JP5653080B2 - Focus detection apparatus and focus detection method - Google Patents

Description

  The present invention relates to a focus detection apparatus such as a camera.

  2. Description of the Related Art Conventionally, a camera automatic focus detection apparatus using a phase difference detection method is known. In the phase difference detection method, a light beam from a subject that has passed through different exit pupil regions of the photographing lens is imaged on a pair of AF sensors, and a relative position between a pair of subject images obtained by photoelectric conversion is calculated. (Hereinafter, referred to as phase difference calculation), the defocus amount of the taking lens is detected.

  There is also a subject tracking automatic focus detection device using a multi-part light receiving sensor other than the AF sensor. The subject tracking automatic focus detection apparatus stores a tracking target image in a photographed image as a reference image, and has a subject recognition function for detecting a subject position from repeatedly captured images and reference images, and performs focus detection on the subject position. .

  Patent Document 1 discloses an apparatus that automatically tracks a moving object and performs focus detection or focus adjustment. The apparatus of Patent Document 1 tracks a moving subject position using an image from a second imaging element different from the ranging element, and uses a signal from the ranging element corresponding to the detected subject position. Adjust the focus.

JP 2009-010672A

  However, the focus detection apparatus of Patent Document 1 requires a second image sensor, which increases costs. Even if the photometric sensor and the subject recognition sensor are used together, the photometric sensor needs to be configured with multi-divided pixels in order to realize the subject recognition function. For this reason, the cost is significantly increased as compared with the conventional photometric sensor.

  Accordingly, the present invention provides a focus detection apparatus having a subject recognition function and a tracking function at low cost.

A focus detection device according to one aspect of the present invention includes an area sensor that has a light-shielding pixel and accumulates an image signal obtained from a light beam that has passed through a photographing lens in a plurality of pixels, and the area sensor is a first block. Selection means for selecting one pattern from a first pattern to be configured and a second pattern to be configured as a second block larger than the first block; and a pixel signal of the first pattern selected by the selection means First detection means for detecting the first signal using the largest signal and the light-shielded pixel signal obtained from the light-shielded pixel, and the pixel signal in each block of the second pattern selected by the selection means Second detection means for detecting the second signal using the largest signal and the smallest signal,
Based on the storage time, storage control means for controlling the storage time of the first block based on the first signal, control of the storage time of the second block based on the second signal, and Storage means for storing the pixel signal of the first pattern obtained as a subject recognition signal and storing the pixel signal of the second pattern obtained based on the accumulation time as a focus detection signal; Calculation means for recognizing the position of the subject based on the recognition signal and calculating a defocus amount based on the focus detection signal in one block of the second pattern corresponding to the position of the subject.

  Other objects and features of the present invention are illustrated in the following examples.

  According to the present invention, it is possible to provide a focus detection device having a subject recognition function and a tracking function at low cost.

It is a block diagram of the AF sensor in a present Example. It is a block diagram of the camera in a present Example. It is an optical block diagram of the camera in a present Example. It is an optical block diagram in the case of the focus detection in a present Example. FIG. 3 is an arrangement diagram of area sensors in the AF sensor of the present embodiment. FIG. 6 is a layout diagram of a focus detection visual field and a subject recognition visual field in the present embodiment. It is a flowchart of the focus detection operation | movement in a present Example. It is a subroutine flowchart of the accumulation | storage operation | movement in a present Example. It is explanatory drawing of the PB signal process in a present Example. It is an example of the image of the focus detection operation | movement in a present Example. It is explanatory drawing of the output process of the signal for focus detection in a present Example.

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

  First, the camera in the present embodiment will be described. FIG. 2 is a block diagram showing the configuration of the camera in this embodiment. A signal input circuit 204 for detecting signals from various switch groups 214 provided in the camera is connected to the camera microcomputer (hereinafter referred to as “CPU”) 100. The CPU 100 is connected to an image sensor 206, a photometric sensor 207 (AE sensor), a shutter control circuit 208 for controlling the shutter magnets 218a and 218b, and the AF sensor 101. Further, the CPU 100 transmits a signal 215 to and from a photographing lens (not shown) via the lens communication circuit 205, and controls the focal position and the diaphragm. The operation of the camera is determined by the setting of the switch group 214.

  The AF sensor 101 has an area sensor configured by arranging pixels in two dimensions. The CPU 100 detects the defocus amount based on the contrast distribution of the subject obtained by the area sensor, and controls the focal position of the photographing lens. The AF sensor 101 also has a function of capturing an image for subject recognition.

  The CPU 100 determines the aperture value and shutter speed of the photographing lens based on the luminance of the subject detected by the photometric sensor 207. The CPU 100 performs aperture value control via the lens communication circuit 205, and controls shutter speed control by controlling the energization time of the shutter magnets 218a and 218b via the shutter control circuit 208. Further, the CPU 100 controls the image sensor 206 to perform a shooting operation. The CPU 100 incorporates a subject recognition circuit 209 for performing subject recognition and tracking operations based on the image from the AF sensor 101. The CPU 100 has a built-in storage circuit 210 such as a ROM storing a program for controlling camera operations, a RAM for storing variables, and an EEPROM (electrically erasable and writable memory) for storing various parameters. Has been.

  Next, the optical configuration of the camera in the present embodiment will be described. FIG. 3 is an optical configuration diagram of the camera. Most of the light beam from the subject incident through the photographing lens 300 is reflected upward by the quick return mirror 305 and forms an image on the finder screen 303. The user of the camera observes this image through the pentaprism 301 and the eyepiece lens 302.

  A part of the imaging light beam passes through the quick return mirror 305 and is bent downward by the rear sub-mirror 306, passes through the field mask 307, the field lens 311, the diaphragm 308, and the secondary imaging lens 309, and then on the AF sensor 101. To form an image. The CPU 100 can detect the focus state of the photographic lens 300 by processing an image signal obtained by photoelectrically converting this image. At the time of shooting, the quick return mirror 305 jumps up, and all of the shooting light beam is imaged on the image sensor 206, and the subject image is exposed.

  As a focus detection method in this embodiment, a known phase difference detection method is used, and it is possible to detect the focus states of a plurality of different areas in the screen. FIG. 4 is a detailed configuration diagram of the optical system in focus detection. The light beam from the subject that has passed through the photographic lens 300 is reflected by the sub-mirror 306 and once forms an image in the vicinity of the field mask 307 in a plane conjugate with the imaging surface. FIG. 4 shows a developed optical path reflected by the sub-mirror 306 and turned back. The field mask 307 is a member used to shield extra light other than the focus detection area (ranging point) in the screen.

  The field lens 311 has a function of forming an image of each aperture of the diaphragm 308 near the exit pupil of the photographing lens 300. A secondary imaging lens 309 including a pair of two lenses is disposed behind the stop 308, and each lens corresponds to each opening of the stop 308. Each light beam that has passed through the field mask 307, the field lens 311, the stop 308, and the secondary imaging lens 309 forms an image on the area sensor on the AF sensor 101. The AF sensor 101 is configured to be able to form an image of light beams from different subjects in the shooting screen.

  Here, the field of view of the area sensor on the AF sensor 101 in the shooting screen will be described with reference to FIGS. 5 and 6. FIG. 5 is a layout diagram of area sensors in the AF sensor 101. The area sensors 102a and 102b are configured by two-dimensionally arranging a plurality of pixels, and accumulate an image signal obtained from a light beam that has passed through the photographing lens 300 in the plurality of pixels. The area sensors 102 a and 102 b are optically paired by the secondary imaging lens 309. FIG. 6 is a diagram showing the positional relationship between the finder field and the AF sensor field (subject recognition field, focus detection field). FIG. 6A shows a subject recognition visual field 502 in the area sensors 102 a and 102 b, and receives light from substantially the entire finder visual field 501. On the other hand, FIG. 6B is a focus detection visual field 503 in the area sensors 102a and 102b, and the size of the entire visual field is the same, but the area sensors 102a and 102b are 5 × 5 in blocks of a predetermined size. To form a field of view of 25 blocks. Thus, the focus detection visual field 503 is provided corresponding to the 25 distance measuring points. In the following description, FIG. 6A is referred to as a first pattern, and FIG. 6B is referred to as a second pattern.

  Next, a detailed circuit configuration of the AF sensor 101 will be described with reference to the block diagram of FIG. The area sensors 102a and 102b are composed of aperture pixels for receiving a subject image and light-shielded OB pixels (light-shielded pixels). The subject image formed by the secondary imaging lens 309 is photoelectrically converted by the area sensors 102a and 102b, and signals converted into voltages are accumulated in the area sensors 102a and 102b.

  The block selection circuit 103 (selection means) divides the blocks of the area sensors 102a and 102b with a predetermined pattern (division pattern). The division patterns are the two division patterns (first pattern and second pattern) described with reference to FIGS. 6 (a) and 6 (b). That is, the block selection circuit 103 selects one pattern from the first pattern that configures the area sensors 102a and 102b as a single block and the second pattern that configures as a plurality of blocks. The block selection circuit 103 uses the P-B signal detection circuit 104, the P-OB signal detection circuit 105, the first frame memories 108a and 108b, and the second frame memory 109a to be described later. , 109b.

  The P-B signal detection circuit 104 processes a pixel signal in each block of the second pattern (BLK1 to BLK25) to detect a P-B signal. Here, a method for detecting the P-B signal will be described in detail with reference to FIG. FIG. 9 is a diagram showing image signals obtained by processing the pixel signals of each block while paying attention to the pixel groups of the blocks (BLK1 to BLK3) of the area sensor 102a. In each block, a plurality of pixels are arranged two-dimensionally. The configuration of the area sensor 102b is the same as that of the area sensor 102a. In FIG. 9, if the vertical direction is defined as a column and the horizontal direction is defined as a row, the P-B signal detection circuit 104 calculates a projection signal obtained by adding pixel signals included in the same column. The P-B signal detection circuit 104 detects a difference signal (P-B signal) between the largest signal (peak signal) and the smallest signal (bottom signal) among the projection signals in each block.

  In FIG. 1, the P-OB signal detection circuit 105 has a first pattern, that is, a difference signal (P) between the largest signal (peak signal) and the OB pixel signal (light-shielded pixel signal) among all the pixel signals of the area sensors 102a and 102b. -OB signal) is detected. The accumulation stop determination circuit 106 determines the accumulation stop timing based on the P-B signal or the P-OB signal.

  The amplifying circuit 107 (amplifying means) is a circuit for amplifying the pixel signal when the pixel signal of the area sensor 102b is stored in the first frame memory 108b and the second frame memory 109b. The amplifier circuit 107 is configured to be switchable between two amplification factors (first amplification factor and second amplification factor) of × 1 and × 4. For example, the amplifier circuit 107 generates a subject recognition signal by amplifying the pixel signal obtained by selecting the first pattern with the first amplification factor, and the pixel signal obtained by selecting the second pattern is the first. A focus detection signal is generated by amplification at a second amplification factor different from the amplification factor. The first amplification factor and the second amplification factor are not limited to x1 and x4, and other amplification factors may be used.

  The first frame memories 108a and 108b (storage means) are circuits for storing a subject recognition signal. The signals for the entire screen (pixel signals) obtained by selecting the first pattern are collectively displayed as a subject recognition signal. Remember as. The first frame memories 108 a and 108 b also store the OB pixel signal detected by the P-OB signal detection circuit 105 at the same time as the pixel signal. The second frame memories 109a and 109b (storage means) are circuits for storing focus detection signals, and each block (BLK1 to BLK25) obtained by selecting a second pattern corresponding to 25 distance measuring points. (Pixel signal) is stored as a focus detection signal. The second frame memories 109a and 109b also store the bottom signal for each block detected by the P-B signal detection circuit 104 at the same time as the pixel signal.

  By driving the first shift register 110 under the control of the CPU 100, the subject recognition signals stored in the first frame memories 108a and 108b are selected in order and sent to the OB reference output circuit 111. The OB reference output circuit 111 clamps the pixel signals stored in the first frame memories 108a and 108b to the OB signal reference stored in the first frame memories 108a and 108b, performs signal amplification processing, and outputs the result. The pixel signal output from the OB reference output circuit 111 is AD converted by an AD converter (not shown) in the CPU 100.

  By driving the second shift register 112 by the CPU 100, the focus detection signals stored in the second frame memories 109a and 109b are sequentially selected and sent to the bottom reference output circuit 113. The bottom reference output circuit 113 clamps the pixel signals stored in the second frame memories 109a and 109b to the bottom signal reference stored in the second frame memories 109a and 109b, performs signal amplification processing, and outputs the result. The pixel signal output from the bottom reference output circuit 113 is AD converted by an AD converter (not shown) in the CPU 100.

  The focus detection operation by the focus detection apparatus (camera) configured as described above will be described in detail with reference to the flowchart of FIG. First, in step S101, a signal accumulation subroutine is executed by the AF sensor 101 that has received a control command from the CPU 100. That is, when the CPU 100 receives a focus detection start signal by operating the switch group 214, the CPU 100 controls the AF sensor 101 to perform a signal accumulation operation.

  FIG. 8 is a flowchart of a signal accumulation subroutine by the AF sensor 101. In the signal accumulation subroutine (step S101), first, in step S201, signal accumulation of the area sensors 102a and 102b is started by the control circuit in the AF sensor 101. In step S202, the division pattern by the block selection circuit 103 is set to the second pattern.

  In step S203, the storage stop determination of each block of the second pattern (25 divisions) selected in step S202 is performed. The block selection circuit 103 transmits the signal of each block (BLK1 to BLK25) to the P-B signal detection circuit 104, and the P-B signal detection circuit 104 detects the P-B signal. The P-B signal of each block is compared with the stop level by the accumulation stop determination circuit 106. If the P-B signal is equal to or higher than the stop level, the process proceeds to step S204. On the other hand, if the P-B signal has not reached the stop level, it is determined that the accumulated signal is insufficient, and the process proceeds to step S206. In step S204, the amplification factor (second amplification factor) of the amplifier circuit 107 is set to x1.

  In step S205, the pixel signal of the block determined to stop accumulation in step S203 is stored in the second frame memories 109a and 109b. At the same time, the bottom signal detected by the P-B signal detection circuit 104 is stored in the second frame memories 109a and 109b. In the second frame memories 109a and 109b, pixel signals having a sufficient P-B signal amount for each block divided into 25 for focus detection are stored. Therefore, the signal between blocks becomes a discontinuous signal.

  In step S206, the division pattern by the block selection circuit 103 is set to the first pattern. In step S207, accumulation stop determination is performed with the first pattern (no division) selected in step S206. The block selection circuit 103 sends the signals of all the pixels of the area sensors 102a and 102b to the P-OB signal detection circuit 105, and the P-OB signal detection circuit 105 detects the P-OB signal. The P-OB signal is compared with the stop level by the accumulation stop determination circuit 106. If the P-OB signal is equal to or higher than the stop level, the process proceeds to step S208. On the other hand, if the P-OB signal has not reached the stop level, it is determined that the accumulated signal is insufficient, and the process proceeds to step S210. In step S208, the amplification factor (first amplification factor) of the amplifier circuit 107 is set to x4.

  In step S209, all the pixel signals determined to stop accumulation in step S207 are stored in the first frame memories 108a and 108b. At the same time, the OB signal detected by the P-OB signal detection circuit 105 is stored in the first frame memories 108a and 108b. In the first frame memory 108a, all pixel signals of the area sensor 102a are simultaneously stored for subject recognition. For this reason, the pixel signal is a continuous signal. In the first frame memory 108b, all pixel signals of the area sensor 102b are subjected to x4 amplification processing for subject recognition and stored simultaneously. Since this pixel signal is amplified to x4, it is suitable for dark object recognition processing.

  In step S210, it is determined whether or not the storage operation to the first and second frame memories has been performed for all the blocks of the first pattern and the second pattern. If the storage operation of the pixel signal of any block is not completed, the process returns to step S202, and the accumulation and accumulation stop determination operations are repeated for the block for which the storage operation is not completed. On the other hand, when the storage operation of all the blocks is completed, the process returns to the main flow of FIG.

  In FIG. 7, in step S102, the CPU 100 reads out pixel signals stored in the first frame memories 108a and 108b as subject recognition signals. By driving the first shift register 110 in accordance with a command from the CPU 100, the pixel signals in the first frame memories 108a and 108b are selected in order for each pixel and sent to the OB reference output circuit 111. In addition, the pixel signal clamped to the OB signal reference stored in the first frame memories 108a and 108b and subjected to the signal amplification processing is AD converted by an AD converter (not shown) in the CPU 100. In this way, the subject recognition signal is generated based on the signal obtained from the light-shielded pixel.

  In step S103, subject recognition processing by the subject recognition circuit 209 in the CPU 100 is performed based on the pixel signal (subject recognition signal) obtained in step S102. The subject recognition circuit 209 detects the position of the subject by generating an image based on the pixel signal obtained in step S102 and comparing it with a reference image of the subject stored in advance. A distance measuring point is determined from the detected subject position. In step S104, the image signal corresponding to the position of the subject detected in step S103 is updated and stored as a new subject reference image.

  In step S105, the CPU 100 reads out pixel signals stored in the second frame memories 109a and 109b as focus detection signals. The second shift register 112 is driven in order to read out the pixel signal from the block of distance measuring points determined in step S103. At this time, pixel signals for each pixel in the second frame memories 109 a and 109 b are sequentially selected and sent to the bottom reference output circuit 113. Further, the pixel signal clamped to the bottom signal reference stored in the second frame memories 109a and 109b and subjected to signal amplification processing is AD-converted by an AD converter (not shown) in the CPU 100. As described above, the focus detection signal is generated based on the smallest signal in the plurality of blocks constituting the second pattern.

  Next, in step S106, the CPU 100 calculates a defocus amount by performing a correlation calculation based on the pixel signal subjected to AD conversion in step S105. In this way, the CPU 100 is a calculating unit that recognizes the position of the subject based on the subject recognition signal and calculates the defocus amount based on the focus detection signal in one block of the second pattern corresponding to the subject position. is there. In step S107, the CPU 100 performs drive control of the focus lens of the photographing lens 300 via the lens communication circuit 205 based on the defocus amount calculated in step S106.

  In step S108, it is determined whether or not to continue the focus detection operation. When the CPU 100 receives a focus detection continuation signal by operating the switch group 214, the CPU 100 returns to step S101 and repeats the subject recognition operation and the focus detection operation. On the other hand, when the CPU 100 receives a focus detection end signal by operating the switch group 214, the CPU 100 ends a series of focus detection operations.

  Next, the flow of the focus detection operation in FIG. 7 will be described using an example of an actual pixel signal with reference to FIGS. 10 and 11. FIGS. 10A to 10C are images generated from pixel signals read from the first frame memory 108a by performing a series of operations in steps S101 to S108. The pixel signal obtained from the first frame memory 108a is not divided into blocks and is a signal based on the OB pixel signal. For this reason, the images shown in FIGS. 10A to 10C are images having a high correlation in luminance distribution with each other.

  First, a pixel signal is read from the first frame memory 108a in the first signal accumulation operation (steps S101 and S102). The CPU 100 determines the position of the subject by generating the image of FIG. 10A from the read pixel signal and comparing it with a reference image stored in advance. In the present embodiment, it is detected that a subject exists near BLK9, and an image near BLK9 is stored as a new reference image (steps S103 and S104).

  Next, the BLK9 signal is read from the second frame memory 109a, and a focusing operation is performed based on the read signal (steps S105 to S107). FIG. 11 shows pixel signals from BLK9 in the second frame memory 109a. The pixel signals read from the second frame memory 109a are subjected to accumulation control so that the P-B signal in the block becomes equal to or higher than the stop level. In correlation calculation at the time of focus detection, a DC component is an unnecessary signal. On the other hand, the P-B signal, which is a contrast component, affects the accuracy of correlation calculation. For this reason, in this embodiment, the focus detection signal is clamped with the bottom reference and amplified to cut the DC component, and the P-B signal is amplified.

  Next, the pixel signal is read from the first frame memory 108a in the second accumulation operation (steps S101 to S102). The image shown in FIG. 10B is generated from the read signal, and compared with a reference image stored in advance, it is detected that a subject exists in the vicinity of BLK12. Then, the image near BLK12 is stored as a new reference image (steps S103 to S104). Further, the BLK12 signal is read from the second frame memory 109a, and the focus operation is performed based on the read signal (steps S105 to S107).

  Further, the pixel signal is read from the first frame memory 108a in the third accumulation operation (steps S101 to S102). The image shown in FIG. 10C is generated from the read signal, and compared with a reference image stored in advance, it is detected that a subject exists in the vicinity of BLK19. Then, the image near BLK19 is stored as a new reference image (steps S103 to S104). Further, the BLK19 signal is read from the second frame memory 109a, and the focus operation is performed based on the read signal (steps S105 to S107).

  In this embodiment, an image is generated from the signal stored in the first frame memory 108a. If it is determined that the reference image is dark, the first frame memory in which the amplified signal is stored. An image may be generated by the signal from 108b. Further, an image may be generated by combining signals stored in the first frame memories 108a and 108b.

  As described above, by configuring the AF sensor 101 so that the focus detection pixel signal and the subject recognition signal can be accumulated, the moving subject position is tracked and detected without providing a subject recognition sensor separately. Focus adjustment corresponding to the subject position can be performed. Further, the focus detection signal can be detected with high accuracy by dividing the area sensors 102a and 102b into blocks, controlling the accumulation for each block, and further amplifying the P-B signal based on the bottom signal. . Furthermore, subject recognition signals are collectively controlled by the entire area sensors 102a and 102b and output on the basis of the OB pixel signal, so that images obtained continuously have high correlation in luminance distribution and high accuracy. The subject can be recognized. Further, by providing a plurality of frame memories, both the focus detection signal and the subject recognition signal can be obtained by a single signal accumulation operation, and the time for focus detection can be shortened.

  In this embodiment, the still camera has been described. However, the focus detection device and the photometry device of the present technology are not only a still camera, but also a camcorder (movie camera) that performs moving image shooting, various inspection cameras, a surveillance camera, an endoscope camera, It can also be applied to robot cameras. Moreover, you may apply to the apparatus provided with the focus detection function in the imaging sensor. This embodiment provides a camera focus detection device, and is particularly useful for a still camera and a movie camera.

  According to the present embodiment, by using the subject recognition function and the focus detection function together with one AF sensor, a focus detection device having a subject tracking function can be provided at low cost. Furthermore, by providing the AF sensor with a subject recognition signal accumulation control mode and output mode and a focus detection signal accumulation control mode and output mode, both subject recognition accuracy and focus detection accuracy can be improved.

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

100: CPU
101: AF sensor 102a, 102b: Area sensor 103: Block selection circuit 108a, 108b: First frame memory 109a, 109b: Second frame memory

Claims (5)

An area sensor that has a light-shielding pixel and accumulates an image signal obtained from a light beam that has passed through a photographing lens in a plurality of pixels;
Selection means for selecting one pattern from a first pattern that constitutes the area sensor as a first block and a second pattern that constitutes a second block that is larger than the first block ;
First detection means for detecting a first signal using the largest signal among the pixel signals of the first pattern selected by the selection means and the light-shielded pixel signal obtained from the light-shielded pixels;
Second detection means for detecting a second signal using the largest signal and the smallest signal among the pixel signals in each block of the second pattern selected by the selection means;
Accumulation control means for controlling the accumulation time of the first block based on the first signal, and controlling the accumulation time of the second block based on the second signal;
Storing said pixel signal of said first pattern obtained based on the accumulated time stored as object recognition signal, and stores the pixel signal of the second pattern obtained on the basis of the accumulation time as the focus detection signal Means,
Calculating means for recognizing the position of the subject based on the subject recognition signal and calculating a defocus amount based on the focus detection signal in one block of the second pattern corresponding to the position of the subject;
A focus detection apparatus comprising: Before Symbol object recognition signal, focus detecting apparatus according to claim 1, characterized in that it generated the light shielding pixel No. signal as a reference. The focus detection apparatus according to claim 1, wherein the focus detection signal is generated on the basis of the smallest signal in the second block constituting the second pattern.   The pixel signal obtained by selecting the first pattern is amplified by a first amplification factor to generate the subject recognition signal, and the pixel signal obtained by selecting the second pattern is first amplified. 4. The focus detection apparatus according to claim 1, further comprising amplification means for generating the focus detection signal by amplifying at a second amplification factor different from the rate. 5. Accumulating an image signal obtained from a light beam that has passed through a photographing lens in a plurality of pixels of an area sensor having a light-shielding pixel ;
Selecting one pattern from a first pattern that constitutes the area sensor as a first block and a second pattern that constitutes a second block that is larger than the first block ;
Detecting the first signal using the largest signal among the selected pixel signals of the first pattern and the light-shielded pixel signal obtained from the light-shielded pixel;
Detecting a second signal using a largest signal and a smallest signal among pixel signals in each block of the selected second pattern;
Controlling the accumulation time of the first block based on the first signal, and controlling the accumulation time of the second block based on the second signal;
The pixel signal of the first pattern obtained on the basis of the accumulation time is stored as an object recognition signal, and stores the pixel signal of the second pattern obtained on the basis of the accumulation time as the focus detection signal Steps,
Recognizing the position of the subject based on the subject recognition signal, and calculating a defocus amount based on the focus detection signal in one block of the second pattern corresponding to the subject position. A focus detection method characterized by the above.