converts the image shift amount into a defocus amount and drives the focus lens of the optical system unit 101 (performs focus control).

[0033] The image shift between image A and image B is calculated for each rectangular region obtained by dividing the unit pixel cells 200, which are arranged in a two-dimensional matrix as shown in Figure 2(b), into N×M sections. The region used for the calculation is determined according to the output of the distance measurement region selection unit 109 of the previous frame. Furthermore, the position of the subject is selected based on the coordinates (position) of the distance measurement region of the current frame from the obtained N×M distance measurement calculation results. This type of control makes it possible to perform lens drive control with low latency (low delay time) without waiting for processing by the subject detection unit 108 of the current frame. The range and position of the distance measurement region are determined according to the output signal from the distance measurement region selection unit 109.

[0034] Here, the principle of focus detection using phase difference will be explained with reference to Figures 3(a) to 3(c). Figures 3(a) to 3(c) are explanatory diagrams of the principle of focus detection using the image sensor 102. The image sensor 102 shown in Figures 3(a) to 3(c) has a plurality of unit pixel cells 200 as explained with reference to Figures 2(a) and 2(b). The positions of the unit pixel cells 200 are indicated by unit pixel cells P1 to P13, respectively. Also, pixels 1a and 1b of the unit pixel cells 200 are indicated as pixels A and B, respectively.

[0035] Multiple unit pixel cells P, each containing divided pixels A and B, are arranged beneath a single microlens 2. Pixels A and B, positioned beneath the microlens 2, are pixels divided into pupils, with the microlens 2 acting as the exit pupil. During distance measurement (focus detection), signals output from multiple pixels A (pixel group for image A) and signals output from multiple pixels B (pixel group for image B) are combined in the column (or row) direction to generate images A and B as outputs of the same-color unit pixel cell groups. The offset of corresponding points is then calculated using SAD calculation. The result of the SAD calculation is obtained by the following equation (1).

[0036] C = Σ|YAn - YBn| …(1) In equation (1), n ​​is the number of microlenses in the range where horizontal SAD (Single-Aspected Diode) calculations (correlation calculations) are performed. YAn and YBn are the luminance signals of image A and image B, respectively, corresponding to the nth microlens. The value obtained by shifting the corresponding pixel relative to the luminance signal YBn of image B is plotted, and the position where the amount of shift is minimized is the focus position.

[0037] In the focused state shown in Figure 3(a), the imaging optical system forms an image at pixels A and B under the microlens of unit pixel cell P7, so the pixel group for image A and the pixel group for image B are approximately the same. This indicates that the image shift amount d(a) between the pixel group for image A and the pixel group for image B, which is obtained by correlation calculation, is approximated to 0.

[0038] In the back-focused state shown in Figure 3(b), the imaging optical system forms images at the following positions: pixels A and B under the microlenses of unit pixel cell P5 for image A and unit pixel cell P9 for image B. At this time, an image shift amount d(b) occurs between the pixel group for image A and the pixel group for image B, which is determined by correlation calculation.

[0039] In the front-focused state shown in Figure 3(c), the pixels for image A are located under the microlenses of unit pixel cell P9, and the pixels for image B are located under the microlenses of unit pixel cell P5, as the position where the imaging optical system forms an image. At this time, the image displacement amount d(c) between the pixel group for image A and the pixel group for image B, which is determined by correlation calculation, is the opposite of the image displacement amount in the back-focused state shown in Figure 3(b).

[0040] This means that in the in-focus state, the pixel group for image A and the pixel group for image B are looking at the same subject, but in the back-focused and front-focused states, the pixel group for image A and the pixel group for image B are looking at a subject shifted by an image displacement amount d. Based on the minimum image displacement amount d obtained from the correlation calculation result and the baseline length, the amount of defocus can be determined using well-known techniques, and the focusing operation on the subject can be performed.

[0041] Next, referring to Figure 6, the operation of the imaging device 100 from the detection of a subject by the subject detection unit 108 during live view shooting to the completion of lens drive control unit 111 is described. Figure 6 is a flowchart showing the operation of the imaging device.

[0042] First, in step S600, the subject detection unit 108 detects a subject and generates subject information. Next, in step S601, the motion correction area selection unit 112 refers to the subject and subject information detected in step S600 to determine the subject and the blur correction area to be corrected (selection of the blur correction target).

[0043] Here, with reference to Figure 7, the process of step S601 (selection of image stabilization target) will be described in detail. Figure 7 is a flowchart of the image stabilization target selection process.

[0044] First, in step S700, the motion correction area selection unit 112 checks the reliability of all detected subjects and assigns flag information indicating a candidate for motion correction to subjects whose reliability value exceeds a threshold. Next, in step S701, the motion correction area selection unit 112 checks the type of subject and determines whether the subject is of a type that should be subject for motion correction. If the subject is of a type that should be subject for motion correction, the motion correction target candidate flag is passed. On the other hand, if the subject is not of a type that should be subject for motion correction, the motion correction target flag is changed to invalid. Here, the type of subject that should be subject for motion correction is a person, animal, vehicle, etc., selected by a user-indicated member (not shown).

[0045] Next, in step S702, the motion correction area selection unit 112 determines whether the same subject that was determined to undergo motion correction in the previous frame exists among the subjects for which the motion correction target candidate flag is enabled in the current frame. If the same subject exists, it is selected as the subject for motion correction. On the other hand, if the same subject does not exist, the motion correction target subject with the highest reliability is selected (subject transfer confirmation). If all motion correction target candidate flags are disabled, the entire field of view may be selected as the motion correction target subject.

[0046] Next, in step S703, the motion correction amount calculation unit 113 calculates a motion vector relating to the region of the subject to motion correction in the previous frame, based on the region information of the subject to motion correction. If the same subject does not exist in the previous frame, the value of the motion vector is set to 0.

[0047] Next, in step S704, the motion correction amount calculation unit 113 predicts the amount of subject blur in the next frame from the motion vector amount of the previous frame and the motion vector amount of the current frame, and determines the subject blur correction position. A correction amount is calculated that allows the sensitivity to be controlled using a low-pass filter so that the correction value does not become a steep value due to the predicted amount of blur, and this flow ends.

[0048] Next, in step S602 of Figure 6, the image sensor drive control unit 114 controls the image sensor 102 to complete its drive before the exposure of the next frame begins, based on the predicted amount of subject blur (subject blur correction). Subsequently, in step S603, the motion correction amount calculation unit 113 notifies the distance measurement area selection unit 109 of the selected motion correction subject (blur correction subject) that it refers to.

[0049] Next, in step S604, the distance measurement area selection unit 109 selects a subject for distance measurement (distance measurement subject) from the subjects acquired by the subject detection unit 108. The selection of the distance measurement subject is controlled to make it easier to select a subject in relation to the selection of the blur correction target performed in step S601. In addition, the control is set to give greater importance to the priority given to the type of subject. For example, if a person's face is detected and a person's pupil is detected, the pupil is given greater priority.

[0050] Next, in step S605, the distance measurement area selection unit 109 selects a valid calculation result from the distance measurement results that have been pre-calculated by N × M, based on the subject information acquired in step S600 and the information of the blur-correcting subject (correction area) generated in step S603. In other words, the distance measurement area selection unit 109 changes the distance measurement area (calculation area) based on the subject information and the correction area information.

[0051] Here, with reference to Figure 8, the process of step S605 (changing the distance measurement area) will be described in detail. Figure 8 is a flowchart of the process of changing the distance measurement area.

[0052] First, in step S800, the distance measurement area selection unit 109 determines whether the subject blur correction area (blur correction area) is the entire field of view (the entire captured image). If the blur correction area is the entire field of view, the process proceeds to step S802. On the other hand, if the blur correction area is not the entire field of view, that is, if a specific subject is to be corrected, the process proceeds to step S801.

[0053] In step S801, the distance measurement area selection unit 109 determines whether the blur correction area and the distance measurement area represent the same area. If the blur correction area and the distance measurement area represent the same area, the process proceeds to step S803. On the other hand, if the blur correction area and the distance measurement area do not represent the same area, the process proceeds to step S802. Here, the same area includes not only areas that are perfectly identical, but also areas that are evaluated as substantially identical (approximately identical areas). Note that if the entire field of view is not considered the blur correction area, the distance measurement area may represent a part of the blur correction area.

[0054] In step S802, the distance measurement area selection unit 109 narrows the range of the distance measurement area to a size smaller than the subject detection position (subject area) acquired by the subject detection unit 108, and then terminates this flow. By setting the distance measurement area to be smaller than the subject area, it is possible to reduce the mixing of subjects at multiple distances within the distance measurement area.

[0055] In step S803, the distance measurement area selection unit 109 sets the range of the distance measurement area to be the same as the subject detection position (subject area) acquired by the subject detection unit 108, and terminates this flow. Here, the same range includes not only a range that is exactly the same, but also a range that is evaluated as substantially the same (approximately the same range). By setting a wide distance measurement area (to be the same as the subject area), as described above, it becomes possible to take a wide calculation range for subjects with high ISO, low light, and low contrast, and it becomes possible to improve the focusing accuracy.

[0056] Next, in step S606 of Figure 6, the image shift amount calculation unit 110 calculates the image shift amount based on the distance measurement area output from the distance measurement area selection unit 109 (distance measurement calculation). Subsequently, in step S607, the lens drive control unit 111 converts the image shift amount into a defocus amount and drives the optical system unit 101 (lens drive), thus ending this flow.

[0057] Next, with reference to Figure 4, the sequence from exposure to lens drive control of the imaging device 100 will be described. Figure 4 is a timing chart showing the operation related to distance measurement of the imaging device 100.

[0058] First, at timing t400, the image sensor drive control unit 114 uses the image sensor 102 to perform an exposure operation on the image formed via the optical system unit 101. Subsequently, at timing t401, the image sensor drive control unit 114 completes the exposure operation by the image sensor 102 and performs a readout operation. That is, it outputs the signal from the image sensor 102 to the subject detection unit 108, the image displacement amount calculation unit 110, and the motion correction amount calculation unit 113 via the A / D conversion unit 103 and the capture unit 104. Subsequently, at timing t402, the subject detection unit 108 performs subject detection processing. At approximately the same timing, the image displacement amount calculation unit 110 performs calculation processing on a rectangular region divided into N × M sections.

[0059] Next, at timing t403, the motion correction amount calculation unit 113 calculates the subject blur correction amount based on the output of the subject detection unit 108 and notifies the distance measurement area selection unit 109 of information regarding the subject for which subject blur correction has been performed (information on the subject subject to motion correction). Next, at timing t404, the image sensor drive control unit 114 starts driving the image sensor 102 to a position within the image height and completes the drive by timing t405, which is the exposure start timing for the next frame. At timing t405, the image sensor 102 performs the exposure operation for the next frame on the image formed via the optical system unit 101. At the following timings t406 and t407, the same processing as at timings t401 and t402 is performed.

[0060] Next, at timing t408, the motion correction amount calculation unit 113 calculates the subject blur correction amount based on the output of the subject detection unit 108 and notifies the distance measurement area selection unit 109 of information regarding the subject subject for which the subject blur correction was performed. The distance measurement area selection unit 109 then selects the subject information for motion correction obtained at timing t403 and the distance measurement calculation result obtained at timing t407 from the calculation results of the rectangular area divided into N×M at timing t407.

[0061] Next, at timing t409, the image sensor drive control unit 114 and the lens drive control unit 111 start driving the image sensor 102 to a position within the image height and starting the focusing drive so that the optical system unit 101 focuses on the subject. Then, they complete the drive until timing t405, which is the exposure start timing for the next frame. The operation from timing t410 onward is a repetition of the operation from timing t405 to t409.

[0062] According to this embodiment, it is possible to perform highly accurate focus detection while correcting for subject blur and reducing the frequency of the subject being removed from the distance measurement area. [Examples]

[0063] Next, Embodiment 2 of the present invention will be described. Embodiment 1 described the selection control of the distance measurement area (calculation area) when the subject to motion correction (motion correction area) and the subject to distance measurement (distance measurement subject area) are different. In this embodiment, with reference to Figures 10(a) and (b), the selection control of the calculation area when a relationship is recognized between the motion correction area and the distance measurement subject area will be described. Figures 10(a) and (b) are explanatory diagrams of the distance measurement area based on the relationship between the motion correction area and the distance measurement subject area. In Figures 10(a) and (b), the dashed line shows the motion correction area, the dotted line shows the distance measurement subject area, and the calculation area is shown as a gray area enlarged to the right of each figure.

[0064] Figure 10(a) shows the case where the motion correction area 1001a is a face and the distance measurement subject area 1002a is the pupil. In this case, even if the pupil area is excluded due to camera shake and subject blur, it can be inferred that the surrounding area is at the same distance, so the calculation area 1003a where the distance measurement calculation is performed is controlled to be the same size as or wider than the distance measurement subject area 1002a. In this embodiment, the relationship between a person's face and pupil is used for explanation, but it is not limited to this, and the same applies, for example, if the motion correction target area is a vehicle and the area where the distance measurement calculation is performed is the relationship between a person riding in the vehicle. In other words, this control is performed when the motion correction area is an area that does not indicate the entire field of view, and a part of it is designated as the distance measurement subject position.

[0065] Figure 10(b) shows the case where the motion correction area 1001b is the torso and the distance measurement subject area 1002b is the face (when the motion correction area 1001b and the distance measurement subject area 1002b are related subject types). Generally, the torso represents a larger area than the face and its movement is more stable. Therefore, it is suitably selected as the motion correction area. On the other hand, generally, the subject for which distance measurement is to be performed is often the face. In this case, it is assumed that the XY coordinates within the image height of the face move in accordance with the movement of the torso, and it is inferred that this indicates the same subject. The calculation area 1003b is controlled to be the same size as or wider than the face area (distance measurement subject area 1002b). These controls can be performed based on coordinates (position), or they can be determined by adding information indicating the relationship between subjects to the index information at the time of subject detection.

[0066] As described above, in each embodiment, the focus detection device (imaging device 100) has a first determination means (motion correction area selection unit 112) and a second determination means (distance measurement area selection unit 109). The first determination means determines the blur correction area based on subject information. The second determination means determines the focus detection area (distance measurement subject area) based on subject information and determines the calculation area for focus detection (distance measurement area) from the focus detection area. The second determination means also determines the calculation area based on the relationship between the blur correction area and the focus detection area.

[0067] Preferably, the second determination means determines the calculation area based on the positional relationship between the blur correction area and the focus detection area. Also preferably, the first and second determination means acquire subject information from the output signal of the image sensor 102. Also preferably, the focus detection device has correction means (motion correction amount calculation unit 113, image sensor drive control unit 114) that perform blur correction based on the blur correction area. Also preferably, the focus detection device has calculation means (image shift amount calculation unit 110) that calculates the amount of image shift of the image signal in the calculation area.

[0068] Preferably, the subject information includes at least one of the following: the position (coordinates), size, confidence level, or type of the subject. Also preferably, if the blur correction area is the entire image, the calculation area is smaller than the subject in the subject information. Also preferably, if the blur correction area and the focus detection area are the same (or substantially the same), the calculation area is larger than if they were different. Also preferably, if the focus detection area is a part of the blur correction area, the calculation area is larger than if they were different. Also preferably, if the blur correction area and the focus detection area are related subject types, the calculation area is larger than if they were different.

[0069] In this way, by performing processing based on the relationship between the motion correction area and the distance measurement subject area, it becomes possible to appropriately select the calculation area (distance measurement area) even if the motion correction area and the distance measurement subject area are set to different areas. As a result, it becomes possible to perform focus detection while correcting subject blur, while reducing the frequency of the subject being removed from the distance measurement area.

[0070] (Other examples) The present invention can also be realized by supplying a program that implements one or more of the functions of the above-described embodiments to a system or device via a network or storage medium, and by having one or more processors in the computer of that system or device read and execute the program. It can also be realized by a circuit (e.g., an ASIC) that implements one or more functions.

[0071] According to the present invention, it is possible to provide a focus detection device, an imaging device, a focus detection method, and a program that enable highly accurate focus detection while correcting subject blur and reducing the frequency of the subject being removed from the distance measurement area.

[0072] Although preferred embodiments of the present invention have been described above, the present invention is not limited to these embodiments, and various modifications and changes are possible within the scope of its gist. [Explanation of symbols]

[0073] 100 Imaging device (focus detection device) 109 Distance measurement area selection unit (second determination means) 112 Motion correction area selection unit (first determination means)

Claims

1. A first determination means for determining the blur correction area based on subject information, A second determination means that determines a focus detection region based on the subject information and determines a calculation region for focus detection from the focus detection region, It includes a calculation means for calculating the amount of image shift of the image signal in the calculation domain, The second determination means is a focus detection device characterized by determining the calculation area based on the relationship between the blur correction area and the focus detection area.

2. The focus detection device according to claim 1, characterized in that the second determination means determines the calculation area based on the positional relationship between the blur correction area and the focus detection area.

3. The focus detection device according to claim 1 or 2, characterized in that, when the blur correction area is the entire image, the calculation area is smaller than the subject of the subject information.

4. The focus detection device according to any one of claims 1 to 3, characterized in that when the blur correction area and the focus detection area are the same, the calculation area is larger than when the blur correction area and the focus detection area are different.

5. The focus detection device according to any one of claims 1 to 4, characterized in that when the focus detection area is part of the image stabilization area, the calculation area is larger than when the image stabilization area and the focus detection area are different.

6. The focus detection device according to any one of claims 1 to 5, characterized in that when the image stabilization area and the focus detection area are of the same subject type corresponding to a part of the same subject, the calculation area is larger than when the image stabilization area and the focus detection area are different.

7. A first determination means for determining the blur correction area based on subject information, The system includes a second determination means for determining a focus detection region based on the subject information and determining a calculation region for focus detection from the focus detection region, The focus detection device is characterized in that, when the blur correction area is the entire image, the calculation area is smaller than the subject of the subject information.

8. A first determination means for determining the blur correction area based on subject information, The system includes a second determination means for determining a focus detection region based on the subject information and determining a calculation region for focus detection from the focus detection region, A focus detection device characterized in that, when the image stabilization region and the focus detection region are the same, the calculation region is larger than when the image stabilization region and the focus detection region are different.

9. A first determination means for determining the blur correction area based on subject information, The system includes a second determination means for determining a focus detection region based on the subject information and determining a calculation region for focus detection from the focus detection region, A focus detection device characterized in that, when the focus detection area is part of the image stabilization area, the calculation area is larger than when the image stabilization area and the focus detection area are different.

10. A first determination means for determining the blur correction area based on subject information, The system includes a second determination means for determining a focus detection region based on the subject information and determining a calculation region for focus detection from the focus detection region, A focus detection device characterized in that, when the image stabilization area and the focus detection area are of the same subject type corresponding to a part of the same subject, the calculation area is larger than when the image stabilization area and the focus detection area are different.

11. The focus detection device according to any one of claims 1 to 10, characterized in that the first determination means and the second determination means acquire the subject information from the output signal of the image sensor.

12. The focus detection device according to any one of claims 1 to 11, further comprising correction means for performing blur correction based on the blur correction region.

13. The focus detection device according to any one of claims 1 to 12, characterized in that the subject information includes at least one of the following: the position, size, reliability, or type of the subject.

14. Image sensor and An imaging device characterized by having a focus detection device according to any one of claims 1 to 13.

15. A first determination step in which the blur correction area is determined based on subject information, A second determination step involves determining a focus detection region based on the subject information and determining a calculation region for focus detection from the focus detection region, The calculation step includes calculating the amount of image shift of the image signal in the calculation domain, A focus detection method characterized in that, in the second determination step, the calculation area is determined based on the relationship between the blur correction area and the focus detection area.

16. A first determination step in which the blur correction area is determined based on subject information, The system includes a second determination step of determining a focus detection region based on the subject information and determining a calculation region for focus detection from the focus detection region. A focus detection method characterized in that, when the blur correction area is the entire image, the calculation area is smaller than the subject of the subject information.

17. A first determination step in which the blur correction area is determined based on subject information, The system includes a second determination step of determining a focus detection region based on the subject information and determining a calculation region for focus detection from the focus detection region. A focus detection method characterized in that, when the blur correction region and the focus detection region are the same, the calculation region is larger than when the blur correction region and the focus detection region are different.

18. A first determination step in which the blur correction area is determined based on subject information, The system includes a second determination step of determining a focus detection region based on the subject information and determining a calculation region for focus detection from the focus detection region. A focus detection method characterized in that, when the focus detection area is part of the image stabilization area, the calculation area is larger than when the image stabilization area and the focus detection area are different.

19. A first determination step in which the blur correction area is determined based on subject information, The system includes a second determination step of determining a focus detection region based on the subject information and determining a calculation region for focus detection from the focus detection region. A focus detection method characterized in that, when the blur correction area and the focus detection area are of the same subject type corresponding to a part of the same subject, the calculation area is larger than when the blur correction area and the focus detection area are different.

20. A program characterized by causing a computer to execute the focus detection method described in any one of claims 15 to 19.

Images