JP2014102290A - Automatic focusing apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable a focusing operation in which the difference between individuals and difference between photographing environments, of a predetermined shift amount for correcting a focusing position, required for the spacial frequency of an object are absorbed.SOLUTION: An automatic focusing apparatus includes imaging means for photoelectrically converting an object image formed by a photographic optical system including a focus lens and generating and outputting an imaging signal, and control means for performing a scan operation for obtaining an evaluation signal indicating the focusing state of the focus lens, while moving the focus lens, on the basis of the imaging signal and controlling the position of the focus lens, on the basis of the obtained evaluation signal. The automatic focusing apparatus includes a first mode for reading the imaging signal, while adding the imaging signal, and a second mode for reading the imaging signal, without adding the imaging signal, as a mode for reading the imaging signal from the imaging means. The control means limits the reading of the imaging signal through the use of the second mode, in the case of a second illuminance lower than a first illuminance.

Description

  The present invention relates to an automatic focus adjustment apparatus that performs focus adjustment.

  Conventionally, various means have been proposed for reducing the influence of optical aberrations on automatic focus adjustment. For example, in Patent Document 1, color information of a distance measurement area is detected, and the focus position corresponding to the distance measurement area is shifted by a predetermined amount determined by the detected color information and / or characteristic information of the focus lens. It is disclosed that the focus lens is driven to the position.

  Further, Patent Document 2 aims to enable focus detection for a subject having a low frequency pattern and a high frequency subject having a fine pattern. To achieve this object, a sample pitch for sampling the subject image formed on the light receiving surface is selected based on the result of comparing the spatial frequency of the subject image signal with a predetermined value, and the selected subject image signal It is disclosed to perform focus detection based on the above.

  Patent Document 3 aims to perform focus adjustment with high accuracy regardless of the spatial frequency component included in the subject. For this purpose, a detailed search is performed after the coarse search, and when the maximum focus evaluation value obtained by the coarse search is equal to or greater than a predetermined value, the high-pass filter F1 used in the coarse search is used in the detailed search. Also, it is disclosed that high frequency components are extracted using a high-pass filter F2 having a high cut frequency.

JP 2004-347665 A Japanese Patent Laid-Open No. 11-14900 JP 2004-325517 A

  However, in Patent Document 1, it is not possible to absorb the difference between individuals of a predetermined shift amount determined by the detection information of the color information detection means and / or the characteristic information of the focus lens and the difference due to the photographing environment. In order to absorb this difference, it is effective to perform focus adjustment on a subject having a spatial frequency as high as possible. However, when trying to find the in-focus position at high speed, it may be necessary to perform horizontal pixel addition reading of the image sensor, and in that case, the focus lens position characteristics of the high spatial frequency of the subject will be affected. can not get. For this reason, it is impossible to absorb a difference between individuals of a predetermined shift amount and a difference due to an imaging environment and perform a focus adjustment operation at high speed.

  Patent Document 2 is an invention relating to phase difference AF, and can select a plurality of sample pitches for sampling a subject image formed on a light receiving surface. In the initial state, the spatial frequency of a subject image signal is set with a fine sampling pitch. To get. For this reason, there is no disclosure about a predetermined shift amount determined by speeding up of contrast AF or focus lens characteristic information.

  In Patent Document 3, the sensor reading mode is the same for the coarse search and the detailed search. Here, if the horizontal pixel non-addition readout mode is used in order to absorb the difference between the predetermined shift amounts depending on the individual and the photographing environment, a high-speed focus adjustment operation cannot be performed. On the other hand, if the horizontal pixel addition readout mode is used for speeding up, it is not possible to absorb differences between individual shift amounts and differences due to shooting environments.

  Therefore, an object of the present invention is to make it possible to perform a focus adjustment operation that absorbs differences between individual shift amounts for correcting a focus position required by the spatial frequency of a subject and differences due to an imaging environment.

  In view of the above object, the first aspect of the present invention is based on an imaging unit that photoelectrically converts a subject image formed by a photographing optical system including a focus lens to generate and output an imaging signal, An automatic focus adjustment apparatus having a control unit that performs a scanning operation for acquiring an evaluation signal indicating an in-focus state of the focus lens while moving the focus lens, and that controls a position of the focus lens based on the acquired evaluation signal. The control means has a first mode for reading out the image pickup signal from the image pickup means and a second mode for reading out the image pickup signal without adding the image pickup signal. In the case of the second illuminance lower than the first illuminance, the reading of the imaging signal using the second mode is limited.

  According to the present invention, it is possible to enable a focus adjustment operation that absorbs differences between individual shift amounts for correcting a focus position required by the spatial frequency of an object and differences due to an imaging environment.

Block diagram for explaining imaging devices 1 to 3 of the embodiment Explanatory drawing of operation procedure of Examples 1-3 The figure explaining the reliability evaluation method of AF evaluation value Operation explanatory diagram of scan AF processing of Embodiment 1 The figure which shows an example of the relationship between a subject's spatial frequency and a focus position Explanatory drawing of scan AF operation of Example 1 Program diagram for determining exposure in Example 1 Program diagram for determining exposure in Example 1 Explanatory drawing of operation procedure of embodiment 2 Operation explanatory diagram of scan AF processing of embodiment 2

Example 1
FIG. 1 shows a block diagram of an imaging apparatus 1 embodying the present invention. The photographic lens barrel 31 includes a zoom lens group 2, a focus lens group 3, and a photographic optical system that includes a diaphragm that is an amount of light adjusting means that controls the amount of light transmitted through these lens groups and is an exposure means.

  First, the light flux of the subject that has passed through the taking lens barrel 31 of the image pickup apparatus 1 is imaged on the light receiving surface of the sensor 5 after the amount of light is adjusted by the diaphragm 4. This subject image is converted into an electrical signal (imaging signal) by photoelectric conversion processing of a solid-state imaging device (sensor) 5 such as a CCD or CMOS, and is output to the imaging circuit 6. In the imaging circuit 6, various signal processing is performed on the input signal, and a predetermined image signal is generated. This image signal is output to the A / D conversion circuit 7, converted into a digital signal (image data), and then temporarily stored in a memory (VRAM) 8 such as a buffer memory. The image data stored in the VRAM 8 is output to the D / A conversion circuit 9, converted into an analog signal and converted into an image signal in a form suitable for display, and then the image data is displayed on the image display device (LCD) 10. Is displayed.

  On the other hand, the image data stored in the VRAM 8 is also output to the compression / decompression circuit 11. The compression / expansion circuit 11 includes a compression circuit and an expansion circuit. The compression circuit reads out the image signal temporarily stored in the VRAM 8 and performs compression processing and encoding processing of the image data in order to make it suitable for storage in the storage memory 12. The decompression circuit performs a decoding process, a decompression process, and the like for making the image data stored in the storage memory 12 an optimum form for reproducing and displaying. The image data output from the VRAM 8 is subjected to compression processing by the compression circuit in the compression / decompression circuit 11, converted to image data in a form suitable for storage, and stored in the storage memory 12 including a semiconductor memory or the like. The

  The storage memory 12 that is a storage medium for image data or the like is a fixed semiconductor memory such as a flash memory, or a card-type flash memory that has a card shape or a stick shape and is detachable from the device. A semiconductor memory is applied. In addition, various forms such as a magnetic storage medium such as a hard disk or a floppy disk are applied.

  As the operation switch 24, there are a main power switch for starting the imaging apparatus 1 and supplying power, and a release switch for starting a photographing operation (storage operation). In addition, there are a playback switch for starting a playback operation, a zoom switch for moving the zoom lens group 2 of the photographing optical system to perform zooming, an optical viewfinder (OVF) electronic viewfinder (EVF) changeover switch, and the like. The release switch has a first stroke (hereinafter referred to as SW1) for generating an instruction signal for starting an AE process and an AF process performed before the photographing operation and a second stroke (hereinafter referred to as SW2) for generating an instruction signal for starting an actual exposure operation. And a two-stage switch.

  For example, when a reproduction switch (not shown) of the operation switches 24 is operated and turned on, the reproduction operation is started. Then, the image data stored in a compressed form in the storage memory 12 is output to the compression / expansion circuit 11, subjected to decoding processing, expansion processing, etc. in the expansion circuit, and then output to the VRAM 8 and temporarily stored. The Further, the image data is output to the D / A conversion circuit 9, converted into an analog signal, converted into an image signal in a form suitable for display, and then displayed on the LCD 10 as an image.

  On the other hand, the image data digitized by the A / D conversion circuit 7 includes an AE processing circuit 13 that performs automatic exposure (AE) processing and an AF processing circuit 14 that performs automatic focus adjustment (AF) processing, separately from the VRAM 8 described above. And the face detection circuit 36. First, the AE processing circuit 13 receives the input digital image signal and performs arithmetic processing such as cumulative addition on the luminance value of the image data for one screen. Thereby, the AE evaluation value corresponding to the brightness of the subject is calculated. This AE evaluation value is output to the CPU 15.

  The AF processing circuit 14 receives the output from the A / D conversion circuit 7 and generates an AF evaluation value indicating the in-focus state. The AF processing circuit 14 receives the input digital image signal, extracts a high-frequency component of the image data through a high-pass filter (HPF) or the like, and performs arithmetic processing such as cumulative addition, so that the high frequency side is processed. An AF evaluation value signal corresponding to the contour component amount and the like is calculated. Specifically, high-frequency components of image data corresponding to a partial area of the screen designated as the AF area are extracted via a high-pass filter (HPF) or the like, and arithmetic processing such as cumulative addition is performed. Thereby, an AF evaluation value signal corresponding to the contour component amount on the high frequency side and the like is calculated. When this AF area is one place in the central part or any part on the screen, or in the central part or any part on the screen and a plurality of adjacent parts, or in a plurality of places distributed discretely and so on. As described above, the AF processing circuit 14 serves as a high-frequency component detection unit that detects a predetermined high-frequency component from the image signal generated by the sensor 5 in the process of performing the AF processing.

  The face detection circuit 36 receives the digital image signal input from the A / D conversion circuit 7 and searches the image for portions characterizing the face such as eyes and eyebrows, and determines the position of the person's face on the image. Ask. Further, the size and inclination of the face are obtained from the positional relationship such as the interval between the parts characterizing the face.

  On the other hand, a predetermined timing signal is output from the timing generator (TG) 16 to the CPU 15, the image pickup circuit 6, and the sensor driver 17, and the CPU 15 performs various controls in synchronization with this timing signal. The imaging circuit 6 receives the timing signal from the TG 16 and performs various image processing such as separation of color signals in synchronization with the timing signal. Further, the sensor driver 17 receives the timing signal of the TG 16 and drives the sensor 5 in synchronization therewith.

  The CPU 15 controls the first motor driving circuit 18, the second motor driving circuit 19, and the third motor driving circuit 20, respectively, so that the diaphragm 4, the focus driving motor 22, and the zoom driving motor 23 are controlled. The focus lens group 3 and the zoom lens group 2 are driven and controlled. That is, the CPU 15 controls the first motor drive circuit 18 based on the AE evaluation value calculated by the AE processing circuit 13 to drive the aperture drive motor 21 and adjust the aperture amount of the aperture 4 so as to be appropriate. AE control is performed. Further, the CPU 15 controls the second motor drive circuit 19 based on the AF evaluation value signal calculated by the AF processing circuit 14 to drive the focus drive motor 22 to move the focus lens group 3 to the in-focus position. I do. When a zoom switch (not shown) among the operation switches 24 is operated, the CPU 15 receives this and controls the third motor drive circuit 20 to drive the zoom drive motor 23, thereby controlling the zoom lens group 2. Is moved to perform a zooming operation of the photographing optical system.

  The EEPROM 25 is an electrically rewritable read-only memory in which programs for performing various controls and the like and data used for performing various operations are stored in advance. 26 is a battery. The switching circuit 27 controls the flash emission of the strobe light emitting unit 28. The speaker 30 performs voice guidance and warning. The LED 29 performs an in-focus display or a warning display based on an instruction from the CPU 15. The AF auxiliary light emitting unit 33 is composed of a light source such as an LED, which is an illuminating unit that illuminates all or part of the subject when acquiring an AF evaluation value. The AF auxiliary light driving circuit 32 drives the AF auxiliary light emitting unit 33 based on an instruction from the CPU 15. The shake detection sensor 35 detects camera shake and the like, and the shake detection circuit 34 processes the signal of the shake detection sensor 35.

  Next, the actual photographing operation of the image pickup apparatus of the present embodiment will be described with reference to the flowchart shown in FIG. In this embodiment, the operation of acquiring the AF evaluation value while driving the focus lens group 3 to a predetermined position is called a scan operation. Also, the focus lens position interval from which the AF evaluation value is acquired is the scan interval, the position at which the AF evaluation value is acquired is the scan point, the number of AF evaluation value acquisition is the scan point number, and the AF evaluation value acquisition range is scanned. An area for acquiring an image signal for detecting the range and the in-focus position is called an AF frame.

  When the main power switch of the imaging apparatus 1 is in the ON state and the operation mode of the imaging apparatus is in the shooting (recording) mode, the shooting process sequence is executed. First, in step S201, the CPU 15 displays an image formed on the sensor 5 through the photographing lens barrel 31 as an image on the LCD.

  Next, in step S202, the state of the release switch is confirmed. When the photographer operates the release switch and the CPU 15 confirms that SW1 (first stroke of the release switch) is turned on, the process proceeds to the next step S203, and normal AE processing is executed.

  Subsequently, in step S204, the AF processing circuit 14 acquires a plurality of AF evaluation values and performs a scan AF process for detecting a focus position from the results. Details of this processing will be described later.

  If the reliability of any AF evaluation value obtained in step S204 is sufficient, AFOK display is performed in step S205. The AFOK display is performed by turning on the LED 29 and the like, and processing such as displaying a green frame on the LCD 10 is performed. If the reliability of all AF evaluation value signals is low in step S204, the focus position is not detected and the process proceeds to step S205 where AFNG display is performed. This is performed by, for example, blinking the LED 29, and simultaneously performing processing such as displaying a yellow frame on the LCD.

  In step S206, the CPU 15 confirms SW2 (second stroke of the release switch). If SW2 is on, the CPU 15 proceeds to step S207 and executes actual exposure processing.

  Here, a method for evaluating the reliability of the AF evaluation value signal will be described. Since a specific method is described in Japanese Patent Registration No. 4235422 and Japanese Patent Registration No. 4185541, the outline thereof will be described with reference to FIG.

  The AF evaluation signal has a mountain shape as shown in FIG. 3 when the horizontal axis indicates the distance and the vertical axis indicates the AF evaluation value, except for a special case of near-far competition. Therefore, whether or not the AF evaluation signal has a mountain shape is determined from the difference between the maximum value and the minimum value of the AF evaluation signal, the length of the inclined portion with an inclination greater than a certain value, and the gradient of the inclined portion. By doing so, the reliability of the AF evaluation signal is determined. As shown in FIG. 3, points D and E that are recognized as being inclined from the top of the mountain (point A) are obtained. This is performed by examining the difference between the AF evaluation values acquired at points adjacent to the acquisition point. This check starts from the top of the mountain (point A). The difference in the AF evaluation values between the top of the mountain (point A) and the adjacent point on the D point side (infinity side) is examined, and if the top of the mountain (point A) is larger, it is determined that the slope is inclined. Next, the difference between the AF evaluation value of the point compared to the peak of the mountain (point A) and its adjacent point is examined, the AF evaluation value of the point closer to the peak of the mountain (point A) is large, and the difference is a predetermined value ( If it is equal to or greater than (SlopeThr), it is determined that the vehicle is inclined. This check is continued until the magnitude relationship between the AF evaluation values of adjacent points is reversed or the difference is smaller than a predetermined value. In FIG. 3, since the difference between the AF evaluation values of the point D and the point on the infinity side adjacent to the point D is smaller than a predetermined value, the point up to the point D is determined to be a portion inclined at a certain value or more. . Similarly, the point E side (closest side) is examined. In FIG. 3, it is determined that the portion up to point E is inclined at a certain inclination or more. The width between points D and E is defined as the width L of the mountain, the difference SL1 between the AF evaluation values at points A and D and the difference SL2 between the AF evaluation values at points A and E are obtained, and the sum SL1 + SL2 is determined as the slope of the mountain SL.

  The coefficients for determining the reliability of the AF evaluation value obtained in this way are compared with respective threshold values, and if all the conditions are satisfied, it is determined that the AF evaluation value is reliable. First, the difference between the maximum value and the minimum value of the AF evaluation value is compared with the predetermined value, and if it is smaller than the predetermined value, it is determined that the reliability is low. The maximum value and the minimum value in this case are the maximum value and the minimum value of the entire AF evaluation value. In the example of FIG. 3, the maximum value is the AF evaluation value acquired at point A, but the minimum value is the value at point E. It is not a value but a value obtained at a point closer to the closest side. The length L of the portion that is inclined at an inclination of a certain value or more is compared with the predetermined value Lo, and if it is smaller than the predetermined value, it is determined that the reliability is low. Then, the average value SL / L of the inclined portion is compared with the predetermined value SLo / Lo, and if it is smaller than the predetermined value, it is determined that the reliability is low. When the above three conditions are satisfied, it is determined that the reliability of the AF evaluation value is high.

  Here, the concept of the scan AF process for detecting the in-focus position in step S204 in FIG. 2 will be described with reference to FIG. The scan AF is performed by obtaining the position of the focus lens group 3 where the high frequency component of the imaging signal output from the sensor 5 is the largest in each BPF. Actually, data in a plurality of BPFs are acquired and processed by moving the focus lens group 3 once, but in FIG. 4, only the output waveform related to one BPF is shown to make the drawing easier to see.

  The CPU 15 controls the focus drive motor 22 via the second motor drive circuit 19 so that the focus lens group 3 is moved from a position corresponding to infinity (“A” in FIG. 4) to a close distance set in each shooting mode. Drive to the corresponding position ("B" in FIG. 4). Then, while driving the focus lens group 3, the output (AF evaluation value signal) of the AF processing circuit 14 is acquired for each of a plurality of set BPFs, and the AF evaluation value acquired when the driving of the focus lens group 3 is completed. From the signal, a position (“C” in FIG. 4) at which it becomes maximum is obtained for each of the plurality of BPFs.

  The acquisition of the output of the AF processing circuit 14 is performed at predetermined steps without performing stop positions of all the focus lens groups 3 in order to increase the speed of scan AF. In this case, the AF evaluation value signal may be acquired at points a1, a2, and a3 shown in FIG. In such a case, the in-focus position C is obtained by calculation from the point where the AF evaluation value signal becomes the maximum value and the points before and after that point.

  Next, details of the scan AF process in step S204 of FIG. 2 will be described. First, the output (AF evaluation value signal) of the AF processing circuit 14 will be described. In the general description, “high-frequency components are extracted via a high-pass filter (HPF) or the like, and further arithmetic processing such as cumulative addition is performed, and an AF evaluation value signal corresponding to the contour component amount on the high frequency side is calculated. However, a plurality of filters having different transmission characteristics are prepared. In addition, a bandpass filter (BPF) is configured to remove noise components near the Nyquist frequency and prevent aliasing due to high-frequency signals, and an AF evaluation value signal corresponding to the contour component amount on the high-frequency side is calculated. Yes.

  Next, the BPF setting method and how to obtain the in-focus position will be described. In general, when aberration such as spherical aberration of an optical system occurs, the focus position differs according to the frequency of the subject. Therefore, when an AF evaluation value signal is acquired while driving the focus lens group 3 using a filter that extracts high-frequency spatial frequencies of a plurality of subjects having different characteristics, for example, as shown in FIG.

  FIG. 5 shows a focus in a BPF having such characteristics that the transmittance of the filter is maximum at 6.7%, 13.3%, 16.7%, 26.7%, 50%, and 80% of the Nyquist frequency. The relationship between the position and the best focus position (focus position) is shown. When the frequency characteristics of the filters are different, the position where the AF evaluation value is maximized is different. In FIG. 5, the position where the AF evaluation value is maximized is distributed on the right side (that is, the side where the focus is close to the focus) as the frequency at which the transmittance of the BPF becomes maximum increases. Since this distribution is due to the characteristics of the photographic lens barrel 31 including the focus lens group 3, it is not always as shown in FIG. In some cases, the characteristics are reversed, and in other cases, the positions where the difference is small and the AF evaluation value is maximized are distributed at almost the same position. In addition, the amount may vary depending on the focal length, shooting distance, etc. associated with zooming of the taking lens barrel.

  When performing the automatic focus adjustment operation, the one that maximizes the transmittance of the filter at the highest possible frequency is used, and the focus lens group 3 is driven to a position where the AF evaluation value is maximized. This makes it possible to obtain the position of the focus lens group 3 that the photographer feels as the best focus position. This is because when shooting a subject with a mixture of high and low spatial frequencies, the low spatial frequency subjects do not feel blurred even if they are slightly out of focus, but the spatial frequency is high. This is because in the case of a subject, it is easy to feel that the image is blurred even with a slight shift.

However, even if an attempt is made to obtain a position where the AF evaluation value at a high frequency is maximized using a BPF having the maximum transmittance of the filter, sufficient AF evaluation value signal reliability cannot be obtained for the following reason. In some cases, the position where the AF evaluation value is maximized cannot be obtained.
(1) In order to read out signals from the sensor at high speed, pixel addition is performed at the time of reading, and the number of read pixels is reduced from the actual number of pixels, so that a high frequency component of a subject existing when no addition is performed is lost.
(2) Since the high-frequency signal includes noise superimposed on the sensor or the circuit system, a false signal may be generated. This tendency is particularly noticeable at low illumination.
(3) A high frequency component of the subject may be lost due to camera shake of the photographer or movement of the subject.
(4) A low spatial frequency component always exists if there is an end of the subject, but a high frequency component itself may not exist if a fine pattern does not exist in the subject.

  Although (2) to (4) may not occur depending on the situation at the time of shooting, (1) is a necessary requirement for reducing the time required for AF, and therefore it always occurs. Most recent image sensors read and add 3 pixels in the horizontal and vertical directions respectively, so the high-frequency component obtained during AF can only be obtained at 33.3% or less of the Nyquist frequency during shooting. . On the other hand, since the focus position of 50% to 80% of the Nyquist frequency at the time of shooting is required as the best focus position (focus position), it is not possible to focus as it is.

  Therefore, as in Patent Document 1 of the background art, a filter having the maximum transmittance at a relatively low frequency is used, and a predetermined amount determined by the characteristic information of the focus lens from the position where the AF evaluation value is maximized. A method of driving the focus lens group 3 at a shifted position has been adopted. However, in this method, there are error components such as individual differences, subject dependency, and light source dependency in a predetermined amount determined by the characteristic information of the focus lens. Therefore, when this error component is large, sufficient focusing accuracy cannot be obtained.

  Therefore, in the present invention, first, scanning is performed at a relatively rough interval using a readout mode capable of high-speed readout that performs horizontal pixel addition, and AF evaluation values are acquired with a plurality of BPFs from a low range to a high range. From each AF evaluation value, the focus lens position where the value becomes maximum is obtained and each reliability is evaluated. From each AF evaluation value and its reliability, it is judged whether AF in horizontal non-addition readout mode is possible, and if it is judged to be possible, detailed scanning is performed at relatively fine intervals in horizontal non-addition readout mode, Acquisition of AF evaluation values in a plurality of BPFs from low to high. Then, the focus lens position where the value is maximized is obtained from each AF evaluation value, and at the same time, each reliability is evaluated. If the reliability in the highest range is high, the focus lens group is obtained from the AF evaluation value obtained as a result. 3 peak position is selected as the in-focus position. When the reliability of the highest range is low as a result of the detailed scan in the horizontal non-addition readout mode, the position obtained by adding the BP correction amount (best focus correction amount) to the BPF result obtained with the highly reliable AF evaluation value signal Is the in-focus position.

  Also, if it is determined that AF in the horizontal non-addition readout mode is impossible from the result of scanning at a relatively coarse interval, a detailed scan is performed at relatively fine intervals in the horizontal addition readout mode. AF evaluation values in a plurality of BPFs are acquired up to the area. Of the BPFs obtained as a result of the AF evaluation value signal having high reliability, the peak position of the focus lens group 3 is obtained by the AF evaluation value obtained by the BPF set at the highest frequency, and the BP correction amount is obtained. The position that takes into account the focus position.

  Here, the BP correction amount is a correction amount for correcting a difference in image formation position due to a difference in spatial frequency and focusing on a subject having a high spatial frequency. Depending on the characteristic information of the focus lens, the imaging position of a high spatial frequency subject and a low frequency subject may differ. In this case, by focusing on a subject having a high spatial frequency, it is possible to focus on the entire subject having various spatial frequencies. This is because an object with a high spatial frequency feels blurred even with a slight focus position shift, but an object with a low spatial frequency does not feel blurred even with a slight shift. Therefore, when the AF evaluation value is acquired in a low frequency band, it is necessary to finally correct the focus lens control position. Specifically, the difference between the imaging position in the frequency band used for acquiring the AF evaluation value and the imaging position in the high frequency band is used as a BP correction amount, and the BP is set at the peak position of the acquired AF evaluation value. The position where the correction amount is added is set as the in-focus position, and the focus lens is controlled.

  Since this BP correction amount is determined by the characteristics of the photographic lens barrel, it is generally different when the frequency at which the BPF transmittance is maximized is different, and also varies depending on the focal length, photographic distance, etc. of the photographic lens barrel. Sometimes. Therefore, for each BPF used, a correction amount corresponding to the focal length of the photographic lens barrel and the photographic distance is recorded as data in a memory (not shown). For example, the BPF setting is 10%, 20%, 40%, 50% with respect to the input signal, and 3.3%, 6.7%, 13.3%, 16.7 in the case of horizontal three-pixel addition. The BPF is selected so that the transmittance peak is in%.

  Next, a specific operation procedure will be described with reference to FIG. When the scan AF process is started, first, in step S601, it is confirmed whether or not the illuminance at the time of shooting is brighter than a predetermined illuminance. When it is darker than the predetermined illuminance (second illuminance), the process proceeds to step S620, and when bright (first illuminance), the process proceeds to step S602. The predetermined illuminance may be set to a brightness at which a signal sufficient to detect the peak of the AF evaluation value in non-addition readout is obtained. For this determination, the output of the AE processing circuit 13 is used.

  As described above, at low illuminance, high-frequency signals include noise superimposed on sensors and circuit systems, and may generate false signals, so that sufficient illuminance can be obtained without generating false signals. If there is no error, there is a high possibility that the peak of the AF evaluation value cannot be detected in non-addition readout. Therefore, in the case of low illuminance, the process proceeds to step S620 in order to perform processing that does not use (limit) non-addition readout.

  In step S620, the sensor reading mode is changed. The readout mode considering power consumption for LCD display is changed to a readout mode for performing scan AF at low illumination. In this readout mode, the number of readout lines is increased by performing pixel addition so that the SN ratio of the output signal becomes high even at low illuminance. For example, from a readout mode of about 30 fps for LCD display, horizontal 3 pixels, vertical 2 pixels addition, and readout line number 347 lines, scan AF 30 fps, horizontal 3 pixels, vertical 3 pixels addition, readout line number 1018 lines at low illuminance Change to read mode. Accordingly, the drive frequency of the sensor is also changed if necessary.

  On the other hand, in step S602, the reading mode considering the power consumption for LCD display is changed to the high-speed reading mode. For example, the readout mode is changed from about 30 fps for LCD display to 120 fps for high-speed scan AF. Accordingly, the drive frequency of the sensor is also changed if necessary.

  Next, in step S603, the contrast of the subject is determined. If the contrast is high, the process proceeds to step S604, and if the contrast is low, the process proceeds to step S621. When the contrast is low, the amount of signal for generating the AF evaluation value is reduced. Therefore, as in the case of low illuminance, the high-frequency signal has a relatively increased noise superimposed on the sensor or circuit system, which is a false signal. May generate a signal. In this case, there is a high possibility that the peak of the AF evaluation value cannot be detected in non-addition readout. Therefore, in order to perform processing that does not use non-addition reading, the process proceeds to step S621. For the determination of the contrast, among the evaluation values obtained by the AF processing circuit 14, the difference between the maximum value and the minimum value of the signal value before performing the high pass filter (HPF) processing is obtained for each line in the AF frame. The value obtained by extracting the largest one is used. The contrast level is determined by whether or not the difference between the maximum value and the minimum value of the signal value is higher than a predetermined value.

  In step S604, a first scan (first scan operation) is performed. The first scan is performed in a horizontal addition readout mode (first mode). In this scan, the scan interval is an interval for obtaining an AF evaluation value sufficient to determine a scan range when performing a scan by non-addition readout at high speed. For example, the distance can be achieved at the highest speed of the focus lens group 3 drive mechanism, which is 5 times (hereinafter referred to as 5 depths) or more and 15 times (hereinafter referred to as 15 depths) or less of the focal depth at the open F value. In this case, acceleration / deceleration driving may be performed. When the scan interval exceeds 15 depths, the error in the peak position of the AF evaluation value obtained from the AF evaluation value obtained in the first scan becomes large, the scan range in non-addition readout becomes wide, and high speed cannot be achieved. there is a possibility. On the other hand, when the scan interval is less than 5 depths, the time of this first scan becomes long.

  When the scan interval is determined, the CPU 15 first controls the focus drive motor 22 via the second motor drive circuit 19 to drive the focus lens group 3 at the highest speed in FIG. 4A. Next, the focus lens group 3 is driven from “A” to “B” in FIG. 4, and an AF evaluation value is acquired for each of a plurality of BPFs set at a predetermined scan interval determined as described above. And the position ("C" in FIG. 4) where it becomes the maximum for every BPF and its reliability are calculated | required from the acquired AF evaluation value.

  In step S605, it is determined from the first scan result whether or not AF in the horizontal non-addition readout mode (second mode) is possible. If it is determined that AF is possible, it is determined from step S606 to step S607. If not, the process proceeds to step S630.

  Whether or not AF (second scan) in the horizontal non-addition readout mode is possible is determined as follows. The BPF setting in the first scan is set to the actual frequency in which the transmittance of the BPF is set to a relatively high frequency, and the BPF is set to a relatively low frequency in the second scan. The actual frequency at which the transmittance of the BPF thus obtained reaches a peak substantially matches. For example, in the first scan, if the BPF setting is set so that the transmittance peaks at 10%, 20%, 40%, and 50% with respect to the input signal, then in the case of horizontal three-pixel addition, 3. The transmission peak is set at 3%, 6.7%, 13.3%, and 16.7%. In the second scan, if the BPF setting is set so that the transmittance peaks at 13%, 17%, 33%, and 50% with respect to the input signal, two BPFs on the high frequency side of the first scan And the frequency at which the transmittance of the two BPFs on the low frequency side of the second scan peaks substantially coincide. In this way, the BPF is set, and it is determined whether the second scan is possible from the AF evaluation value obtained by each BPF and its reliability.

  First, it is a condition that the reliability of AF evaluation values obtained by the two BPFs on the first scan high frequency side is higher than a predetermined value. When the reliability is low, it indicates that the signal component from the subject in this band is small. In the second scan, the AF evaluation value is generated using a signal equivalent to or higher than the high frequency side, so that high reliability cannot be obtained from the AF evaluation value obtained from each BPF of the second scan. Thus, the in-focus position cannot be obtained.

  Furthermore, in order to perform the second scan, it is necessary that the AF evaluation value output on the first scan high frequency side is higher than the output on the first scan low frequency side. The higher output on the high frequency side indicates that there is a subject corresponding to the frequency on the high frequency side of the first scan. Since this subject may contain a higher frequency, it is meaningful to perform the second scan and obtain the focus position from the peak of the AF evaluation value for the high-frequency subject. Since there is a subject corresponding to at least the low frequency in the second scan, a highly reliable AF evaluation value can be obtained in the second scan.

  Conversely, when the output on the low frequency side is high, it indicates that the subject corresponding to the low frequency is dominant even if the AF evaluation value obtained by the BPF on the high frequency side is highly reliable. In this case, it cannot be expected that the AF evaluation value obtained by the BPF on the high frequency side of the second scan is high, and the AF evaluation value obtained by the BPF on the low frequency side is used when the second scan is performed. To obtain the in-focus position. In this case, the second scan is not performed because there is a demerit in that the second scan performs a time loss in switching the read mode and the read rate becomes slower than the mode used in the first scan.

  If it is determined in step S606 that the second scan is possible, the sensor reading mode is changed in step S607. The high-speed readout mode used in the first scan AF is changed to a non-horizontal addition readout mode for obtaining the in-focus position of the subject corresponding to the high frequency. For example, the readout mode is changed to a horizontal non-addition 30 fps readout mode. At the same time, the exposure is changed according to the sensor readout mode change. In the first scan, when the horizontal three-pixel addition readout mode is used and the vertical pixel addition element at that time is one pixel (non-addition), a total of three pixels are added. When the vertical addition number in the horizontal non-addition readout mode in the second scan is 1 pixel (non-addition), the total addition number is 1 pixel, so the second scan requires an exposure amount three times that in the first scan. It becomes.

  Therefore, the exposure is changed according to the program diagram of FIG. In the program diagram, the upper limit is appropriate luminance with the minimum diameter of the diaphragm 4 and the fastest value of the electronic shutter. Although this luminance depends on the sensitivity of the sensor, if the horizontal non-addition readout mode sensitivity is equivalent to ISO 100, the minimum diameter of the diaphragm 4 is F8, and the fastest electronic shutter speed is 1/2000 seconds, then Lv17. This becomes the tracking limit of the program diagram. Therefore, at Lv17 or higher, the exposure is Av = F8 (Av value = 6) and Tv = 1/2000 seconds (Tv value = 11). Every time the luminance decreases from this point, the Av value is decreased to the full aperture. If the maximum aperture is reached, the Tv value is decreased as the brightness decreases. Then, the digital gain is applied from the time when the lower limit of the shutter speed (Tv value = 5 = 1/30 seconds in the case of non-addition readout) is reached. However, there is a limit to this gain amount.

The same applies to the readout mode of horizontal addition, but the Tv value, Av value, and gain amount that achieve appropriate exposure at the same luminance are different because the sensitivity is increased by addition. For example, in the case of horizontal three-pixel addition readout, the sensitivity is three times that of non-addition readout, so if the sensitivity of non-addition readout is equivalent to ISO 100, it is equivalent to ISO 300. Therefore, the following limit of the program diagram is Lv = 15.41 (= 17−LOG ₂ (300/100)). Further, since the S / N ratio can be improved as the sensitivity increases, a gain is applied when the Tv value is large (the shutter speed is fast). For example, the gain is started to be applied when the Tv value = 7 (= 1/120 seconds), and the Tv value is decreased again from the time when the gain is multiplied by half (for example, 2.5 steps). The digital gain is applied from the time when the lower limit of the shutter speed (Tv value = 5 = 1/30 seconds) is reached.

  The exposure is changed in consideration of this sensitivity difference. When the horizontal three-pixel addition readout mode is changed to the horizontal non-addition readout mode, more exposure is required, so either the Tv value or the Av value is reduced or the gain is multiplied by the sensitivity difference. . If the luminance is such that the Tv value is the upper limit value, the Av value is reduced by 1.59. If the Av value before change is Av0, the Av value is changed to Av0-1.59. However, when the Av value reaches the open F value, the remaining change amount is performed by the Tv value. If the open F value is reached at Av0-1, the Tv value is reduced to -0.59. If the Av value is the open F value and no gain is applied, the Tv value is reduced by 1.59. However, when the Tv value reaches the lower limit, the remaining change amount is performed by gain. If the lower limit is reached when the first stage Tv value is lowered, the gain is multiplied by 1.5. In the case of luminance in a state where gain is applied, the gain amount is tripled and further gain is applied. However, if the gain limit is reached, no further gain is applied.

  In addition, in the case of luminance in which the Tv value multiplied by the gain is different between horizontal addition reading and horizontal non-addition reading, the Av value remains the open F value, but the Tv value and gain are changed. For example, when Av value = 3 (F2.8), Tv value = 7 (= 1/120 seconds), and gain = 3 steps (8 times), Av value = 3 (F2.8), Tv value = 5 ( = 1/30 seconds), gain is changed to 2.59 stages (about 6 times) (see FIG. 8).

  In the case of a CMOS sensor, in order to increase the reading speed, an average of a plurality of pixels may be taken instead of pixel addition to reduce the number of output pixels in the horizontal direction. For the sake of easy understanding, a reading mode that takes a pixel average is called a horizontal addition reading mode, and a reading mode that does not take a pixel average is called a horizontal non-addition reading mode. It is the same that high frequency information is lost. In this case, the required exposure amount does not change, but there is a concern that the S / N ratio may decrease due to a decrease in the number of pixels involved in signal generation. Therefore, in the case of non-additive reading, the gain is applied after the Tv value becomes smaller (the shutter speed is slow). For example, in the horizontal addition reading, gain is applied at a Tv value = 7 (= 1/120 seconds), and the Tv value is decreased again from the time when the gain is multiplied by half (for example, 2.5 stages). The digital gain is applied from the time when the lower limit of the shutter speed (Tv value = 5 = 1/30 seconds) is reached. In the case of horizontal non-addition readout, the digital gain is multiplied from the time when Tv value = 5 = 1/30 seconds.

  In step S608, the in-focus position is obtained from the AF evaluation value obtained by performing the second scan (second scan operation). The second scan interval is set to an appropriate interval of 5 depths or less. Here, since it is determined that the second scan is possible, the reliability of the AF evaluation values obtained with the two high-frequency BPFs is high, so the high-frequency BPF set in the first scan was obtained. A position where the AF evaluation value is maximized is obtained. Therefore, the scan interval is obtained from the difference between the maximum value and the AF evaluation value obtained by the highest BPF at the scan point adjacent to the maximum value.

The value obtained by normalizing the maximum value by the number of lines in the AF frame is d [i], and the value obtained by normalizing the AF evaluation value at the adjacent scan point by the number of lines in the AF frame is d [i−1], d. [I + 1], the difference between the maximum value and the minimum value of the signal value before BPF processing is obtained for each line in the AF frame, and the value obtained by extracting the maximum value among them is defined as MM.
Δ = max (d [i] −d [i−1], d [i] −d [i + 1]) ÷ MM (where max (a, b) is a function for selecting the maximum value of a and b) . This value is a value obtained by normalizing the gradient near the peak position of the AF evaluation value with the contrast of the subject. A large value indicates that the AF evaluation value shows a certain value only in the vicinity of the peak position, and the AF evaluation value is very small at a scan point away from the AF evaluation value. This indicates that a subject having a high spatial frequency is dominant. Conversely, when this value is small, it indicates that there is some AF evaluation value even at a scan point away from the peak position. This indicates that there is a subject with a low spatial frequency.

Therefore, the larger the value of Δ, the finer the scan interval is set. For example, the setting is as follows.
Δ> Δ1.5: Opening depth (1 depth)
1.5 ≧ Δ> 1: twice the open depth (2 depths)
1 ≧ Δ> 0.5: 3 times the open depth (5 depths)
0.5 ≧ Δ: 5 times the open depth (5 depths)
The scan range is determined by an error in obtaining a position where the AF evaluation value obtained by the highest BPF set in the first scan is maximized, and an AF resulting from a difference in frequency characteristics of the BPF used in obtaining the AF evaluation value. Determined by taking into account the deviation of the peak position of the evaluation value. The error in obtaining the peak position is mainly determined by the scan interval of the first scan. The error is about half of the scan interval. If you are scanning at 10-depth intervals, you can expect about 5 depths. The shift of the peak position of the AF evaluation value resulting from the difference in the frequency characteristics of the BPF is due to the optical characteristics of the photographing lens barrel 31. From the design tolerance, it is obtained by obtaining the difference between the peak position of the AF evaluation value in the BPF in the highest region of the first scan and the peak position of the AF evaluation value in the BPF in the highest region of the second scan. This amount is obtained by simulation based on manufacturing errors and design values of the optical elements constituting the photographing lens barrel 31 because optical characteristics differ depending on manufacturing errors even in the same designed lens barrel. What is necessary is to simulate the case where each optical element has an error on the side where the difference is displaced to the feeding side and the case where the optical element has an error on the side displaced to the feeding side, compared to the case where the optical element is manufactured as designed. .

  For example, when the photographic lens barrel 31 is manufactured according to the design value, the peak of the AF evaluation value in the BPF in the highest band of the second scan is the peak position of the AF evaluation value in the BPF of the highest band in the first scan. This is a case where the position is X-displaced to the retracting side. When each optical element has an error on the side displaced to the feeding side, the peak position of the AF evaluation value in the BPF in the highest region of the X-α second scan is displaced to the feeding side. If each optical element has an error on the side that is displaced toward the feeding side, the peak position of the AF evaluation value in the BPF in the highest region of the X + Β second scan is displaced toward the feeding side.

The scan range in this case is in the range from PX-β to PX-α with respect to the position P at which the AF evaluation value obtained by the highest BPF set in the first scan is maximum. The position where the AF evaluation value obtained by the highest BPF set in the second scan is maximized is distributed.
The scan range necessary for obtaining the peak position of the AF evaluation value is added to the two factors obtained in this way to obtain the scan range. That is,
PX-β-scan range from 1-second scan interval of the first scan to PX-α + 1/2 of the scan interval of the first scan + scan interval of the second scan And

  Also, the BPF setting is a highly reliable AF evaluation even when there is a transmittance peak at a high frequency sufficient to detect the in-focus position of the high frequency subject and the contrast of the high frequency subject is insufficient. Set to obtain a value. For example, a BPF having a transmittance peak at 10%, 20%, 40%, or 50% with respect to the input signal may be selected.

  The AF evaluation value obtained by controlling the focus lens group 3 and acquiring the AF evaluation value by the setting of the scan interval, the scan range, and the BPF, and the AF evaluation value obtained by each BPF having high reliability. Each time, the position where it is maximized and its reliability are obtained. When the reliability in the highest range is high, the peak position of the focus lens group 3 is obtained from the AF evaluation value obtained as a result, and is set as the in-focus position, and the focus lens group 3 is controlled to that position. When the reliability of the highest region is low, the peak position of the focus lens group 3 is obtained from the AF evaluation value obtained with the highest BPF among the BPFs that have obtained the highly reliable AF evaluation value signal, The position where the BP correction amount in the BPF is added is set as the in-focus position, and the focus lens group 3 is controlled to that position.

  Here, the processing after step S621 will be described. When it is determined that the light intensity is low or the contrast of the subject is low, processing in steps S621 to S622 (fourth scan) is performed. In step S621, the scan interval is set to about 5 times the open depth (5 depths), the scan range is the same as the first scan ("A" to "B" in FIG. 4), and the BPF is set to the same as the first scan. The focus lens group 3 is controlled to perform scan AF, and an AF evaluation value is acquired for each BPF. Then, the position where the AF evaluation value obtained with each BPF having the highest reliability of the AF evaluation value obtained with each BPF and the reliability thereof are obtained are obtained. In step S622, the peak position of the focus lens group 3 is selected from the AF evaluation values obtained from the highest BPF among the BPFs that have obtained highly reliable AF evaluation value signals. A position obtained by adding the BP correction amount in the BPF to the selected peak position value is set as an in-focus position, and the focus lens group 3 is controlled to that position.

  Here, the processing after step S630 will be described. If it is determined that the second scan is not possible, the processing of steps S630 to S631 (third scan) is performed. In step S630, when the AF evaluation value obtained by any BPF in the first scan is high, the scan interval is set to about five times the open depth (5 depths), and the BPF is the same as the first scan. Set. The scan range is set around the peak position P1 of the AF evaluation value obtained by the highest-reliability BPF among the AF evaluation values obtained as a result of the first scan. Specifically, centering on the peak position P1, the error E1 and the scan of the third scan when obtaining the position where the AF evaluation value obtained by the highest BPF set in the first scan before and after the peak position P1 is maximum. Set by adding an interval R3. That is, the scan range is P-E1-R3 to P + E1 + R3. When the reliability of AF evaluation values obtained with any BPF in the first scan is low, the scan interval is set to about 5 times the open depth (5 depths), and the BPF is set to be the same as the first scan. The scan range is set to be the same as the first scan ("A" to "B" in FIG. 4).

  Then, the focus lens group 3 is controlled to perform scan AF, and an AF evaluation value is obtained for each BPF. For each AF evaluation value obtained by the BPF, the AF evaluation value obtained by each BPF is highly reliable. Find the position where the maximum is and its reliability. In step S631, the peak position of the focus lens group 3 is selected from the AF evaluation values obtained from the highest BPF among the BPFs that have obtained highly reliable AF evaluation value signals. A position obtained by adding the BP correction amount in the BPF to the selected peak position value is set as an in-focus position, and the focus lens group 3 is controlled to that position.

  Although the first embodiment has been described by taking a compact digital camera as an example, the present invention can also be applied to a digital video camera and a digital SLR.

(Example 2)
Example 2 is an embodiment in which the horizontal pixel non-addition readout mode is fast enough to be regarded as equivalent to the horizontal addition mode (for example, 60 fps to 120 fps).

  The actual shooting operation will be described with reference to the flowchart shown in FIG. Parts that perform the same operations as in the first embodiment are given the same numbers, and detailed descriptions are omitted.

  When the main power switch of the imaging apparatus 1 is in the ON state and the operation mode of the imaging apparatus is in the shooting (recording) mode, the shooting process sequence is executed. After displaying the LCD in step S901 in FIG. 9, in step S902, the AE process is executed to measure the luminance of the subject, and the exposure when obtaining the AF evaluation value is optimized.

  In step S903, an AF evaluation value is acquired. Note that until the SW1 is determined to be on in step S904, the sensor reading mode is a reading mode in consideration of power consumption for LCD display. For example, the reading speed is about 30 fps, and the horizontal three-pixel vertical two-pixel addition reading mode is set. Is set. Using the AF evaluation value acquired in step S903, the readout mode and BPF used in the scan AF process in step S906 are determined.

  In step S904, when the CPU 15 confirms that SW1 (first stroke of the release switch) has been turned on, normal AE processing is executed in S905. Thereafter, a plurality of AF evaluation values are acquired in step S906, and scan AF processing is performed to detect the in-focus position from the result. Details of this processing will be described later.

  If the reliability of any AF evaluation value obtained in step S906 is sufficient, AFOK display is performed in step S907, and if the reliability of all AF evaluation value signals is low, the focus position is detected. In step S907, AFNG display is performed.

  In step S908, the CPU 15 confirms SW2 (second stroke of the release switch). If SW2 is turned on, the CPU 15 proceeds to step S909 and executes actual exposure processing.

  Here, details of the scan AF processing in step S906 will be described. In this embodiment, it is determined whether or not AF is possible in the horizontal non-addition readout mode from a plurality of AF evaluation values acquired before SW1 and its reliability. Scanning is performed to acquire AF evaluation values in a plurality of BPFs from low to high. Then, the focus lens position where the value is maximized is obtained from each AF evaluation value, and at the same time, each reliability is evaluated. If the reliability in the highest range is high, the focus lens group is obtained from the AF evaluation value obtained as a result. 3 peak position is selected as the in-focus position. As a result of scanning in the horizontal non-addition readout mode, if the reliability of the highest range is low, a position obtained by adding a BP correction amount to the BPF result obtained with a highly reliable AF evaluation value signal is set as the in-focus position. . If it is determined from the result of scanning at relatively coarse intervals that AF in the horizontal non-addition readout mode is impossible, scanning is performed in the horizontal addition readout mode, and a plurality of BPFs from low to high are scanned. An AF evaluation value is acquired. Of the BPFs obtained as a result of the AF evaluation value signal having high reliability, the peak position of the focus lens group 3 is obtained by the AF evaluation value obtained by the BPF set at the highest frequency, and the BP correction amount is obtained. The position that takes into account the focus position.

  The setting of the BPF in steps S903 and S906 is performed as follows. The frequency at which the transmittance at the time of AF evaluation value acquisition in step S903 is set to a relatively high frequency is set to a relatively high frequency, and the BPF is set to a relatively low frequency in the scan in step S906. The actual frequency at which the transmittance of the BPF thus obtained reaches a peak substantially matches. Here, the scan AF in step S906 is performed in a horizontal non-addition readout mode. For example, in the acquisition of the AF evaluation value in step S903, the BPF setting is set so that the transmittance peaks at 10%, 20%, 40%, and 50% with respect to the input signal. Then, when the readout mode during LCD display is horizontal three-pixel addition, the transmission peak is set at 3.3%, 6.7%, 13.3%, and 16.7%. Become. In the scan of step S906, if the BPF setting is set so that the transmittance peaks at 13%, 17%, 33%, and 50% with respect to the input signal, the two high-frequency bands of the former scan are set. The frequency at which the transmittance of the BPF and the two BPFs on the low frequency side of the latter scan peak is substantially the same.

  Next, a specific operation procedure will be described with reference to FIG. When the scan AF process is started, first, in step S1001, the illuminance at the time of shooting is confirmed. If it is darker than the predetermined illuminance, the process proceeds to step S1007, and if it is bright, the process proceeds to step S1002. The predetermined illuminance is set to such a brightness that a signal sufficient to detect the peak of the AF evaluation value in non-addition readout can be obtained. For this determination, the output of the AE processing circuit 13 in step S905 is used.

  In step S1007, the sensor reading mode is changed. The readout mode considering power consumption for LCD display is changed to a readout mode for performing scan AF at low illumination. In this readout mode, the number of readout lines is increased by performing pixel addition so that the SN ratio of the output signal becomes high even at low illuminance. For example, from a readout mode of about 30 fps for LCD display, horizontal 3 pixels, vertical 2 pixels addition, and readout line number 347 lines, scan AF 30 fps, horizontal 3 pixels, vertical 3 pixels addition, readout line number 1018 lines at low illuminance Change to read mode. Accordingly, the drive frequency of the sensor is also changed if necessary. At the same time, the exposure change accompanying the sensor readout mode change is performed according to the diagram for the addition readout mode shown in FIG.

  In step S1002, the subject contrast is determined. If the contrast is high, the process proceeds to step S1003. If the contrast is low, the process proceeds to step S1010. When the contrast is low, the amount of signal for generating the AF evaluation value is reduced. Therefore, as in the case of low illuminance, the high-frequency signal has a relatively increased noise superimposed on the sensor or circuit system, which is a false signal. May generate a signal. In this case, there is a high possibility that the peak of the AF evaluation value cannot be detected in non-addition readout. Therefore, the process proceeds to step S1010 in order to perform processing that does not use non-addition reading. For the determination of the contrast, the difference between the maximum value and the minimum value of the signal value before the high pass filter (HPF) processing is performed for each line in the AF frame among the evaluation values obtained by the AF processing circuit 14 in step S903. Find the maximum value and use it.

  In step S1003, it is determined whether or not a non-addition scan (fifth scan) is possible. If possible, the process proceeds to step S1004, and if not possible, the process proceeds to step S1010. Whether the fifth scan is possible is determined using the AF evaluation value acquired in step S903. In step S903, the frequencies at which the transmittances of the two BPFs on the low frequency side of the non-addition scan and the two BPFs on the high frequency side of the non-addition scan peak are set to substantially match. Therefore, it is a condition that the outputs of the two BPFs on the high frequency side are higher than the outputs of the two BPFs on the low frequency side. However, if the position is far from the in-focus position, the output of the high-frequency BPF is very small even if there is a subject corresponding to the high-frequency side. Therefore, in step S903, the focus lens group 3 is driven before SW1, AF evaluation values are acquired at a plurality of positions, and used as numerical values when determining the maximum value. If the timing for turning on SW1 is early and a sufficient range for driving the focus lens group 3 cannot be secured, it is determined that the fifth scan is possible. For example, when the focus lens group 3 is driven for less than 2/3 of the entire scan range, it is determined that the fifth scan is possible.

  In step S1010, the sensor reading mode is changed. The reading mode considering power consumption for LCD display is changed to a high-speed reading mode for performing scan AF. For example, the readout mode is changed from about 30 fps for LCD display to 120 fps for high-speed scan AF. Accordingly, the drive frequency of the sensor is also changed if necessary. At the same time, the exposure change associated with the sensor readout mode change is performed according to the diagram for the addition readout mode shown in FIG.

  Also in step S1004, the sensor reading mode is changed. The readout mode considering the power consumption for LCD display is changed to a non-addition and relatively high-speed readout mode for performing non-addition scan AF. For example, the readout mode is changed from about 30 fps for LCD display to 60 to 120 fps for non-addition scan AF. Accordingly, the drive frequency of the sensor is also changed if necessary. At the same time, the exposure change associated with the sensor readout mode change is performed according to the diagram for the non-additive readout mode shown in FIG.

  In step S1005, a fifth scan AF is performed. In this scan, the scan interval is set to about 5 times the open depth (5 depths), and the scan range is set to the entire range (“A” to “B” in FIG. 4) that should be the focus position in the set shooting mode. . The BPF is set so that the two low-frequency sides have approximately the same peak in transmittance as the two high-frequency BPFs at the time of AF evaluation value acquisition in step S902, and the high frequency corresponds to the frequency on the high-frequency side. The characteristic is set to have a transmittance peak in a sufficiently high range to detect the in-focus position. For example, in the acquisition of the AF evaluation value in step S902, the BPF setting is set so that the transmittance peaks at 10%, 20%, 40%, and 50% with respect to the input signal. Then, when the readout mode during LCD display is horizontal three-pixel addition, the transmission peak is set at 3.3%, 6.7%, 13.3%, and 16.7%. Become. Therefore, if the BPF is set so that the transmittance peaks at 13%, 17%, 33%, and 50% with respect to the input signal, the two BPFs on the high frequency side of the former scan and the latter The frequencies at which the transmittances of the two BPFs on the low frequency side of the scan peak are substantially the same. By setting a BPF having a transmittance peak at 50%, it is possible to detect the in-focus position of the subject corresponding to the high frequency side.

  With the above settings, the focus lens group 3 is controlled to perform scan AF, and an AF evaluation value is acquired for each BPF. And the position where it becomes the maximum for every BPF and its reliability are calculated | required from the acquired AF evaluation value. When the reliability in the highest range is high, the peak position of the focus lens group 3 is obtained from the AF evaluation value obtained as a result, and is set as the in-focus position, and the focus lens group 3 is controlled to that position. When the reliability of the highest region is low, the peak position of the focus lens group 3 is obtained from the AF evaluation value obtained with the highest BPF among the BPFs that have obtained the highly reliable AF evaluation value signal, The position where the BP correction amount in the BPF is added is set as the in-focus position, and the focus lens group 3 is controlled to that position.

In step S1006, the reliability of the fifth scan result is determined. When the reliability of all BPFs was low, it was impossible to detect the in-focus position by scanning in the horizontal non-addition readout mode, but it was possible to detect the in-focus position by scanning in the horizontal addition readout mode. There is sex. The reason is as follows.
(A) Since the high-frequency signal includes noise superimposed on the sensor and the circuit system, a false signal is generated.
(B) The high frequency component of the subject was lost due to the camera shake of the photographer or the subject moving.
(C) As a result of erroneously determining that there is a subject corresponding to the high frequency side, there is no high frequency component in the subject.

  For these reasons, if an AF evaluation value is acquired with a lower BPF, the in-focus position may be detected.

  Accordingly, for the most reliable AF evaluation values obtained from all BPFs, the percentage of the threshold value at which the reliability value is considered to be high is examined. When the value obtained as a result is equal to or larger than a predetermined value, there is a possibility that the peak position of the AF evaluation value obtained by the scan in the horizontal addition reading mode exists near the peak of the AF evaluation value obtained by the BPF. high. Conversely, when the value is less than the predetermined value, it is impossible to estimate the peak position of the AF evaluation value obtained by the scan in the horizontal addition reading mode. This predetermined value may be set to about 70% of the threshold value for which the reliability is high.

  Therefore, if the reliability value is greater than or equal to a predetermined value, the scan range of the sixth scan in step S1012 is limited to the vicinity of the peak position detected in the fifth scan. The value may be about 2 to 3 times the scan interval of the sixth scan before and after the peak position. On the other hand, when the reliability is less than the predetermined value, the scan range is set to the entire range (“A” to “B” in FIG. 4) that should be the focus position in the set shooting mode.

  If it is determined in step S1006 that the reliability of the fifth scan result is low, the process advances from step S1006 to step S1011. On the other hand, if it is determined that the reliability is high, the process proceeds to step S907 in FIG. 9 to perform AFOK display.

  In step S1011, the sensor reading mode is changed. The readout mode for non-addition scan AF is changed to a high-speed readout mode by addition for performing addition scan AF. For example, the readout mode is changed from 60 to 120 fps for non-addition scan AF to 120 fps for high-speed scan AF. Then, along with the sensor readout mode change, the exposure is changed according to the diagram for the non-addition readout mode shown in FIG. 7 and the diagram for the addition readout mode.

  Subsequently, in step S1012, a sixth scan in the horizontal addition reading mode is performed. The scan interval is set to about 5 times the open depth (5 depths), the BPF is the same as when the AF evaluation value is acquired in step S902, and the scan range is set as described above. Then, the focus lens group 3 is controlled to perform scan AF, and an AF evaluation value is acquired for each BPF. For each AF evaluation value obtained by the BPF with high reliability of the AF evaluation value obtained by each BPF. Find the position where it is maximized and its reliability. The peak position of the focus lens group 3 is obtained from the AF evaluation value obtained by the highest BPF among the BPFs that have obtained the highly reliable AF evaluation value signal, and the position obtained by adding the BP correction amount in the BPF to the value is obtained. The in-focus position is set, and the focus lens group 3 is controlled to that position.

  Here, the processing after step S1008 will be described. This operation is the same as in the first embodiment. If it is determined that the illumination is low or the contrast of the subject is low, processing (fourth scan) similar to steps S621 to S622 in FIG. 6 is performed.

  First, the scan interval is about 5 times the open depth (5 depths), the scan range is the same as the first scan of the first embodiment ("A" to "B" in FIG. 4), and the BPF setting is the first of the first embodiment. Set to the same as the scan, the focus lens group 3 is controlled to perform scan AF, and an AF evaluation value is acquired for each BPF. Then, the position where the AF evaluation value obtained with each BPF having the highest reliability of the AF evaluation value obtained with each BPF and the reliability thereof are obtained are obtained. Then, the peak position of the focus lens group 3 is selected from the AF evaluation values obtained with the highest BPF among the BPFs having obtained highly reliable AF evaluation value signals, and the BP correction amount in the BPF is added to that value. The focus position is set as the in-focus position, and the focus lens group 3 is controlled to that position.

(Example 3)
The third embodiment is a case where the number of BPFs that can be set is limited to two. The BPF in the first scan was set to a relatively low frequency side in the second scan, and the actual frequency in which the transmittance of the BPF peaked in the second scan was set to a relatively high frequency. The actual frequency at which the BPF transmittance reaches a peak is set to substantially match. For example, in the first scan, if the BPF is set to have a transmittance peak at 20% and 50% with respect to the input signal, it is 6.7% and 16.7% in the case of horizontal three-pixel addition. The transmission is set to have a peak. In the second scan, if the BPF is set to have a transmittance peak of 17% and 50% with respect to the input signal, the BPF on the high frequency side of the first scan and the low frequency side of the second scan The frequencies at which the BPF transmittance peaks are substantially the same.

  In this way, the BPF is set, and it is determined whether the second scan is possible from the AF evaluation value obtained by each BPF and its reliability. First, it is a condition that the AF evaluation value obtained by the BPF on the first scan high frequency side is highly reliable. Furthermore, in order to perform the second scan, it is necessary that the AF evaluation value output on the first scan high frequency side is higher than the output on the first scan low frequency side. The BPF settings for the third and fourth scans are the same as for the first scan.

  As described above, the first to third embodiments have been described by taking the compact type digital camera as an example, but the present invention can also be applied to a digital video camera and a digital SLR.

1 imaging device,
3 Focus lens group 5 Solid-state image sensor (sensor)
6 Imaging circuit 13 AE processing circuit 14 AF processing circuit 15 CPU with built-in memory for calculation

Claims (6)

Imaging means for photoelectrically converting a subject image formed by a photographing optical system including a focus lens to generate and output an imaging signal;
Control means for performing a scanning operation for acquiring an evaluation signal indicating an in-focus state of the focus lens while moving the focus lens based on the imaging signal, and controlling the position of the focus lens based on the acquired evaluation signal An automatic focus adjustment device comprising:
As a mode for reading the imaging signal from the imaging means, it has a first mode for reading while adding the imaging signal, and a second mode for reading without adding the imaging signal,
The control unit is configured to restrict reading of the imaging signal using the second mode when the second illuminance is lower than the first illuminance.   The control means performs a first scan operation for acquiring the imaging signal using the first mode, and in the case of the first illuminance, a second based on a result obtained from the first scan operation. The automatic focus adjustment apparatus according to claim 1, wherein the scanning operation is performed.   When the result obtained from the first scan operation satisfies a predetermined condition, the control unit reads the imaging signal in the second mode when performing the second scan operation. The automatic focus adjustment apparatus according to claim 2.   The automatic focus adjustment apparatus according to claim 3, wherein the predetermined condition is that the reliability of the evaluation signal obtained in the first scanning operation is higher than a predetermined value.   When the result obtained from the first scan operation does not satisfy a predetermined condition, the control unit reads the image pickup signal in the first mode when performing the second scan operation. The automatic focus adjustment apparatus according to claim 2.   In the case of the first illuminance, when the contrast based on the evaluation signal is higher than a predetermined value, the imaging in the first mode is faster in the first scan operation than in the case of the second illuminance. 6. The automatic focus adjustment apparatus according to claim 2, wherein a signal is read out.

Images