JP5293447B2 - Electronic camera - Google Patents

Description

  The present invention relates to an electronic camera that acquires an image by receiving subject light.

  2. Description of the Related Art In recent years, electronic cameras such as digital cameras that have been widely used perform main shooting in a state in which subject light is received and light-shielded shooting in a state in which subject light is blocked, and obtain a difference between images obtained by the shooting. Thus, it has been devised to acquire an image from which noise components have been removed (see Patent Document 1). In addition, by obtaining the difference between the signal level of each pixel immediately after resetting the image sensor incorporated in the camera body and the signal level of each pixel when the subject light is received, reset noise and each pixel generated in each pixel can be reduced. It has also been devised to remove fixed pattern noise (FPN: Fixed Pattern Noise) peculiar to circuit elements such as photodiodes.

  As a method for removing the noise component described above, for example, an average signal level obtained by averaging signal levels of pixels in the same pixel column among a plurality of pixels arranged in a matrix on the image sensor is calculated and calculated. For example, the average signal level is subtracted from the signal level of the pixel obtained during the main photographing.

Japanese Patent No. 4146945

  However, when a plurality of pixels arranged in the image sensor includes, for example, pixels having a signal level higher than the signal level of other pixels (so-called white point pixels), the average signal level of the pixel column is different from that of other pixels. It becomes higher than the average signal level of the pixel column. For this reason, when the subtraction process using the average signal level is executed, a value larger than the actual noise component is subtracted, and a pixel defect called a vertical stripe is generated in the noise-removed main image. . Thus, when the white point pixel is included in the plurality of pixels arranged in the image sensor, the above-described noise removal cannot be performed with high accuracy.

  The present invention provides an electronic camera capable of performing highly accurate noise correction on an image obtained at the time of photographing even when white point pixels are included as a plurality of pixels arranged in an image sensor. The purpose is to provide.

In order to solve the above-described problems, an electronic camera according to the present invention includes an imaging device in which a plurality of pixels that photoelectrically convert incident light are arranged in a matrix, and each of the plurality of pixels arranged in the imaging device. The incident light received by each of the plurality of pixels arranged in the image sensor, and is output as a first pixel signal a pixel signal based on the charge accumulated in a state where the image sensor is shielded from light A control unit that outputs a pixel signal based on the electric charge obtained when the signal is photoelectrically converted as a second pixel signal, and a first pixel signal having a preset signal value among the first pixel signals output from the plurality of pixels. when including the first pixel signal exceeds a second threshold value larger than the threshold value, the first pixel signal exceeds the second threshold value, which is disposed on the periphery of the pixel which outputs the first pixel signal periphery After corrected based on the first pixel signal output from the element, using at least pre excluding the set the first pixel signal exceeds a first threshold value and the first pixel signal remaining included in each pixel row A correction unit that calculates a correction value for each pixel column and corrects the second pixel signal for each pixel column using the calculated correction value for each pixel column. And

The correction unit may replace a first pixel signal exceeding the second threshold with a first pixel signal output from any one of the peripheral pixels.

The correction unit may correct the first pixel signal exceeding the second threshold by performing an interpolation process using the first pixel signal output from the peripheral pixel.

  The plurality of pixels block a photoelectric conversion unit that photoelectrically converts the incident light, a charge storage unit that stores charges photoelectrically converted by the photoelectric conversion unit, and the photoelectric conversion unit and the charge storage unit. Or a switching element to be connected, and the control unit outputs a pixel signal output in a state where the photoelectric conversion unit and the charge storage unit are blocked by the switching element, and the photoelectric conversion unit and the charge. It is preferable to output the first pixel signal and the second pixel signal by obtaining a difference from the pixel signal output in a state where the storage unit is connected.

  Further, the imaging device includes a transfer unit that transfers a pixel signal from each pixel of the plurality of pixels in a pixel column direction, and an amplifier that amplifies the pixel signal read by the transfer unit for each pixel column, Further, it is preferable that the correction unit calculates the correction value by averaging the signal value of the first pixel signal for each pixel column amplified by the amplifier.

  According to the present invention, it is possible to perform highly accurate noise removal even when white point pixels are included as the plurality of pixels arranged in the image sensor.

It is a figure which shows the structure of the digital camera of this invention. It is a circuit diagram which shows the structure of the pixel of an image pick-up element. It is a figure which shows the procedure which reads the pixel signal of each pixel. It is explanatory drawing which shows an example of noise correction. It is explanatory drawing which shows the case where the correction value used for noise correction is calculated | required when the pixel whose signal level is higher than a 2nd threshold value arises. It is a flowchart which shows the flow of a process of the digital camera at the time of imaging | photography. It is explanatory drawing which shows the case where the correction value used for noise correction is calculated | required except the row | line | column in which the pixel whose signal level is higher than a 2nd threshold value is included.

  FIG. 1 shows a configuration of a digital camera using the present invention. As is well known, the digital camera 10 photoelectrically converts the subject light captured by the imaging optical system 15 by the imaging device 16 and acquires image data from the electrical signal (pixel signal) after the photoelectric conversion. The imaging optical system 15 includes a lens group including a zoom lens, a focus lens, and the like that are not shown. These zoom lens and focus lens are moved in the direction of the optical axis L by a lens driving mechanism (not shown). For example, a CMOS image sensor is used as the image sensor 16. The image sensor 16 is driven by a driver 17.

  The AFE (Analog Front End) circuit 21 includes an AGC circuit and a CDS circuit (not shown). The AFE circuit 21 receives an image signal acquired by the imaging process. The imaging process includes a first discharge process and a second discharge process in addition to an exposure time according to a preset shutter speed and an exposure process in which each pixel of the image sensor 16 receives subject light.

  The first discharge process is executed in a state where it is shielded by a shutter (not shown). Since unnecessary charges are accumulated in each pixel of the image sensor 16 even in a light-shielded state, fixed pattern noise (FPN) is generated by discharging unnecessary charges accumulated in each pixel. It is possible to acquire an image (hereinafter referred to as FPN image) indicating In the first discharge process, the charge accumulated in the capacitor 64 (see FIG. 2) provided in the image sensor 16 in a state where the transfer transistor 65 (see FIG. 2) of each pixel provided in the image sensor 16 is off. And a process of discharging charges accumulated in the photodiode 63 (see FIG. 2) and the capacitor 64 while the transfer transistor 65 provided in the image sensor 16 is on. By performing the first discharge process, it is possible to acquire an FPN image.

  Further, the second discharge process is executed in a state where it is shielded from light by the shutter after the exposure process. Similarly to the first discharge process, this second discharge process is also stored in the capacitor 64 (see FIG. 2) provided in the image sensor 16 with the transfer transistor 65 of each pixel provided in the image sensor 16 turned off. And a process of discharging charges accumulated in the photodiode 63 and the capacitor 64 while the transfer transistor 65 provided in the image sensor 16 is on. That is, two image signals are output from the image sensor 16 in each of the first discharge process and the second discharge process. By performing the second discharge process, the main image can be acquired.

  The AFE circuit 21 performs a correlated double sampling process on the two image signals obtained by the first discharge process to generate an image signal indicating an FPN image (hereinafter referred to as an FPN image signal). Similarly, a correlated double sampling process is performed on the two image signals obtained by the second discharge process to generate an image signal indicating the main image (hereinafter, the main image signal). Then, the generated FPN pixel signal and the main image signal are subjected to analog processing such as gain control and noise removal, and then the analog signal is converted into a digital signal.

  A DFE (Digital Front End) circuit 22 performs defect correction processing and noise correction processing on the input FPN image signal and main image signal. The DFE circuit 22 has functions of a defect correction unit 23 and a noise correction unit 24.

  The defect correction unit 23 identifies a pixel that becomes a white point (hereinafter, a white point pixel) generated in the FPN image or the main image as a defective pixel, and corrects the defective pixel. Note that the white point pixel indicates a case where the pixel level is relatively high even when light is not incident. The defect correction unit 23 determines whether the signal value of each pixel (hereinafter referred to as pixel value) in the input FPN image signal or the main image signal exceeds a preset defect detection threshold value. In this determination, a pixel having a pixel value exceeding the defect detection threshold is determined to be a white point pixel, and the white point pixel is specified as a defective pixel.

  When the defective pixel is specified, the defect correcting unit 23 replaces the pixel value of the defective pixel with one of the pixel values of the same color arranged at the peripheral edge of the defective pixel, or the peripheral edge of the defective pixel. Interpolation processing using the pixel values of the arranged pixels of the same color is performed to perform defect correction on the defective pixels.

  The noise correction unit 24 performs noise correction on the main image based on the FPN image. Specifically, a correction value for each pixel column is obtained from each pixel of the FPN image, and the pixel value of each pixel of the main image is corrected for each pixel column using the obtained correction value. A method for obtaining a correction value for each pixel column will be described later. The image signal subjected to the noise correction is recorded in the buffer memory 30 as image data. Reference numeral 25 denotes a timing generator (TG), which is provided for controlling the drive timing of the AFE circuit 21, the DFE circuit 22, and the driver 17.

  The image processing circuit 35 performs image processing such as white balance processing, color interpolation processing, contour compensation processing, and gamma processing on the image data stored in the buffer memory 30. After these processes, the image processing circuit 35 performs a compression encoding process using a preset compression rate. By executing this compression encoding process, for example, JPEG compressed image data is generated. Further, the image processing circuit 35 generates thumbnail image data by performing resolution changing processing.

  A storage medium 37 such as a memory card, an optical disk, or a magnetic disk is detachable from the media slot 36. The storage medium 37 can record image files such as still images and moving images. For example, if the image file is based on a still image, in addition to the compressed image data and thumbnail image data generated by the image processing circuit 35, there is one piece of information indicating the shooting conditions at the time of shooting and information of the digital camera 10. The Exif format image files compiled in the above are written in the storage medium 37.

  The LCD 38 is one form of a display device, and displays a through image and an image obtained at the time of shooting. In addition, the LCD 38 displays an image for setting when the digital camera 10 is set. Reference numeral 39 denotes a display control circuit that performs drive control of the LCD 38.

  The CPU 41 is electrically connected to the buffer memory 30, the image processing circuit 35, the media slot 36, the display control circuit 39, the built-in memory 43, and the like via the bus 42. The CPU 41 is connected to a release button 44, a setting operation unit 45, and the like. The CPU 41 controls each unit of the digital camera 10 based on an operation request on these operation members and a control program stored in the built-in memory 43. To do.

FIG. 2 is a circuit diagram illustrating the configuration of the pixels of the image sensor 16. As described above, a CMOS image sensor is used as the image sensor 16. The imaging device 16 includes the above-described plurality of pixels A _ij , a vertical shift register 51, a sample hold (SH) circuit 52 _{j, a} horizontal shift register 53, and an output amplifier 54. Each of the plurality of pixels A _ij is electrically connected by a plurality of signal lines 55 _i , 56 _i , 57 _i , 58 _i extending in the horizontal direction from the vertical shift register 51. Of the plurality of signal lines, the signal line 55 _i is a transfer signal line, the signal line 56 _i is a reset signal line, the signal line 57 _i is a selection signal line, and the signal line 58 _i is a power supply line. It is a signal line. Of the plurality of pixels A _ij, the pixel signals output from the pixels arranged in the same column direction (pixel A ₁₁ and pixel A ₂₁ in FIG. 2) are _supplied to the column amplifier 62 _j by the vertical signal line 61 _j. To the SH circuit 52 _j .

Like the conventional CMOS image sensor, the plurality of pixels A _ij include a photodiode 63 that generates charges by photoelectric conversion, a capacitor 64 that functions as a floating diffusion (FD) that stores charges generated by the photodiode 63, a photo A transfer transistor 65 that transfers the electric charge accumulated in the diode 63 to the capacitor 64, an amplification transistor 66 whose gate is connected to the capacitor 64 and converts the potential fluctuation of the capacitor 64 into an electric signal, and a pixel for reading the signal are selected in units of rows. And a reset transistor 68 for resetting the capacitor potential to the power supply potential (Vdd). The selection transistor 67 is connected to the vertical signal line 61 _j wired for each pixel column.

The SH circuit 52 _j includes selection transistors 71 and 72 for selecting pixels from which signals are read out in units of columns, and capacitors 73 and 74 for storing signals that are swept out when the gates of the selection transistors 71 and 72 are opened. Consists of Note that when the gate of the transfer transistor 65 described above is cut off, the selection transistor 71 and the capacitor 73 are used when the charge stored in the capacitor 64 is swept away. It is used when sweeping out the electric charge accumulated in the capacitor 64 when the gate of 65 is opened. The SH circuit 52 _j is connected to the horizontal signal line 75, and the pixel signal output from the SH circuit 52 _j is the pixel signal held in the SH circuit 52 _j through the horizontal signal line 75 and the output amplifier 54. Is output via. A collection of pixel signals output through the output amplifier 54 is an image signal (image data).

Next, the image sensor 16 described above is driven by the following operation. The vertical shift register 51 outputs a reset signal to the target pixel _Aij . When this reset signal is output, the reset transistor 68 of the target pixel A _ij is turned on. As a result, a voltage based on the electric charge accumulated in the capacitor 64 becomes the gate voltage of the amplification transistor 66. In this state, when a selection signal is output, the selection transistor 67 is turned on. Since a predetermined amount of voltage is applied to the power supply voltage line 58 _i , the gate voltage of the amplification transistor 66 is output to the vertical signal line 61 _j in response to the selection transistor 67 being turned on. The gate voltage output to the vertical signal line 61 _j is input to the SH circuit 52 _j via the column amplifier 62 _j . Note that the horizontal shift register 53 outputs an ON signal to the selection transistor 71. As a result, the selection transistor 71 is turned on, and the voltage amplified by the column amplifier 62 _j is accumulated in the capacitor 73 as electric charge.

Thereafter, the vertical shift register 51 outputs a transfer signal and a reset signal. In response to this, the transfer transistor 65 and the reset transistor 68 are turned on. As a result, the voltage based on the charges accumulated in the photodiode 63 and the capacitor 64 becomes the gate voltage of the amplification transistor 66. When a selection signal is output in this state, the selection transistor 67 is turned on. As a result, the gate voltage of the amplification transistor 66 is output to the vertical signal line 61 _j . The voltage output to the vertical signal line 61 _j is input to the SH circuit 52 _j through the column amplifier 62 _j . Note that the horizontal shift register 53 outputs an ON signal to the selection transistor 72. As a result, the selection transistor 72 is turned on, and the voltage amplified by the column amplifier 62 _j is accumulated in the capacitor 74 as electric charge. The charges stored in the capacitors 73 and 74 of the SH circuit 52 _j are simultaneously output to the horizontal signal line 75 and output through the output amplifier 54.

As shown in FIG. 3, the pixels of the image sensor 16 are arranged in a Bayer array composed of any one of R, G, and B colors. In FIG. 3, the R color pixel is indicated as R _n , the G color pixel is indicated as Gr _n , Gb _n , and the B color pixel is indicated as B _n . The pixel signal of each pixel in the image sensor 16 is read out for each pixel in the same row when the horizontal direction in FIG. 3 is a row and the vertical direction in FIG. 3 is a column. That is, pixel signals are read out in the order of the pixels R ₀₀ , Gr ₀₀ , R ₀₁ , Gr ₀₁ . Thereafter, pixel signals are read out in the order of the pixels Gb ₀₀ , B ₀₀ , Gb ₀₁ , B ₀₁ ... Arranged in the second row.

Next, a method for calculating a correction value used for noise correction will be described. As described above, the correction value used for noise correction is calculated based on the FPN image signal. As the image signal output from the image sensor 16, the pixel signal of each pixel output to the vertical signal line 61 _j is output via the horizontal signal line 75. In addition, each pixel of the image sensor 16 has pixels of any one of R color, G color (specifically, Gr color and Gb color), and B color arranged in a Bayer array. Considering these reasons, the correction value used for noise correction is a pixel having the same color component among the pixel signals output from the plurality of pixels and the pixel signals output through the same vertical signal line 61 _j. It is desirable to obtain using the pixel signal. As shown in FIG. 4, if the pixel is an R color pixel, the correction value of the R color pixel in the same pixel column is obtained by averaging the pixel values in the pixels R ₀₀ , R ₀₇ , R ₁₄ , and R ₂₁ . It is done. Similarly, the correction values of the pixels for the Gr color, Gb color, and B color are obtained in the same manner. Note that, as the correction value for each pixel column, the pixel values of all the pixels included in the same pixel column may be used, or some pixels may be used.

  In this way, correction values for the same pixel column are obtained during noise correction. However, in the pixel signal of each pixel output from the image sensor 16, the defect correction unit 23 described above does not determine a white point in addition to a pixel having a signal level determined to be a white point, but other pixels Pixels having a signal level that is clearly higher than the signal level are also included. When the pixel signal output from such a pixel is used, the correction value of the pixel column including the pixel is higher than the other correction values. For this reason, when such a correction value is used, noise correction on the main image is not performed with high accuracy, and vertical stripes are generated in the main image after noise correction.

  Thereby, when calculating | requiring the correction value for every pixel row | line using an FPN image, it is determined whether the pixel value of each pixel exceeds the preset threshold value for noise correction. Note that the noise correction threshold value is less than the above-described defect detection threshold value. When the pixel value of each pixel is equal to or less than the noise correction threshold value in this determination, a correction value for each pixel column is obtained by the method described above. On the other hand, if it is determined in the above-described determination that the pixel value exceeds the noise correction threshold, the noise correction is performed among the pixel columns including the pixel having the pixel value exceeding the noise correction threshold. The correction value for each pixel column is obtained using the pixel values of the remaining pixels, excluding the pixels that have pixel values that exceed the threshold for use.

As shown in FIG. 5, when it is determined that the pixel value of the pixel Gr ₀₇ exceeds the noise correction threshold value, the correction value is calculated using the pixel values of the pixels Gr ₀₀ , Gr ₁₄ , and Gr _21. calculate. Using the correction value for each pixel column obtained in this way, noise correction for the main image is executed for each pixel column.

  Next, the flow of processing in the digital camera 10 at the time of shooting will be described based on the flowchart of FIG. This flowchart is executed when the operation of the release button 44 is not executed, which is a so-called standby state.

  Step S101 is processing for determining whether or not the release button 44 has been operated. When the release button 44 is operated, the operation signal is input to the CPU 41. The CPU 41 determines that the release button 44 has been operated when an operation signal for the release button 44 is input. In this case, the determination process in step S101 is Yes, and the process proceeds to step S102. On the other hand, when the operation signal of the release button 44 is not input to the CPU 41, the CPU 41 determines that the release button 44 is not operated. In this case, the determination process in step S101 is No. In this case, the shooting standby state is maintained until the determination process in step S101 is Yes. When the determination process in step S101 is executed, the AE process and the AF process (not shown) are executed, and then the process in step S102 is executed. In addition, the process of step S102 and step S103 becomes an imaging process.

  Step S102 is processing to acquire an FPN image. When step S102 is executed, first, a first discharge process is executed. As described above, the first discharge process is executed in a state where the image sensor 16 is shielded from light by a shutter (not shown). The first discharge process is performed after the process of discharging the charge accumulated in the capacitor 64 in a state where the transfer transistor 65 of each pixel provided in the image sensor 16 is off, and then the transfer provided in the image sensor 16. A process of discharging the charge accumulated in the photodiode 63 and the capacitor 64 is performed in a state where the transistor 65 is on. By performing these processes, two types of image signals are input to the AFE circuit 21. The AFE circuit 21 performs a correlated double sampling process using two pixel signals corresponding to each pixel. After this correlated double sampling processing is executed, analog processing such as gain control and noise removal is performed. Finally, the analog pixel signal is converted into a digital pixel signal. A collection of digital pixel signals subjected to this processing is stored in the buffer memory 30 as an FPN image signal.

  Step S103 is processing for acquiring the main image. The CPU 41 opens the shutter based on the shutter speed determined by the AE process or set by the setting operation unit 45. Thereby, the exposure process is executed, and the subject light captured via the imaging optical system 15 is irradiated to each pixel of the imaging element 16 for a preset exposure time. In response to this, photoelectric conversion is performed by the photodiode 63 that constitutes each pixel of the image sensor 16, and signal charges are accumulated. After this exposure process is executed, a second discharge process is executed. The second discharge process is performed on the image sensor 16 after the process of discharging the charge accumulated in the capacitor 64 with the transfer transistor 65 of each pixel provided on the image sensor 16 turned off is executed. In the state where the transfer transistor 65 is on, a process of discharging the charges accumulated in the photodiode 63 and the capacitor 64 is executed. By performing these processes, two types of image signals are input to the AFE circuit 21. The AFE circuit 21 performs a correlated double sampling process using two pixel signals corresponding to each pixel. After this correlated double sampling processing is executed, analog processing such as gain control and noise removal is performed. Finally, the analog pixel signal is converted into a digital pixel signal. A collection of digital pixel signals subjected to this processing is stored in the buffer memory 30 as a main image signal.

  Step S104 is a process for performing defect correction. The DFE circuit 22 reads the FPN image signal recorded in the buffer memory 30 and determines whether or not the signal value (pixel value) in the read FPN image signal exceeds a defect detection threshold. In this determination, if there is a pixel value that exceeds the defect detection threshold, the pixel having that pixel value is identified as a defective pixel. Then, the pixel value of any pixel of the same color located at the periphery of the defective pixel is replaced with the pixel value of the pixel determined to be a defective pixel, or the pixel of the same color located at the periphery of the defective pixel The pixel value of the pixel determined to be a defective pixel is replaced by performing an interpolation process using the pixel value. The same processing is executed for the main image signal.

  Step S105 is processing for calculating a correction value for each pixel column. The DFE circuit 22 reads out the defect-corrected FPN image signal and determines whether or not the pixel value of each pixel in the FPN image signal exceeds a noise correction threshold value. When the pixel value exceeds the noise correction threshold, the address of the pixel having the pixel value is stored. After the comparison with the threshold value for noise correction is performed for all pixel values, the DFE circuit 22 calculates a correction value for each pixel column used for noise correction.

As shown in FIG. 4, when the pixels included in the same pixel column are the pixels R ₀₀ , R ₀₇ , R ₁₄ , R _{21 and} the pixel values of all these pixels are equal to or less than the noise correction threshold value, the DFE The circuit 22 calculates the correction value in the pixel row by averaging the pixel values of these pixels. On the other hand, as shown in FIG. 5, the pixels included in the same pixel column are pixels Gr ₀₀ , Gr ₀₇ , Gr ₁₄ , Gr ₂₁ , and among these pixels, the pixel value of the pixel Gr ₀₇ is a threshold value for noise correction. Is exceeded, the pixel values of the remaining pixels excluding the pixel Gr ₀₇ (in this case, the pixel Gr ₀₀ , the pixel Gr ₁₄ , and the pixel Gr ₂₁ ) are averaged to obtain a correction value in the pixel column. calculate. In this way, pixel values for each pixel column are obtained.

  Step S106 is a noise correction process for the main image. The DFE circuit 22 performs noise correction on the main image signal stored in the buffer memory 30 using the correction value for each pixel column. This noise correction is a process of subtracting the pixel value of a pixel included in a predetermined pixel column by the correction value of the corresponding pixel column. As a result, a main image signal from which fixed pattern noise has been removed is generated.

  Step S107 is image processing. Noise correction is performed in step S106. The image processing circuit 35 reads the main image signal subjected to noise correction. Then, image processing such as white balance processing, color interpolation processing, contour compensation processing, and gamma processing is performed on the read main image signal. After these processes, the image processing circuit 35 performs a compression encoding process using a preset compression rate. By executing this compression encoding process, for example, JPEG compressed image data is generated. Further, the image processing circuit 35 generates thumbnail image data by performing resolution changing processing.

  Step S108 is processing to record an image. In addition to the compressed image data and thumbnail image data generated in step S <b> 107, the CPU 41 collects information indicating shooting conditions and information such as a camera model as an image file, and writes the image file to the storage medium 37. When this process ends, one shooting is completed.

  When shooting is performed using the digital camera 10, an FPN image made up of unnecessary charges accumulated in each pixel without receiving the subject light is acquired before acquiring the main image obtained by receiving the subject light. Have acquired. When calculating a correction value in noise correction for each pixel column from this FPN image, a pixel having a pixel value having a high pixel value is excluded from a target to be used when calculating the correction value, and the correction value is set using the remaining pixels. Calculated. According to this, since it is not necessary to use a pixel value whose pixel value (signal level) is clearly higher than that of other pixels, it is possible to obtain a correction value with high accuracy. This suppresses the occurrence of vertical streaks in the main image from which the fixed pattern noise has been removed.

  Further, when acquiring the FPN image signal, the difference between the pixel signal obtained when the transfer transistor 65 is turned off and the pixel signal obtained when the transfer transistor 65 is turned on is calculated. Since the signal charge accumulated by the diode 63 can be considered as fixed pattern noise, more accurate noise correction can be performed on the main image.

In the present embodiment, the correction value for each pixel column is calculated after excluding the pixel value that exceeds the threshold value for noise correction. However, the present invention is not limited to this. For example, if any of the pixels arranged in the same row has a pixel value that exceeds the threshold value for noise correction, the pixel value of the pixel arranged in the same row sets the threshold value for noise correction. There is a high possibility of exceeding the pixel value. As a result, when a pixel having a pixel value that exceeds the noise correction threshold is specified, a correction value for each pixel column may be obtained after removing all pixels in the same row including the pixel. Is possible. That is, as shown in FIG. 7, when the pixel R ₀₈ has a pixel value exceeding the noise correction threshold, the pixel in the same row including the pixel R ₀₈ (in the figure, the pixel R ₀₇ is The pixel Gr ₀₇ , the pixel R ₀₈ , the pixel Gr ₀₈ , the pixel R ₀₉ , and the pixel Gr ₀₉ ) are excluded from the target when calculating the correction values to be used for noise correction, and the pixels in other rows are used for each pixel column. The correction value is calculated.

  In the present embodiment, only pixels having a pixel value higher than the noise correction threshold are excluded when calculating the correction value, but are arranged at the periphery of the pixel having a pixel value higher than the noise correction threshold. Replaced with the pixel value of the same color pixel or higher than the noise correction threshold value by interpolation using pixels of the same color arranged at the periphery of the pixel having a pixel value higher than the noise correction threshold value. It is also possible to use the correction value for each pixel column after replacing the pixel value of the pixel to be the pixel value.

  In the present embodiment, the noise correction process is performed after the defect correction process. However, the present invention is not limited to this, and the defect correction process can be performed after the noise correction process is performed.

  In this embodiment, an electronic camera is taken as an example, but the present invention is not limited to this. For example, the present invention can be used for a portable terminal such as a portable phone or a portable game machine having a camera function. is there.

  In addition, a program for causing a computer to execute the functions of the defect correction unit and the noise correction unit illustrated in FIG. 1 may be used. In this case, at the time of photographing with the electronic camera, it is also possible to record the FPN image and the main image in association with each other and execute defect correction processing and noise correction processing. These functions are set as one of the functions of the image processing program, and these processes can be executed before the image processing is executed.

DESCRIPTION OF SYMBOLS 10 ... Digital camera, 16 ... Imaging device, 22 ... DFE circuit, 23 ... Defect correction part, 24 ... Noise correction part, 30 ... Buffer memory, 35 ... Image processing circuit, 51 ... Vertical shift register, 52 ... SH circuit, 53 ... Horizontal shift register, 62 ... Column amplifier, 63 ... Photodiode, 64 ... Capacitor, 65 ... Transfer transistor, 67 ... Select transistor, 68 ... Reset transistor

Claims (5)

An image sensor in which a plurality of pixels that photoelectrically convert incident light are arranged in a matrix;
A pixel signal based on the charge accumulated in each of the plurality of pixels arranged in the image pickup device and stored in a state where the image pickup device is shielded from light is output as a first pixel signal, and arranged in the image pickup device. A control unit that outputs, as a second pixel signal, a pixel signal based on charges obtained when photoelectrically converting the incident light received by each of the plurality of pixels.
When the first pixel signal output from the plurality of pixels includes a first pixel signal whose signal value exceeds a second threshold value greater than a preset first threshold value , the second threshold value is set. After the excess first pixel signal is corrected based on the first pixel signal output from the peripheral pixel arranged at the peripheral edge of the pixel that output the first pixel signal, at least a first threshold value set in advance is set. A correction value for each pixel column is calculated using the remaining first pixel signals included in each pixel column excluding the excess first pixel signal, and the pixel column is calculated using the calculated correction value for each pixel column. A correction unit that corrects the second pixel signal for each pixel column every time;
An electronic camera characterized by comprising: The electronic camera according to claim 1,
The electronic camera according to claim 1 , wherein the correction unit replaces a first pixel signal exceeding the second threshold with a first pixel signal output from any one of the peripheral pixels . The electronic camera according to claim 1,
The electronic camera according to claim 1, wherein the correction unit corrects the first pixel signal exceeding the second threshold by performing an interpolation process using the first pixel signal output from the peripheral pixel . The electronic camera according to any one of claims 1 to 3 ,
The plurality of pixels block a photoelectric conversion unit that photoelectrically converts the incident light, a charge storage unit that stores charges photoelectrically converted by the photoelectric conversion unit, and the photoelectric conversion unit and the charge storage unit, or A switching element to be connected,
The control unit is output with the pixel signal output in a state where the photoelectric conversion unit and the charge storage unit are blocked by the switching element, and the photoelectric conversion unit and the charge storage unit are connected. An electronic camera that outputs the first pixel signal and the second pixel signal by obtaining a difference from the pixel signal . The electronic camera according to any one of claims 1 to 4 ,
The imaging device further includes a transfer unit that transfers a pixel signal from each pixel of the plurality of pixels in a pixel column direction, and an amplifier that amplifies the pixel signal read by the transfer unit for each pixel column. ,
The electronic camera according to claim 1, wherein the correction unit calculates the correction value by averaging a signal value of a first pixel signal for each pixel column amplified by the amplifier .

Images