JP2011151549A - Signal processing apparatus, imaging device, and signal processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method of effectively correcting striped noise from an image improved in S/N by means of pixel addition. <P>SOLUTION: An imaging device includes an imaging element having an aperture region for accumulating and outputting charge generated in response to incident light, and a shaded optical black region; an addition means to add pixel signals in the aperture region by a predetermined pixel unit number; a means to calculate an average value of rows and columns in an optional region of a signal obtained from the optical black region; a correction value calculation means to calculate a correction value, by multiplying the average value of the rows or the columns by a correction coefficient corresponding to a pixel addition unit number in the aperture region; and a means to correct the pixel signal in the aperture region by using the correction value. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to signal processing of pixel signals output from an image sensor.

  Imaging devices such as CMOS image sensors are used in imaging devices such as digital cameras and video cameras. In recent years, an increase in the number of pixels of an image sensor has progressed, and as a result, individual pixel sizes have become smaller. In addition, high-speed processing is required in order to maintain the usability such as continuous shooting speed even if an image sensor with a large number of pixels is used. As a high-speed processing method, there is a multi-channel method in which a plurality of output paths of an image sensor are used and a plurality of pixels are read simultaneously.

  On the other hand, since the optical signal becomes smaller due to the reduction in the pixel size, there is a problem that it is easily affected by noise. In addition, there is a problem that the amount of noise varies depending on the output path due to the influence of multi-channel. For example, a multi-channel CMOS image sensor has a plurality of output amplifiers. If there are variations in the characteristics of the output amplifiers, a level difference occurs between columns, resulting in stripe noise in the vertical direction. Will occur. In addition, since the power supply and GND (ground) of each pixel are common, if these power supply and GND fluctuate during reading of the selected row, a level difference occurs in the entire selected row, and as a result, a horizontal stripe shape is generated. Noise is generated.

  Since these striped noises cause deterioration in image quality, a reduction in noise is required. The stripe noise in the vertical direction is fixed noise determined by the output amplifier characteristics, and can be corrected by correcting the variation for each output amplifier. On the other hand, for horizontal stripe noise, if the fluctuations in the power supply and GND are random, the position and intensity at which the stripe noise occurs are also random. As a technique for correcting such random stripe noise, Patent Document 1 describes a method of calculating a correction value from a signal in a light-shielded pixel region and subtracting the correction value from a signal in an aperture pixel region. It is described that the correction value is calculated by multiplying the signal of the non-light-shielded pixel region by the correction coefficient in consideration of the difference in the influence on the power supply and GND variation due to the difference in structure between the light-shielded pixel region and the aperture pixel region. Has been.

JP 2008-67060 A JP-A-8-82005

  By the way, for the purpose of improving image quality and high-speed processing, there has been proposed an imaging apparatus capable of switching between an operation of adding and reading out pixel signals on a signal output line and an operation of reading out without adding signals (for example, see Patent Document 2). . Then, the random noise of the output image can be reduced by averaging the output signals of each row on the signal output line. However, the striped noise generated uniformly in the adding direction cannot be reduced. Further, the reduction of random noise makes the stripe noise more noticeable.

  Even if the correction value is calculated by multiplying the signal of the light-shielded pixel region by the correction coefficient and the aperture pixel region signal is corrected as in Patent Document 1 described above, such an imaging apparatus eliminates the remaining correction and overcorrection. In some cases, the residual noise or overcorrection may worsen the stripe noise. Furthermore, in the image with improved S / N due to pixel addition, the above-mentioned correction remaining and the stripe noise due to overcorrection become more conspicuous.

In order to solve the above-described problems, a signal processing apparatus according to the present invention includes an image pickup device including a pixel region including an opening region and a light-shielding region for outputting a pixel signal corresponding to accumulated charges, and an adding unit for adding the pixel signal. A signal processing device for performing a predetermined process on the pixel signal output from the difference value calculating means for calculating a difference value between an average value of the pixel signals in the light-shielding region and a predetermined black reference level;
Correction value calculation means for calculating a correction value by multiplying the difference value by a coefficient corresponding to the number of pixels added by the addition means, and correction means for correcting a pixel signal output from the aperture region using the correction value. It is characterized by having.

  Further, the signal processing method of the present invention is a pixel output from an image pickup device comprising a pixel region composed of an aperture region and a light shielding region for outputting a pixel signal corresponding to the accumulated charge, and an adding means for adding the pixel signal. A signal processing method for performing predetermined processing on a signal, calculating a difference value between an average value of pixel signals in the light-shielding region and a predetermined black reference level, and calculating a coefficient according to the number of added pixels by the adding means A correction value is calculated by multiplying the difference value, and the pixel signal output from the aperture region is corrected using the correction value.

  According to the present invention, stripe noise can be corrected without erroneous correction.

1 is an overall block diagram of an imaging apparatus according to the present invention. 1 is a schematic diagram of an image sensor according to the present invention. 6 is a circuit of a unit pixel of the image sensor in the present invention. 3 is a readout circuit of an image sensor according to the present invention. 4 is a timing chart of an all-pixel readout mode in the present invention. The timing chart of the horizontal 3 pixel addition average reading mode in this invention. The figure explaining the striped noise correction process in Example 1 of this invention. The figure explaining the striped noise correction process in Example 2 of this invention. FIG. 4 is a diagram illustrating an example of an output image of the imaging apparatus according to the present invention. The block diagram of AFE in this invention.

  Hereinafter, the best mode for carrying out the present invention will be described in detail with reference to examples.

  FIG. 1 is an overall block diagram of a solid-state imaging device (signal processing device) according to Embodiment 1 of the present invention. The CMOS solid-state image sensor 101 captures an image formed by a photographing lens (not shown). An analog front end (AFE) 102 performs horizontal OB clamping and analog-digital conversion processing. A digital front end (DFE) 103 receives the digital output of each pixel and performs digital processing such as image signal correction and pixel rearrangement. The image processing unit 104 performs various image processing such as predetermined pixel interpolation processing, color conversion processing, and striped noise correction processing on the digital output from the DFE 103. In the present embodiment, there is a feature point particularly about the striped noise correction processing, and details thereof will be described later.

  The memory circuit 105 is a working memory for the image processing unit 104, and is also used as a buffer memory in continuous shooting or the like. The control circuit 106 comprehensively controls the entire imaging apparatus, and incorporates a known CPU. The operation circuit 107 electrically receives an operation member in the digital camera. The display unit 108 for displaying an image or the like is, for example, a TFT (Thin Film Transistor) type LCD (Liquid Crystal Display). The recording circuit 109 is a recording medium such as a memory card or a hard disk. A timing generation circuit (TG) 110 generates various timings for driving the image sensor 101.

  FIG. 10 shows the internal configuration of the AFE 102. A gain control amplifier (AMP) 1001 is used for sensitivity adjustment, and amplifies the input image signal according to the set sensitivity. The horizontal OB clamp circuit 1002 performs offset correction so that the difference between the output of the HOB area of each row and the black level reference value is reduced in order to correct the gentle dark shading in the row direction and match the black level reference value. The correction here is fed back to the AMP 1001. Since the correction amount is integrated as the line advances, it follows only a gradual change and is not affected by noise in the HOB region. The horizontal OB clamping process may be performed on the digital output in the DFE 103.

  An analog-to-digital conversion circuit (AD) 1003 applies analog gain to the pixel signal from the image sensor 101 by applying a gain so that the pixel output in the HOB area matches the black level reference value by the AMP 1001 and the horizontal OB clamp circuit 1002. The signal is converted into a 14-bit digital signal, for example.

  FIG. 2 is a configuration example of a pixel region of a CMOS type solid-state imaging device. As shown in FIG. 2, the solid-state imaging device according to the present embodiment includes an aperture pixel area 203 (aperture area), a horizontal optical black area (HOB, horizontal light shielding area) 201, and a vertical optical black area (VOB, vertical direction). A light shielding region) 202. The aperture pixel region 203 accumulates and outputs charges generated according to incident light. A horizontal optical black area (HOB) 201 and a vertical optical black area (VOB) 202 are light-shielded areas provided adjacent to the aperture pixel area 203.

  FIG. 3 is an example of a circuit of unit pixels (for one pixel) of the CMOS image sensor. The photodiode 301 converts light into electric charge. The transfer switch 302 is formed of a MOS transistor and is controlled by a PTX pulse described later. The source follower MOS transistor 304 amplifies the voltage of the floating diffusion (FD) unit 303. The unit pixels are arranged in a matrix, and the pixels in the same column are connected to a common vertical output line 307 via a selection switch 305. The selection switch 305 is controlled by a PSEL pulse described later. A reset switch 306 formed of a MOS transistor resets the PD 301 to the potential VDD of the common power supply VDD 308 via the FD 303 and the transfer switch 302. The reset switch 306 is controlled by a PRES pulse described later.

  FIG. 4 is a diagram for explaining a readout circuit of a CMOS type solid-state imaging device. Here, for convenience of explanation, FIG. 4 shows only a readout circuit from which pixel signals in even columns are read out. A plurality of unit pixels 300 shown in FIG. 3 are arranged in the HOB area and the opening area, respectively. A common signal such as PRES, PTX, PSEL or the like is input to each pixel from a vertical shift register (not shown) for each row. In the example of FIG. 4, the 0th to 5th rows are HOB regions, and the 6th and subsequent rows are open regions. Note that the VOB area is omitted.

  The output of each pixel is input to the capacitor (c0) 401 in the column circuit via a vertical output line to which the current source load 400 is connected. The operational amplifier 403 constitutes an inverting amplifier by the feedback capacitor (cf) 402 and the capacitor (c0) 401. By short-circuiting both ends of the feedback capacitor (cf) 402 by an analog switch (not shown) controlled by a PC0R signal (not shown), the c0 and cf capacitors are reset and the subsequent capacitors cts and ctn are reset. The output of the operational amplifier 403 is held in holding capacitors (cts) 410 and (ctn) 411 via analog switches 404 and 405 driven by PTS and PTN signals, respectively. Here, the reset signal immediately after resetting the FD is held in the holding capacitor (ctn) 411, and then the signal immediately after the signal from the pixel is transferred to the FD is held in the holding capacitor (cts) 410. .

  When the pixel signals for one row are held in the holding capacitor (cts) 410 and the holding capacitor (ctn) 411 for each column, the horizontal scanning circuit 414 sequentially drives PH (n) pulses. Then, the analog switches 412 and 413 are opened and closed, and input to the differential read amplifier 415 at the subsequent stage and output to the outside. Analog switches 406 to 409 (adding means) are switches that connect the storage capacitors of the same color adjacent columns and are driven by PHADD pulses. When the analog switches 406 to 409 are turned on once and then turned off, the addition average potential with the potential held in the holding capacitor in the adjacent column is held in the holding capacitors (cts) 410 and (ctn) 411.

  Here, the analog switches 406 to 409 are arranged in a column in which pixel signals in the opening area are output, and are not arranged in a column in which pixel signals in the HOB area are output. In other words, the OB area is always read without pixel addition regardless of whether or not the pixel is added to the opening area. If the pixels in the HOB area are added and read out, the area that can be used for clamping decreases when the clamping process is performed using the signal in the HOB area in the AFE 102, and the clamping error increases. Therefore, in order to reduce this clamping error, all pixels are read from the HOB area even in the addition reading mode.

  When the horizontal scanning circuit 414 determines the MODE signal and performs horizontal addition reading, only a signal in a specific column can be read from the reading amplifier 415 by thinning out PH (n) pulses. For example, when the MODE signal indicates all-pixel reading, a signal to 0 → 2 → 4 → 6 → 8 →... Is output, and when the MODE signal performs horizontal three-pixel addition average reading, 0 → 2 → 4 → 8 → 14 →... Is determined in advance so as to output a signal to the column and switched.

  FIG. 5 is an example of a timing chart for driving the CMOS image sensor, and shows the timing when all the pixels are read out. However, in order to simplify the description, the timing for starting resetting all pixels and starting accumulation is omitted, and only the timing for the pixel readout operation in the 0th row in FIG. 4 is shown. In FIG. 5, it is assumed that the transistors and switches that are driven when the signal level is the Hi level are turned on, and the transistors and switches that are driven when the signal level is the Lo level are turned off.

  First, when PRES changes from Hi to Lo, the reset switch 306 is turned OFF, and the FD 303 is released from the reset to the power supply VDD 308. Thereafter, when PSEL changes from Lo to Hi, the selection switch 305 is turned ON, and the source follower output of each pixel in the 0th row is connected to each vertical output line. Thereafter, the PC0R signal is set to Hi to turn on a switch for short-circuiting both ends of a column circuit feedback capacitor (cf) (not shown), and the PTS and PTN signals are set to Hi to turn on the analog switches 404 and 405. Then, the storage capacitors (cts) 410, (ctn) 411, the feedback capacitor (cf) 402, and the capacitor (c0) 401 are reset. Thereafter, the PTS and PTN signals are set to Lo, and the PC0R signal is set to Lo, thereby releasing the reset of each capacitor.

  Thereafter, the PTN signal is changed from Lo → Hi → Lo in order to hold the reset output of the FD 303 in the holding capacitor (ctn) 411. Thereafter, the PTS signal is set to Hi to hold the pixel signal output in the holding capacitor (cts) 410, and the PTX signal is changed from Lo → Hi → Lo to transfer the pixel signal to the FD 303. Thereafter, the pixel signal is held in the holding capacitor (cts) 410 by setting the PTS signal to Lo. Thereafter, FD303 is reset again by setting PSEL to Lo and PRES to Hi. Then, the PH (n) signal is sequentially output by the horizontal scanning circuit 414, and each pixel signal in the 0th row is sequentially read to the read amplifier 415. Similarly, the pixel signals in the first and subsequent rows are sequentially read out, and the reading of all the pixels is completed. At this time, since the PHADD signal remains Lo, the output signal is not subjected to horizontal addition averaging.

  FIG. 6 is a timing chart when the CMOS image sensor is driven in the horizontal three-pixel addition average mode. The process until the reset output and the pixel signal output of the FD 303 are held in the holding capacitors (cts) 410 and (ctn) 411 is the same as that in the timing chart of FIG. After that, by changing the PHADD signal from Lo → Hi → Lo, the signals in the open regions 6, 8, 10 and 12, 14, 16 are added on the holding capacitors (cts) 410 and (ctn) 411, respectively. Average. At this time, the horizontal scanning circuit 414 outputs only the PH (n) signals in the columns 0 → 2 → 4 → 8 → 14 →... From the region, the horizontal three-pixel addition average signal is read out every three columns.

  Next, the striped noise correction process performed in the image processing unit 104 in FIG. 1 will be described. The striped noise correction process described here is a process for correcting horizontal line noise that occurs in the entire selected row due to power supply fluctuation during reading of the selected row. The HOB clamp corrects a component that gradually changes in the row direction, whereas the striped noise correction processing is for correcting high-frequency component noise that occurs in the row direction.

  FIG. 9A is an example of an output image in the all-pixel readout mode. If the standard deviation of the random noise amount in the opening region is σ, the standard deviation of the random noise amount in the HOB region is also σ. FIG. 9B is an example of an output image in the horizontal three-pixel addition readout mode. Assuming that the random noise amount at the time of reading all pixels is σ, the random noise amount σ ′ of the image output from the aperture area obtained by averaging three pixels is obtained as shown in Equation 1.

  That is, by averaging the three pixels, the amount of random noise is reduced to 1 / √3 when all the pixels are read. However, since all pixels are read out in the OB area, the amount of random noise in this area is σ. By the way, since the horizontal stripe noise is uniformly generated in the entire readout row due to fluctuations in the power supply and GND during the readout operation, the intensity of the stripe noise does not change depending on whether or not the horizontal pixel is added. However, in the horizontal three-pixel addition mode, the amount of random noise in the aperture region is reduced to 1 / √3, so that stripe noise becomes more conspicuous than in the all-pixel readout mode.

  In addition, since the random noise amount in the HOB area is the same in both the all-pixel readout mode and the horizontal 3-pixel addition readout mode, the correction error when the stripe noise correction is performed using the output of the HOB area is approximately the same. is expected. However, in the horizontal three-pixel addition readout mode, since the random noise in the aperture region is small, the stripe noise caused by the correction error is more conspicuous than in the all-pixel readout mode. Therefore, in the present embodiment, when pixels are added in the opening area, striped noise correction is performed in consideration of the difference in the random noise amount between the HOB area and the opening area.

  FIG. 7 is a diagram illustrating the striped noise correction processing according to the first embodiment. This correction processing is performed by the image processing unit 104 in FIG.

  First, in a process 701 (difference value calculation means), a value A is calculated by subtracting a predetermined black reference level from the row average value of the HOB area of the pixel signal input to the image processing unit 104. Next, in a process 702 (tap processing means), a value A1 obtained by performing a tap process on the value A for each row calculated in the process 701 is calculated.

  Here, the tap process is a process of calculating an average value of pixel signals of a plurality of rows before and after the target row with respect to the pixel signal of the target row. Here, the number of rows used to calculate the average value is referred to as the number of taps. For example, when a tap process with 3 taps is performed on the i-th row, the average value for a total of 3 rows of the i-1th row, the i-th row, and the i + 1-th row is calculated. In this embodiment, lines before and after the target line are used for the tap process, but only the previous line or the subsequent line of the target line may be used.

  As an example of the number of taps, it is appropriate that the number of taps is about 3 taps in the all-pixel readout mode and about 5 taps in the horizontal three-pixel addition readout mode. In the horizontal three-pixel addition readout mode, random noise in the aperture region is reduced by pixel addition, and stripe noise becomes more conspicuous. Therefore, in the addition readout mode, the number of taps is increased as compared with the all-pixel readout mode to correct the relatively low-frequency stripe noise that is visually noticeable. Further, if the number of taps is increased, the number of pixels used for correction value calculation is increased, which has the effect of reducing correction errors.

  In the process 703 (correction value calculation means), the final correction value is calculated by multiplying the value A1 subjected to the tap process in the process 702 by a correction coefficient. An error due to the influence of random noise in the HOB region occurs in the row average value calculated from the HOB region. Therefore, when the difference between the row average value and the predetermined black reference level is directly subtracted from the pixel signal of the aperture region as a correction value, a uniform offset component may be generated in the row direction due to the error. As a result, stripe noise may be worse than before correction.

  In order to reduce the error, the number of HOB columns used to calculate the row average value may be increased. However, there is a demerit that an increase in the number of HOB columns directly leads to an increase in the circuit scale of the image sensor. Therefore, the amount of correction is reduced by multiplying the difference between the row average value and a predetermined black reference level by a correction coefficient of 1 or less, thereby preventing stripe noise from deteriorating due to erroneous correction. Further, the larger the number of pixel addition units (number of added pixels) in the aperture region, the smaller the random noise and the more noticeable stripe noise caused by erroneous correction. Therefore, the correction coefficient has the number of pixel addition units (number of added pixels). It is better to make it smaller as it increases. However, if the correction coefficient is made too small, the effect of correcting stripe noise that is originally desired to be corrected is also reduced. Therefore, in this embodiment, the correction value c for each row is calculated by performing the calculation shown in Expression 2.

  Here, the coefficient k is determined in consideration of the difference in the intensity of the stripe noise due to the difference in structure between the opening area and the HOB area, and the difference in the appearance of the stripe noise due to the difference in the display method for still images and moving images. Correction coefficient. The value A1 is a value calculated by performing the tap process in the process 702. i (n) is a correction coefficient determined by the pixel addition unit number n (addition pixel number) of the aperture region. The correction coefficient i (n) has a value as shown in Equation 3.

  By multiplying the correction coefficient i (n) shown in Expression 3, it is possible to suppress erroneous correction even when random noise in the aperture region is reduced by pixel addition. In addition, when n pixels are added in the aperture area and m pixels are added in the HOB area as this application example, i (n, m) shown in Expression 4 may be used instead of i (n).

  In processing 704 (correction means), the stripe noise is removed by subtracting the correction value c for each row calculated in processing 703 for each row from the opening area. By performing the above processing, even in an image in which random noise components are reduced by pixel addition in the aperture area, stripe noise can be effectively eliminated without increasing new stripe noise due to erroneous correction. Can be removed. In this embodiment, the correction processing is performed by the image processing unit 104 in FIG. 1, but the same correction processing may be performed by the DFE 103.

  Striped noise correction processing performed by the image processing unit 104 according to the second exemplary embodiment of the present invention will be described with reference to FIG. Note that portions other than the striped noise correction processing are the same as those in the first embodiment, and thus are omitted.

  In processing 801 (difference value calculation means), a value A is calculated by subtracting a predetermined black reference level from the row average value for each row in the HOB area of the image input to the image processing unit 104. Next, in the process 802, the tap process is performed on the value A for each row calculated in the process 801 similarly to the process 702 in FIG. 7, and the value A1 is calculated. The value A1 mainly includes a low frequency component of the striped noise. Subsequently, in a process 803 (correction value calculation means), a value A2 obtained by subtracting the value A1 from the value A is calculated. The value A2 mainly includes high frequency components in the striped noise. In process 804 (correction value calculation means), as shown in equation 5, the correction value c1 is calculated by multiplying A1 calculated in process 802 by the correction coefficient k1.

  Here, the coefficient k1 is determined in consideration of the difference in the intensity of the stripe noise due to the difference in structure between the opening area and the HOB area, and the difference in the appearance of the stripe noise due to the difference in the display method for still images and moving images. Correction coefficient. The correction value c1 is mainly for correcting the low frequency component of the striped noise. Here, the value A1 is tapped to increase the number of pixels used for correction value calculation, and the influence of random noise in the HOB area is reduced. Therefore, even if the correction amount of the striped noise is increased, it is unlikely to be erroneously corrected. In process 805 (correction value calculation means), as shown in Expression 6, the value A2 calculated in Process 803 is multiplied by the correction coefficient k2 and the correction coefficient i (n) shown in Expression 3 of the first embodiment. A correction value c2 is calculated.

  Here, the coefficient k2 is determined in consideration of the difference in the intensity of the stripe noise due to the difference in structure between the opening area and the HOB area, and the difference in the appearance of the stripe noise due to the difference in the display method for still images and moving images. Correction coefficient. The correction value c2 is mainly for correcting high-frequency components in the striped noise. Here, since the tap processing is not performed on the value A2, the number of pixels used for calculating the correction value is small, and the value A2 is easily influenced by random noise in the HOB area, and a correction error is likely to occur. Therefore, the correction coefficient i (n) is multiplied so that the amount of correction becomes smaller as the number of pixel addition units (number of added pixels) in the opening area increases, that is, as the image has less random noise. As a result, it is possible to suppress striped noise caused by erroneous correction.

  In process 806 (correction value calculation means), the final correction value c is calculated by adding the correction value c1 and the correction value c2. In the process 807, the stripe noise component is removed by subtracting the correction value c calculated in the process 806 for each row from the opening area.

  As described above, the striped noise detected from the HOB area is divided into a low-frequency component and a high-frequency component, a correction coefficient close to 1 is used for a low-frequency component with a small correction error, and a high-frequency component that is likely to generate a correction error is A correction coefficient corresponding to the number of pixel addition units (number of added pixels) in the opening area is used. Thereby, the stripe noise can be effectively removed without generating new stripe noise due to erroneous correction.

  The present invention is also realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, or the like) of the system or apparatus reads the program. It is a process to be executed.

101 Solid-state image sensor 102 Analog front end (AFE)
103 Digital Front End (DFE)
104 Image Processing Unit 201 Horizontal Optical Black Area 202 Vertical Optical Black Area 203 Open Pixel Area 1002 Horizontal OB Clamp Circuit

Claims (9)

A signal processing device that performs a predetermined process on a pixel signal output from an imaging device, comprising: a pixel region composed of an aperture region and a light-shielding region for outputting a pixel signal corresponding to the accumulated charge; and an adding means for adding the pixel signal Because
Difference value calculating means for calculating a difference value between an average value of pixel signals in the light-shielding region and a predetermined black reference level;
Correction value calculating means for calculating a correction value by multiplying the difference value by a coefficient corresponding to the number of added pixels by the adding means;
Correction means for correcting the pixel signal output from the aperture region using the correction value;
A signal processing apparatus comprising:   The signal processing apparatus according to claim 1, further comprising a tap processing unit that performs a tap process on the difference value with the number of taps corresponding to the number of pixels added by the adding unit.   The signal processing apparatus according to claim 1, wherein the correction value calculation unit multiplies a high frequency component included in the difference value by a coefficient corresponding to the number of added pixels by the addition unit.   The correction value calculation means calculates the correction value by multiplying the difference value by a coefficient corresponding to a ratio between the number of added pixels of the pixel signal in the opening region and the number of added pixels of the pixel signal in the light shielding region. The signal processing apparatus according to any one of claims 1 to 3.   An image pickup apparatus comprising: the image pickup element; and the signal processing apparatus according to claim 1.   A signal processing method for performing a predetermined process on a pixel signal output from an image pickup device comprising: a pixel region including an opening region and a light shielding region for outputting a pixel signal corresponding to accumulated charge; and an adding unit for adding the pixel signal The difference value between the average value of the pixel signals in the light shielding area and a predetermined black reference level is calculated, and the correction value is calculated by multiplying the difference value by a coefficient corresponding to the number of added pixels by the adding means. And correcting the pixel signal output from the aperture region using the correction value.   The signal processing method according to claim 6, wherein the tap processing is performed on the difference value with the number of taps corresponding to the number of added pixels by the adding unit.   8. The signal processing method according to claim 6, wherein a high frequency component included in the difference value is multiplied by a coefficient corresponding to the number of pixels added by the adding means.   7. The correction value is calculated by multiplying the difference value by a coefficient corresponding to a ratio between the number of added pixels of the pixel signal in the aperture region and the number of added pixels of the pixel signal in the light shielding region. The signal processing method of any one of thru | or 8.

Images