JP2016039462A - Imaging device and control method of the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect a frequency of noise without a significant increase in processing time even though there is no mechanism fro directly detecting a frequency of a noise source.SOLUTION: An imaging device comprises: an imaging element (102) for converting an optical image of a subject to an image signal to output the image signal; an AE sensor (116) for scanning each scan line with a predetermined drive period; an image processing part (105) for detecting a frequency of noise superimposed on the output signal by a signal output from the AE sensor and the predetermined drive period; and a total control part for calculating and setting a drive period for scanning each scan line of the imaging element based on the noise frequency detected by the image processing part. A period of fluctuation of an output of the imaging element, which is caused by noise in a direction perpendicular to the scan line, varies depending on the noise frequency and the drive period. The total control part calculates a drive period where the fluctuation period is larger than the predetermined period and sets the calculated period.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an imaging apparatus and a control method thereof.

  In a conventional imaging device, charges accumulated in a photodiode are converted into an electric signal by a readout circuit, and the converted electric signal is amplified and output. In this process, noise that causes a deterioration in image quality occurs in the image sensor. As a typical example, reset noise in a pixel readout circuit, noise due to variations in components constituting the readout circuit, and noise due to dark current generated in a pixel portion are known.

  In order to reduce these noises, an image sensor that reads out charges as described below is known. First, before reading out the charge from the photodiode, the subsequent circuit in the pixel excluding the photodiode and the pixel readout circuit are reset, and the noise component (N signal) generated during the reset operation is read out. Next, the charge (S signal) accumulated in the photodiode is transferred to a subsequent circuit in the pixel and read out. Then, the noise component is removed by subtracting the N signal from the read S signal (hereinafter referred to as “S-N operation”).

  However, since the sampling timing differs between the N signal and the S signal, if periodic noise is superimposed on the power supply line or ground line supplied to the image sensor during the image reading period of the image sensor, the image is striped by the SN operation. It is known that noise of the shape occurs. This is because if a difference occurs in the reference voltage when transferring the N signal and the S signal due to fluctuations in the power supply during image reading, this difference becomes an offset of the output signal of the image when the SN operation is performed. . Since reading is performed in units of one line, if the offset differs depending on the line to be read, it appears as horizontal stripe pattern noise (horizontal stripes) in the image.

  In order to correct the horizontal stripes generated in such an image, correction is performed by using the output of the horizontal optical black (HOB) area, which is a light-shielded pixel area, in addition to the non-light-shielded pixel area for outputting an image signal. How to do is known. For example, in the HOB area, an average value is calculated in the row direction to extract an offset, and a value obtained by moving and averaging the offset value for a plurality of rows in the column direction is generated as a correction value. By subtracting the correction value thus generated from the output signal of the image, the offset component can be corrected and the horizontal stripe noise can be corrected.

  As described above, the correction using the signal from the HOB area is performed by applying a certain moving average filter to the correction value so as not to be overcorrected or erroneously corrected due to the influence of random noise or micro-scratches in the HOB area. There is a need. For this reason, although correction is effective for removing low-frequency horizontal stripes, high-frequency horizontal stripes cannot be completely removed. Therefore, depending on the frequency of noise that generates horizontal stripes, the above correction may not provide a sufficient effect.

  For example, if the frequency of an oscillation circuit for driving a power source supplied to an image sensor serving as a noise source fluctuates due to a change in temperature or the like, the noise frequency of the reference voltage of the image sensor also fluctuates. When the frequency of the horizontal stripes generated in the image changes to the high frequency side due to the fluctuation of noise frequency, the horizontal stripes may remain.

  On the other hand, Patent Document 1 proposes a method for detecting noise appearing in a recorded image using the output of a second image sensor that is configured separately from the first image sensor for acquiring a recorded image. Yes. The output of the first image sensor and the output of the second image sensor are compared, and it is determined whether dust has adhered to the surface of the first image sensor.

JP 2011-024060 A

  However, the noise detection method disclosed in Patent Document 1 cannot identify the frequency of noise that causes horizontal stripes.

  Noise coming from a noise source outside the imaging apparatus includes magnetic noise and radio wave noise generated from an actuator or power supply circuit provided in a specific lens or accessory. For those noises that do not have a mechanism for directly detecting the noise frequency in the imaging apparatus, there is a method of calculating the noise frequency from the image output in which horizontal stripes are generated.

  However, with this method, it is necessary to drive the image sensor once before taking a recorded image, and horizontal stripes are generated in the first image. Further, since it is necessary to drive the image sensor once more than usual for detection of noise frequency, time and power consumption increase as compared with the case where frequency detection is not performed. In addition, the upper limit of the noise frequency that can be detected is lower than the magnetic noise, and the noise frequency may not be detected accurately.

  The present invention has been made in view of the above problems, and an object thereof is to detect the frequency of noise superimposed on an image signal without increasing the photographing time.

  In order to achieve the above object, an imaging apparatus of the present invention scans each scanning line with a first imaging element that converts an optical image of a subject into an image signal and outputs the image signal, and a predetermined driving cycle. A frequency of noise superimposed on the output signal is detected from a second image sensor different from the first image sensor, a signal output from the second image sensor, and the predetermined driving cycle. And a setting means for setting a driving cycle for scanning each scanning line of the first image sensor based on the frequency of noise detected by the detecting means, and the scanning line is caused by noise. The period of fluctuation of the output of the first image sensor that occurs in the direction perpendicular to the plane changes according to the frequency of the noise and the driving period, and the setting means has a predetermined period of fluctuation. WD that becomes larger than cycle Determined and set the period.

  According to the present invention, it is possible to detect the frequency of noise superimposed on an image signal without increasing the shooting time.

1 is a block diagram showing a schematic configuration of an imaging apparatus according to an embodiment of the present invention. FIG. 6 illustrates a pixel region of an imaging element in an embodiment. 1 is a schematic equivalent circuit diagram of one pixel of an image sensor and a readout circuit in an embodiment. 6 is a timing chart illustrating a normal reading operation of the image sensor according to the embodiment. The figure which shows the specification of the image pick-up element and AE sensor in embodiment. The figure which shows about the offset which arises by SN operation | movement when noise mixes into a power supply line at the time of image reading. The figure which shows the relationship between the frequency of noise, HD period, and horizontal stripes. 6 is a flowchart illustrating an operation of the imaging apparatus according to the first embodiment. The figure which shows the relationship between the HD period with respect to the noise frequency of 125 kHz in 1st Embodiment, and the period (row number) of a horizontal stripe. The figure which shows the specification of the image pick-up element and AE sensor in 2nd Embodiment. The figure which shows the relationship between the HD period with respect to the noise frequency of 500 kHz in 2nd Embodiment, and the period (row number) of a horizontal stripe.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the accompanying drawings.

<First Embodiment>
FIG. 1 is a block diagram showing a schematic configuration of an imaging apparatus 100 according to the first embodiment of the present invention. In FIG. 1, a well-known photographic lens (hereinafter simply referred to as “lens”) 101 is provided with a motor (not shown), and an optical image of a subject is formed on the light receiving surface of the image sensor 102 by focusing. Let me image. The image sensor 102 captures an optical image of the subject formed by the lens 101 as an image signal, and outputs it after performing noise removal processing described later.

  Here, the configuration of the image sensor 102 will be described. As shown in FIG. 2, the pixel area 200 of the image sensor 102 includes a plurality of pixels arranged two-dimensionally, and an effective pixel area 201 that is an exposed area and a horizontal optical black (a light-shielded area). HOB) area 202 (light-shielding area).

  FIG. 3 is a schematic equivalent circuit diagram of one pixel of the image sensor 102 and the readout circuit in the first embodiment. The pixel 301 includes a photodiode (PD) 302, a transfer switch 303, a floating diffusion portion (FD) 304, an amplification transistor 305, a reset switch 306, and a selection switch 307. A signal read from the pixel 301 is input to the reading circuit via the vertical output line 308. A constant current source 309 is connected to the vertical output line 308.

  The read circuit includes a clamp capacitor 310, a reset switch 311 for connecting to a reset power supply for resetting the read circuit, capacitors CTN and CTS, and switches 312 and 313 for controlling writing to the capacitors CTN and CTS. Furthermore, switches 314 and 315 for transferring signals written in the capacitors CTN and CTS to the differential amplifier 320 are included. Note that the readout circuit is shared by a plurality of pixels arranged in the same column.

  FIG. 4 is a timing chart showing a normal readout operation for one pixel in the pixel and readout circuit having the configuration shown in FIG. Here, it is assumed that a predetermined exposure time has ended and charges are accumulated in the PD 302. At time t1, the horizontal synchronization signal SYNC rises, and the control signal φSEL for the selected row (scanning line) changes from L to H. As a result, the selection switch 307 in the selected row is turned on, and a signal obtained by scanning the pixels in the selected row can be output to the vertical output line 308.

  At time t2, the reset pulse signal φRES changes from L to H, the reset switch 306 is turned on, and the potential of the FD 304 is reset. At time t3, the reset pulse signal φRES changes from H to L, the reset switch 306 is turned off, and the reset of the FD 304 is released. The potential of the FD 304 at this time is read out as a noise component (N signal) to the vertical output line 308 via the amplification transistor 305 and input to the clamp capacitor 310. At time t4, the control signal φTN1 is set to H, the switch 312 is turned on, and the clamped N signal is written into the capacitor CTN. Thereafter, at time t5, the control signal φTN1 is set to L, the switch 312 is turned off, and the writing of the N signal to the capacitor CTN is completed.

  Next, at time t6, the control signal φTS1 is set to H, the switch 313 is turned on, and a signal can be written to the capacitor CTS. Subsequently, at time t7, the transfer pulse signal φTX is set to H, and the charge (S signal) accumulated in the PD 302 is read to the FD 304. At time t8, the transfer pulse signal φTX is set to L. As a result, the FD 304 holds a signal (S + N signal) in which a noise component is superimposed on the S signal. The S + N signal held in the FD 304 is converted into a voltage by the amplification transistor 305, appears on the vertical output line 308, is clamped by the clamp capacitor 310, and is written in the capacitor CTS. At time t9, the control signal φTS1 is switched from H to L, the switch 313 is turned off, and writing to the capacitor CTS is completed. Through the operations so far, the N signal is held in the capacitor CTN and the S + N signal is held in the capacitor CTS.

  At time t10, φTS2 and φTN2 are simultaneously turned on, so that the N signal and the S + N signal held in the capacitor CTN and the capacitor CTS are simultaneously output to the differential amplifier 320, and the S signal that is the difference between them is output. (SN operation). At time t11, φTS2 and φTN2 are turned off, the control signal φSEL is changed from H to L, the selection switch 307 is turned off, and the reading of the pixels in the selected row is completed. By performing the differential operation in this manner, noise that differs for each pixel or readout circuit can be removed for each imaging operation, and noise generated in an image can be reduced.

  Returning to the description of FIG. 1, the A / D conversion unit 103 performs analog-digital conversion of the image signal output from the image sensor 102. The signal processing unit 104 performs various corrections, data compression, and the like on the image data output from the A / D conversion unit 103. Further, the signal processing unit 104 acquires a shading (offset) in the vertical direction using a signal obtained from the HOB area 202 of the image sensor 102 illustrated in FIG. Then, a moving average calculation is performed in units of a plurality of pixels in the vertical direction to calculate a correction value, and the calculated correction value is subtracted from the output signal of the pixel in the effective pixel region 201, thereby correcting horizontal stripes.

  The timing generator 113 outputs various drive timing pulses to the image sensor 102, A / D converter 103, and signal processor 104. The image processing unit 105 performs predetermined pixel interpolation processing and color conversion processing on the image signal output from the signal processing unit 104, and performs predetermined arithmetic processing using image data captured as necessary. . The image processing unit 105 also detects the horizontal stripe intensity and calculates the noise frequency. Here, the horizontal stripe strength is determined by, for example, the maximum value of the horizontal stripe amplitude. The memory unit 107 temporarily stores image data.

  The overall control unit 106 controls the entire imaging apparatus 100. The recording medium 109 includes a removable semiconductor memory for recording image data. The flash light emitting unit 111 performs AF auxiliary light projection and flash light control. The power supply unit 115 supplies a necessary voltage to each unit for a necessary period through a DC-DC converter circuit that converts a voltage supplied from a battery or the like into a desired voltage. The display unit 108 is made of, for example, a TFT type LCD.

  Reference numeral 116 denotes an AE sensor, which has the same structure as that of the image sensor 102 although the number of pixels is different. The brightness and color of the light source can be detected using the output of the AE sensor 116. In addition, since the AE sensor 116 and the image sensor 102 have different numbers of pixels in one row, the HD periods are different. FIG. 5 shows the specifications of the image sensor 102 and the AE sensor 116. The image sensor 102 has 6000 pixels arranged in the horizontal direction and 4000 pixels arranged in the vertical direction, and the AE sensor 116 has 400 pixels arranged in the horizontal direction and 300 pixels arranged in the vertical direction. In addition, it is assumed that the HD period TH1 (driving cycle) required to read one row (6000 pixels) of the image sensor 102 is 30 μsec. On the other hand, the AE sensor 116 has 400 pixels in one row and is 1/15 of the image sensor 102. Therefore, if the clock frequency of the read operation of the image sensor 102 and the AE sensor 116 is the same, the HD period TH2 (drive (Period) is 2 μsec.

  Next, the relationship between noise and horizontal stripes appearing in the image will be described. FIG. 6 shows an offset component generated in the SN operation when periodic noise is mixed in the reference voltage of the image sensor 102 at the time of image reading according to the first embodiment. In the image sensor 102, as described above, the timing for reading the N signal and the timing for reading the S + N signal are different (Δt), and if the reference voltage varies during reading, an offset occurs in the output signal for each row. This offset causes periodic horizontal stripes (periodic fluctuations in signal strength) in the image.

  The horizontal stripe Y when the sinusoidal noise is superimposed during such SN operation can be expressed by the following equation.

In the above equation (1), the frequency of the horizontal stripes generated in the image is the second term.

It depends on. That is, the frequency of the horizontal stripes varies depending on the noise frequency f and the one-row readout cycle (HD period) TH. Further, from the equation (2), it can be said that the frequency of the horizontal stripe of the image is equal to the case where the sine wave of the frequency f is sampled at the reciprocal f _h = 1 / TH of the HD period TH.
Note that a frequency higher than the Nyquist frequency f _h / 2 is folded by the sampling theorem.

  A specific example will be described with reference to FIG. FIG. 7 is a conceptual diagram showing the relationship between noise frequency, one-row readout cycle (HD period), and generated horizontal stripes. FIG. 7A shows a case where the noise frequency is smaller than the reciprocal of twice the HD period of the image sensor. FIG. 7B is a diagram illustrating a case where the noise frequency is equal to or greater than the inverse of twice the HD period.

  The image output is obtained by sampling the frequency of noise at the reciprocal frequency of the HD period. Therefore, as shown in FIG. 7A, when the noise frequency is smaller than the inverse of twice the HD period according to the sampling theorem, the noise frequency and the horizontal stripe frequency of the image correspond one-to-one. The frequency of noise can be detected from the horizontal stripe frequency. On the other hand, as shown in FIG. 7B, when the noise frequency is larger than the inverse of twice the HD period, aliasing occurs, and the noise frequency cannot be accurately detected from the horizontal stripe frequency.

  Therefore, the frequency of the horizontal stripe can be obtained by the following equation (3).

If the relationship between the noise frequency f and the HD period TH is expressed by the following equation (4) based on the sampling theorem, the horizontal stripe frequency of the image and the noise frequency f have a one-to-one correspondence.
f <1 / (2 × TH) (4)

  As described above, the relationship between the noise and the horizontal stripes appearing in the image changes depending on the relationship between the noise frequency f and the HD period TH, and when the relationship between the noise frequency f and the HD period TH satisfies Expression (4), the image It can be seen that the noise frequency f can be calculated from the frequency of the horizontal stripes appearing in FIG. In addition, when the relationship between the noise frequency f and the HD period TH does not satisfy Expression (4), it can be seen that the frequency of the horizontal stripes appearing in the image does not coincide with the noise frequency f. That is, by setting the HD period TH (first drive cycle) so that the relationship between the noise frequency and the HD period TH satisfies Expression (4), the noise frequency can be detected from the horizontal stripes appearing in the image. Is possible.

  In the example shown in FIG. 5 described above, since one row of the image sensor 102 is 6000 pixels and the HD period TH1 is 30 μsec, the maximum value of the noise frequency f detectable from the output of the image sensor 102 is about 16.2. 7 kHz. Further, since one row of the AE sensor 116 has 400 pixels and the HD period TH2 is 2 μsec, the maximum value of the noise frequency f that can be detected from the output of the AE sensor 116 is about 250 kHz. Thus, from the output of the image sensor 102, for example, the frequency f of noise of several tens of kHz cannot be accurately detected. On the other hand, it is possible to detect even the noise frequency f of several tens of kHz from the output of the AE sensor 116.

  Next, the operation of the imaging apparatus according to the first embodiment of the present invention will be described with reference to the flowchart of FIG. When the process is started, the image processing unit 105 obtains the output of the AE sensor 116 in S101. Here, the overall control unit 106 executes the operation of the AE sensor 116 when the release button is pressed halfway by an operation from an operation unit not shown in FIG. The AE sensor 116 sends an output to the A / D conversion unit 103, and the A / D converted output further passes through the signal processing unit 104 and reaches the image processing unit 105. In S <b> 102, the image processing unit 105 calculates an average value in the row direction of the output of the AE sensor 116, performs FFT (Fast Fourier Transform) on the calculation result, and calculates a noise frequency f.

In S <b> 103, the image processing unit 105 calculates an HD period T <b> 1 ′ (driving cycle) for performing the reading operation of the image sensor 102 so as not to generate a horizontal stripe with respect to the noise frequency f detected in S <b> 102. As a specific calculation performed by the image processing unit 105, first, a noise cycle Tn that is the reciprocal of the noise frequency f is calculated. If the image sensor 102 is driven at an integer multiple of the noise period Tn, no horizontal stripes appear in the recorded image. Therefore, an HD period T1 ′ that is an integral multiple of the noise period Tn and is equal to or longer than the HD period TH1, which is the HD period of the image sensor 102, is calculated using the following equation (5).
Tn × N = T1 ′ ≧ TH1 (N = 1, 2,...) (5)

  The reciprocal of the HD period T <b> 1 ′ thus obtained is a one-row readout frequency (hereinafter referred to as “HD frequency”) f <b> 1 ′ for performing the readout operation of the image sensor 102 so as not to generate horizontal stripes. In S104, the image pickup device 102 is driven at the HD frequency f1 'calculated in S104 to obtain an output without horizontal stripes, and the process is terminated.

  FIG. 9 shows the relationship between the HD period and the horizontal stripe period (number of rows) when the noise frequency is 125 kHz. As shown in FIGS. 9A and 9B, the period (number of rows) of horizontal stripes appearing in an image becomes infinite every 8 μsec, which is the period of noise. The fact that the period (number of rows) of horizontal stripes appearing in the image is infinite indicates that no horizontal stripes appear in the image. Therefore, if the HD period in which the period (number of rows) of horizontal stripes appearing in the image is infinite is set as the HD period of the image sensor 102, horizontal stripes do not appear in the image. When the HD period necessary for reading out pixels for one row of the image sensor 102 is 30 μsec as shown in FIG. 5, the minimum N satisfying the above equation (5) is 4, and the HD period T1 ′ Is 32 μsec. That is, if the HD period is changed to drive at 32 μsec, the horizontal stripes cannot be seen (FIG. 9C).

  Note that after the calculation of the HD frequency f1 ′ in S103, the HD frequency of the AE sensor 116 is driven to coincide with the detected noise frequency f before the image acquisition in S104, and the noise is again detected based on the output of the AE sensor 116. The frequency f may be detected. If the noise frequency f is not detected from the output of the AE sensor 116 driven at the detected noise frequency f, it means that the noise frequency f detected in S103 is correct.

  As described above, according to the first embodiment, horizontal stripes are generated in the output obtained by the image sensor 102 using the output of the second image sensor (AE sensor 116 in the first embodiment) different from the image sensor 102. The frequency f of noise to be calculated can be calculated. In addition, when the HD period of the second image sensor is shorter than the HD period of the image sensor 102, it is possible to detect higher frequency noise. In addition, it is possible to detect the noise frequency f without separately providing a mechanism for detecting the noise frequency f, and drive so as not to generate horizontal stripes. Furthermore, since the noise frequency f can be detected without operating the image sensor 102, it is possible to detect the noise frequency f in a shorter time and with less power consumption than in the prior art in which detection is performed using the output of the image sensor. Is possible.

<Second Embodiment>
Next, a second embodiment of the present invention will be described. In the second embodiment, noise can be detected more accurately by setting the HD period of the image sensor 102 to an integral multiple of the HD period of the AE sensor 116.

  When the HD period of the AE sensor 116 is an integer multiple of the period of the noise source, a situation may occur in which the horizontal stripes are not visible on the AE sensor 116 but the horizontal stripes are visible on the image sensor 102. However, if the HD period of the image sensor 102 is set to an integral multiple of the HD period of the AE sensor 116 as in the second embodiment described here, the image sensor can be used when no horizontal stripes are generated in the AE sensor 116. No horizontal stripes occur even at 102.

  As a specific example, FIG. 10 shows specifications of the image sensor 102 and the AE sensor 116 in the second embodiment. In the relationship shown in FIG. 10, the HD period of the AE sensor 116 and the HD period of the image sensor 102 are not in an integer multiple relationship. When the HD period becomes longer, the shooting time of one image is affected, so that the driving is performed in the HD period optimum for each element. Therefore, there is no relationship between the HD period of the image sensor 102 and the AE sensor 116.

  FIG. 11 is a diagram showing the horizontal stripe period (number of rows) of an image with respect to a noise frequency of 500 kHz. It can be confirmed that a peak appears for each noise period Tn in the HD period of 2 μsec or more exceeding the noise period Tn (FIG. 11A). The period of noise Tn can be accurately captured when the HD period is less than half the noise period Tn. When the image sensor 102 and the AE sensor 116 are driven under the conditions shown in FIG. 10, the HD period of the AE sensor 116 is 2 μsec, which matches the noise cycle, and horizontal stripes are seen in the output of the AE sensor 116. No (FIG. 11 (b)). However, since the HD period of the image sensor 102 is 27 μsec, as can be seen from FIG. 11C, a fine horizontal stripe having a cycle of about two rows is generated in the output of the image sensor 102.

  Accordingly, the HD period of the image sensor 102 is set to 28 μsec, which is an integral multiple of the HD period of the AE sensor 116, from the beginning. Thereby, even when an image is acquired by the image sensor 102, the period (number of rows) of horizontal stripes is maximized, and an image in which no horizontal stripes are generated can be obtained (FIG. 11C). That is, by setting the HD period of the image sensor 102 to an integer multiple of the HD period of the AE sensor 116 from the beginning, if no horizontal stripes are detected in the output of the AE sensor 116, the horizontal stripes are also detected in the output of the image sensor 102. Is not detected. Accordingly, it is possible to avoid a situation in which horizontal stripes are not detected in the output of the AE sensor 116, but horizontal stripes are generated in the output of the image sensor 102, and noise can be detected more accurately.

  As described above, when there is no mechanism for directly detecting the frequency of the noise source mixed in the image pickup device, the noise frequency can be detected without driving the image pickup device for recording image shooting. Further, it is possible to detect the noise frequency without increasing the shooting time depending on the presence or absence of the noise frequency detection operation.

  In the first and second embodiments, the case has been described in which a multiple of the noise period Tn is the HD period of the image sensor 102, but the present invention is not limited to this. For example, if the noise cycle (number of rows) is larger than the number of rows of the image sensor 102, horizontal stripes can be suppressed, and therefore a predetermined range around a multiple of the noise cycle (number of rows) may be used.

  102: Image sensor, 106: Overall control unit, 104: Signal processing unit, 105: Image processing unit, 113: Timing generation unit, 115: Power supply unit, 116: AE sensor

Claims (8)

A first image sensor that converts an optical image of a subject into an image signal and outputs the image signal;
A second imaging element different from the first imaging element that scans each scanning line at a predetermined driving cycle;
Detection means for detecting a frequency of noise superimposed on the output signal from the signal output from the second image sensor and the predetermined driving cycle;
Setting means for setting a driving cycle for scanning each scanning line of the first image sensor based on the frequency of noise detected by the detecting means;
The period of fluctuation of the output of the first image sensor that occurs in the direction perpendicular to the scanning line due to noise changes according to the frequency of the noise and the driving period, and the setting means An imaging apparatus characterized by obtaining and setting a driving cycle in which a cycle is larger than a predetermined cycle.   When the intensity of the signal output from the second image sensor periodically fluctuates in a direction perpendicular to the scanning line, the detecting means determines the frequency of the fluctuation and the predetermined value. The imaging apparatus according to claim 1, wherein a frequency of the noise is detected based on a driving cycle.   The setting means sets a multiple of the period of the noise as a driving period for scanning each scanning line of the first image sensor based on the frequency of the noise detected by the detecting means. The imaging device according to claim 1 or 2.   4. Each of the scanning lines of the second image sensor is configured by a smaller number of pixels than each scanning line of the first image sensor. 5. Imaging device.   5. The imaging apparatus according to claim 1, wherein the predetermined driving cycle is shorter than a time required for scanning each scanning line of the first imaging element. 6. .   The imaging apparatus according to claim 1, wherein brightness and color of a light source are detected based on an image signal obtained by the second imaging element.   The first image pickup device is driven by using a multiple of the predetermined drive cycle as a drive cycle of the first image pickup device before detecting a noise frequency by the detecting means. Item 7. The imaging device according to any one of Items 1 to 6. The detection means scans each scanning line at a predetermined drive cycle, and the noise output superimposed on the output signal from the signal output from the first image sensor and the predetermined drive cycle. A detection step for detecting the frequency;
Based on the frequency of the noise detected in the detection step, the setting means converts each scanning line of the second image sensor that converts an optical image of the subject into an image signal and outputs the image signal, which is different from the first image sensor. A setting step for setting a driving cycle to be scanned,
The period of fluctuation of the output of the second image sensor that occurs in the direction perpendicular to the scanning line due to noise changes according to the frequency of the noise and the driving period. In the setting step, the fluctuation of the fluctuation A control method for an imaging apparatus, characterized in that a driving cycle in which the cycle is larger than a predetermined cycle is obtained and set.

Images