JP2011097162A - Fpn data acquisition device, solid-state imaging element, and electronic camera - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the following problem: it is conventionally difficult to obtain high-precision FPN correction data when there is shading. <P>SOLUTION: The FPN data acquisition device comprises: an imaging part where pixels each outputting a signal in response to luminous energy entering through an optical system are arranged in a two-dimensional matrix-like form; a retaining part for temporarily retaining signals read in a row sequence from the pixels in the same column; and a comparison part for comparing the signal retained in the retaining part with a signal in process of reading, updating the signal retained in the retaining part with the signal in process of reading when the comparison result is smaller than a threshold, and controlling the signal retained in the retaining part not to be updated when the comparison result is not smaller than the threshold. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an FPN data acquisition device, a solid-state imaging device, and an electronic camera.

  In recent years, electronic cameras using solid-state image sensors have been widely used. However, FPN (fixed pattern noise) generated by the circuit operation of the electronic camera becomes a problem. Therefore, in a general electronic camera, FPN data for reading a signal from the solid-state image sensor is acquired before shooting to create FPN correction data, and correction processing using the FPN correction data is performed at the time of image shooting. Yes. For example, correction data is collected in the OB region, and a threshold value is set in advance for the pixel signal at that time, and a method of replacing a threshold value that exceeds the threshold value is performed (Patent Literature). 1).

JP 2006-157263 A

  On the other hand, in a solid-state imaging device, even when light having a uniform luminance distribution is irradiated, shading occurs in which the peripheral portion of the image is darker than the central portion due to variations in circuit elements. However, the conventional technique has a problem that it is difficult to obtain highly accurate FPN correction data when there is shading.

  In view of the above problems, an object of the present invention is to provide an FPN data acquisition device, a solid-state imaging device, and an electronic camera that can obtain highly accurate FPN correction data even when there is shading.

  An FPN data acquisition apparatus according to the present invention temporarily captures an image pickup unit in which pixels that output signals according to the amount of light incident through an optical system are arranged in a two-dimensional matrix, and signals read out in the row order from the pixels in the same column. If the comparison result is less than a threshold value, the signal held in the holding unit is changed to the signal being read. And a comparison unit that controls to update the signal held in the holding unit when the comparison result is equal to or greater than a threshold value.

  In particular, a reference signal generation unit that generates a reference signal is further provided, and the holding unit holds the reference signal as an initial value before reading a signal from the first pixel.

  Further, the threshold value of the comparison unit is a range in which the signal held in the holding unit is a reference value, (the reference value + the threshold value) is an upper limit value, and (the reference value−the threshold value) is a lower limit value. Provided, the signal held in the holding unit is updated to the signal being read when the comparison result is within the range, and the signal held in the holding unit when the comparison result is out of the range Is controlled not to be updated.

  Furthermore, the comparison unit divides the threshold value into two, a first threshold value and a second threshold value that is relatively larger than the first threshold value, and a signal read from a pixel is held in the holding unit. The threshold value is set as a first threshold value, and the threshold value when the reference signal is held in the holding unit is set as a second threshold value.

  An FPN data solid-state imaging device according to the present invention is a solid-state imaging device equipped with the FPN data acquisition device, and a selection unit that selects a normal imaging mode and an FPN data acquisition mode, and a normal imaging mode is selected by the selection unit. An output unit that outputs a signal read from the imaging unit to the outside when the selection unit is selected, and an output unit that outputs the signal held by the holding unit to the outside when the FPN data creation mode is selected by the selection unit; It is provided.

  The electronic camera according to the present invention is an electronic camera equipped with the FPN data acquisition device, wherein a selection unit that selects a normal shooting mode and an FPN data acquisition mode, and the FPN data acquisition mode is selected by the selection unit. An FPN correction data generating unit that generates FPN correction data from a signal output from the FPN data acquisition device, and an FPN generated by the FPN correction data generating unit when the normal photographing mode is selected by the selection unit. An FPN correction processing unit that corrects a signal read from the imaging unit using correction data, and a storage unit that stores captured image data corrected by the FPN correction processing unit are provided.

  Alternatively, the electronic camera according to the present invention is an electronic camera in which the solid-state image sensor is mounted, and the FPN data acquisition mode is selected by the control unit that controls the selection unit of the solid-state image sensor and the selection unit. An FPN correction data generating unit that generates FPN correction data from a signal output from the FPN data acquisition device, and an FPN generated by the FPN correction data generating unit when the normal photographing mode is selected by the selection unit. An FPN correction processing unit that corrects a signal read from the imaging unit using correction data, and a storage unit that stores captured image data corrected by the FPN correction processing unit are provided.

  The FPN data acquisition apparatus, solid-state imaging device, and electronic camera according to the present invention can obtain highly accurate FPN correction data even when there is shading.

1 is a block diagram of a solid-state image sensor 101. FIG. It is an auxiliary | assistant figure for demonstrating FPN correction | amendment. It is an auxiliary | assistant figure for demonstrating the acquisition process of FPN data. It is a block diagram which shows the structure of the FPN data acquisition part 105 in the solid-state image sensor 101 which concerns on 1st Embodiment. It is a timing chart at the time of FPN data acquisition in a 1st embodiment. It is a block diagram which shows the structure of the FPN data acquisition part 105b in the solid-state image sensor 101b which concerns on 2nd Embodiment. It is a timing chart at the time of FPN data acquisition in the 2nd and 3rd embodiments. It is a block diagram which shows the structure of the FPN data acquisition part 105c in the solid-state image sensor 101c which concerns on 3rd Embodiment. It is a block diagram which shows the structure of FPN data acquisition part 105d in the solid-state image sensor 101d which concerns on 4th Embodiment. 1 is a block diagram illustrating an embodiment of an electronic camera 500. FIG.

  Hereinafter, embodiments of an FPN data acquisition device, a solid-state imaging device, and an electronic camera according to the present invention will be described in detail with reference to the drawings. Since the present invention is characterized by the configuration of the FPN data acquisition apparatus, the configuration of the FPN data acquisition apparatus will be described in detail later by taking some embodiments as examples.

  First, the configuration of the solid-state imaging device 101 including the FPN data acquisition device will be described with reference to FIG. 1, and then a configuration example of a plurality of FPN data acquisition devices will be described, and finally an electronic camera on which the solid-state imaging device 101 is mounted. explain.

  FIG. 1 is a block diagram illustrating a configuration example of the solid-state image sensor 101. In FIG. 1, a solid-state imaging device 101 includes an imaging unit 102 composed of pixels of N rows and M columns (N and M are natural numbers), and M vertical signal lines 103 that output signals read from pixels in the same column. The correlated double processing unit 104 for removing noise due to variations in the characteristics of each pixel, the FPN data acquisition unit 105 corresponding to the FPN data acquisition device according to the present invention, and the signal read for each column It comprises a horizontal output unit 106 that outputs to the outside of the solid-state imaging device 101, and a timing control unit 107 that controls the operation timing of each unit.

  Further, the FPN data acquisition unit 105 includes M column data acquisition units 151 provided for each column from column (1) to column (M), and a mode changeover switch 152. In addition, when designating the one arranged in a specific column among the M FPN data acquisition units 105, a column code ((1) to (M)) is added and described. For example, the FPN data acquisition unit 105 in the (m) column is represented as an FPN data acquisition unit 105 (m). Similarly, when specifying the M mode changeover switches 152 arranged in a specific column, the M mode changeover switches 152 are indicated by adding a column code. For example, the mode changeover switch 152 in the (m) th column is a mode changeover switch. It is written as 152 (m).

  In FIG. 1, the solid-state image sensor 101 switches FPN data acquisition mode / image shooting mode or outputs a timing signal for reading out an image signal at a specified resolution to each unit according to a control signal given from the outside. To do. For example, when FPN data is acquired on the electronic camera side on which the solid-state image sensor 101 is mounted, the FPN data acquisition mode is designated by a control signal for the solid-state image sensor 101. Receiving this, the timing control unit 107 switches the mode changeover switch 152 of the FPN data acquisition unit 105 to the a side. The signals read from the imaging unit 102 in units of rows are read to the horizontal output unit 106 via the column data acquisition unit 151 of each column, and then output to the outside of the solid-state imaging device 101 in the column order. When an image is taken with an electronic camera, an image taking mode is designated by a control signal for the solid-state image sensor 101. Receiving this, the timing control unit 107 switches the mode changeover switch 152 of the FPN data acquisition unit 105 to the b side so that signals read from the imaging unit 102 in units of rows do not pass through the column data acquisition unit 151 of each column. The data is directly read out to the horizontal output unit 106 and then output to the outside of the solid-state imaging device 101 in units of columns.

  Here, the solid-state imaging device 101 according to the present embodiment has a function of acquiring FPN data for correcting FPN (fixed pattern noise) including shading. As shown in FIG. 2, shading 251 and vertical stripe noises 252, 253, 254, and 255 that appear darker in the peripheral portion than the central portion appear in the image 201 before correction. In FIG. 2, the shading characteristics and the width of the vertical stripe noise are drawn with emphasis so as to be easily understood. Therefore, in the solid-state imaging device 101 according to the present embodiment, changes in the row direction such as the shading 251 and the vertical stripe noises 252, 253, 254, and 255 can be acquired as FPN data 256 and can be output to the outside of the solid-state imaging device 101. It has become. As a result, the electronic camera equipped with the solid-state imaging device 101 subtracts the FPN data 256 acquired before shooting from the image data read from the same solid-state imaging device 101 at the time of image shooting in units of columns, and performs shading 251 or An FPN correction process for removing the vertical stripe noises 252, 253, 254, and 255 is performed, and a beautiful image 202 after correction without shading or vertical stripe noise can be obtained.

  Next, a method for generating FPN data 256 will be described with reference to FIG. FIG. 3A shows a method for generating FPN data 256 according to the prior art. The FPN data 256 is obtained by averaging the N rows of data read for each row with the imaging unit 102 in a light-shielded state. However, in FIG. 3A, noise 260 may suddenly appear only in a specific row other than shading 251 and vertical stripe noises 252, 253, 254, and 255 appearing in the same manner in each row. Here, when the level of the noise 260 is small or the number of rows to be averaged is large, the influence on the FPN data 256 generated by the average effect can be almost ignored, but when the level is large or the number of rows to be averaged is small. In this case, the noise remains as noise 261 in the FPN data 256 even after averaging. As a result, the accuracy of the FPN data is deteriorated, and there is a problem that proper FPN correction cannot be performed.

  Therefore, in the FPN data acquisition unit 105 of the solid-state imaging device 101 according to the present embodiment, as illustrated in FIG. 3B, the noise 260 that appears in a specific row is not affected by the generated FPN data 256b. Control is performed so as not to output the data of the row including the noise 260. As a result, the noise 261 of the FPN data 256 due to the noise 260 is removed as shown in FIG. 3A, and the FPN data 256b that does not include the noise 261 can be obtained as shown in FIG. 3B. Since the averaging is performed for each column, only the data in the (3) th row is excluded during the averaging process for the column including the noise 260, and the averaging process for other columns not including the noise 260 is performed. (3) The data on the line is used.

  Hereinafter, the FPN data acquisition unit 105 of the solid-state imaging device 101 for obtaining the FPN data 256b will be described with some embodiments. Here, the FPN data acquisition unit 105 corresponds to the FPN data acquisition apparatus according to the present invention.

(First embodiment)
FIG. 4 is a diagram depicting mainly the configuration of the FPN data acquisition unit 105 of the solid-state imaging device 101 according to the first embodiment. In FIG. 4, the column data acquisition unit 151 (m) includes a holding unit 301, a comparison unit 302, and an AND (logical product gate) 351. For easy understanding, the AND 351 is drawn separately from the comparison unit 302 in FIG. 4, but the AND 351 may be included in the holding unit 301 as shown by the dotted line, or may be included in the comparison unit 302. Absent.

  In FIG. 4, a timing control unit 107 outputs a timing signal and the like necessary for the operation of each unit. For example, a plurality of readout timing signals are output to the imaging unit 102 and signals are read out from each pixel of the imaging unit 102 in units of rows. Further, a holding timing signal is output to the holding unit 301 via the AND 351, and the signal read from the imaging unit 102 is held in the holding unit 301. Note that the output of the comparison unit 302 is added to the AND 351, and a holding timing signal is given to the holding unit 301 when the output of the comparison unit 302 is logic “1”. Further, the timing control unit 107 outputs a comparison timing signal to the comparison unit 302, and the comparison unit 302 is held in the holding unit 301 and the signal of the row being read from the imaging unit 102 (for example, (n) rows). A signal in a row read from the imaging unit 102 in the past (for example, a row before (n−1) rows) is compared. For example, when the level difference between two signals is less than the threshold value, the logic “1” is output to the AND 351 and the holding timing signal is given to the holding unit 301. Control is performed so that the timing signal is not given to the holding unit 301. Note that the same value common to all columns is given from the timing control unit 107 as the threshold value.

  Further, the timing control unit 107 outputs a read mode signal to the mode switch 152 (m), and outputs the signal read from the imaging unit 102 as it is as a signal of the (m) column, or a column data acquisition unit Whether to output the signal held in the holding unit 301 of 151 (m) as the signal of the (m) column is switched. For example, when the FPN data acquisition mode of the solid-state imaging device 101 is designated from the outside, the timing control unit 107 outputs the logic “0” of the readout mode signal and switches the mode switch 152 (m) to the a side, When the image shooting mode is selected, the timing control unit 107 outputs the logic “1” of the readout mode signal and switches the mode switch 152 (m) to the b side.

  Here, in FIG. 4, a dotted circle (a) indicates a circuit example of the holding unit 301, and a dotted circle (b) indicates a circuit example of the comparison unit 302. In the circuit example of the holding unit 301 shown by the dotted circle (a), the signal read from the imaging unit 102 is input to the buffer 361, and the switch transistor 362 is turned on by the holding timing signal input via the AND 351. A signal read from the imaging unit 102 is held in the capacitor 363. Further, in the circuit example of the comparison unit 302 shown by the dotted circle (b), for example, (n) rows of signals (Vn) read from the imaging unit 102 and (n−1) held in the holding unit 301, for example. The signal (Vk) of the row before the row is input to the comparator 371, and it is compared whether or not there is a difference greater than a preset threshold value (Va). This means, for example, determining whether or not the absolute value of the difference between the signal Vn and the signal Vk is greater than or equal to the threshold value Va. When | Vn−Vk | ≧ Va, the output of the comparator 371 becomes logic “0”. When (NG), | Vn−Vk | <Va, the output of the comparator 371 becomes logic “1” (OK). That is, as depicted in the dotted circle (c), it is determined whether or not the signal Vn read from the imaging unit 102 is within the variation of the threshold value Va with reference to the output signal Vk of the holding unit 301. . As described above with reference to FIG. 2, the same threshold value Va may be given to the column data acquisition units 151 of all the columns even when there is shading with different signal levels depending on the columns even in the same row. This means that it is not necessary to prepare a different threshold value for each column. Furthermore, the threshold value can be set to a value smaller than the fluctuation range of shading, and the accuracy can be improved.

  Next, the operation when the FPN data acquisition mode of the solid-state imaging device 101 is selected will be described in detail with reference to FIG. FIG. 5 is a timing chart when acquiring FPN data from the (1) th row to the (N) th row in the (m) th column of the circuit of FIG. 5A is a timing chart in the case where the comparison unit 302 has not detected noise above the threshold over all rows, and FIG. 5B is a row in which there is noise above the threshold in the comparison unit 302 (three rows). It is a timing chart when an eye) is detected.

  First, the timing chart of FIG.

  (Timing T0) The timing control unit 107 sets the read mode signal to logic “1” and switches the mode changeover switch 152 to the a side to set the FPN data read mode. At this time, the holding unit 301 holds some initial value. The output of the comparison unit 302 is assumed to be logic “1” as an initial value.

  (Timing T <b> 1) The timing control unit 107 outputs a read timing signal to the imaging unit 102, and reads the output of the imaging unit 102 (the signal in the (1) row) to the vertical signal line 103.

  (Timing T2) The timing control unit 107 outputs a holding timing signal. Here, since the output of the comparison unit 302 is logic “1”, the holding timing signal is given to the holding unit 301 via the AND 351, and the output of the holding unit 301 is updated from the initial value to the data on the (1) th row. The As shown in the circuit example of the holding unit 301, since the output impedance of the buffer 361 is low, the charge accumulated in the capacitor 363 is reset when the switch transistor 362 is turned on. Alternatively, a reset transistor may be provided at both ends of the capacitor 363 so that the reset signal is output from the timing control unit 107. Further, the holding timing signal is always output at a predetermined position while reading the signal of each row from the imaging unit 102.

  (Timing T <b> 3) The timing control unit 107 outputs a read timing signal to the imaging unit 102, and reads the output of the imaging unit 102 (the signal in the (2) row) to the vertical signal line 103.

  (Timing T4) The timing control unit 107 outputs the comparison timing signal as a logic “1” in a pulse form for a short period to the comparison unit 302, and updates the output of the comparison unit 302.

  The example of FIG. 5A shows a case where the level difference between the (2) th row signal being read from the imaging unit 102 and the (1) th row signal held in the holding unit 301 is less than the threshold value. The same logic “1” as the initial value is continuously output from the comparison unit 302. In FIG. 5A, the case where the level difference between the two signals compared by the comparison unit 302 is greater than or equal to the threshold is indicated as NG, and the case where the level difference is less than the threshold is indicated as OK. Further, as shown in the circuit example of the comparison unit 302, since the output impedance of the comparator 371 is low, the charge accumulated in the capacitor 373 is reset when the switch transistor 372 is turned on. Alternatively, a reset transistor may be provided at both ends of the capacitor 373 so that the reset signal is output from the timing control unit 107.

  (Timing T5) Similar to the timing T2, the timing control unit 107 outputs a holding timing signal. At this time, since the output of the comparison unit 302 is logic “1”, a holding timing signal is given to the holding unit 301 via the AND 351, and the output of the holding unit 301 is changed from the data in the (1) row to the (2) row. It is updated to the data.

  (Timing T <b> 6) Similarly to the timing T <b> 3, the output of the imaging unit 102 ((3) row signal) is read out to the vertical signal line 103.

  (Timing T7) Similar to the timing T4, the output of the comparison unit 302 is updated. The example of FIG. 5A shows a case where the level difference between the (3) th row signal being read from the imaging unit 102 and the (2) th row signal held in the holding unit 301 is less than the threshold value. The same logic “1” as in the previous row is continuously output from the comparison unit 302.

  (Timing T8) Similar to the timing T5, the timing control unit 107 outputs a holding timing signal. At this time, since the output of the comparison unit 302 is logic “1”, a holding timing signal is given to the holding unit 301 via the AND 351, and the output of the holding unit 301 is changed from the data in the (2) row to the (3) row. It is updated to the data.

  (Timing T <b> 9) Similarly to the timings T <b> 3 and T <b> 6, the output of the imaging unit 102 ((4) row signal) is read out to the vertical signal line 103.

  (Timing T10) Similar to the timings T4 and T7, the output of the comparison unit 302 is updated. The example of FIG. 5A shows a case where the level difference between the (4) th row signal being read from the imaging unit 102 and the (3) th row signal held in the holding unit 301 is less than the threshold value. The same logic “1” as in the previous row is continuously output from the comparison unit 302.

  (Timing T11) Similar to the timings T5 and T8, the timing control unit 107 outputs a holding timing signal. At this time, since the output of the comparison unit 302 is logic “1”, a holding timing signal is given to the holding unit 301 via the AND 351, and the output of the holding unit 301 is changed from the data in the (3) th row to the (4) th row. It is updated to the data.

  Thereafter, in the example of FIG. 5A, since all the comparison results of the comparison unit 302 up to the (N) th row are OK (the level difference between the two signals is less than the threshold value), the same operation as above is repeated, and (1) The signals of all rows from the row to the (N) th row are read out and output to the outside of the solid-state image sensor 101 as FPN data of the (m) th column.

  At timing T12, the timing controller 107 changes the read mode signal from logic “1” to logic “0” and switches the mode changeover switch 152 of the FPN data acquisition unit 105 to the b side. As a result, the solid-state image sensor 101 is switched from the FPN data acquisition mode to the image shooting mode. Therefore, when the release button of the electronic camera equipped with the solid-state image sensor 101 is pressed, the image from the imaging unit 102 of the solid-state image sensor 101 is displayed. The signal can be read out.

  Next, the timing chart of FIG. The timing chart of FIG. 5B is a timing chart when the comparison unit 302 detects a row (third row) having noise equal to or higher than the threshold. The operations other than the third row are exactly the same as those in FIG. is there. Here, only a different part from FIG. 5A is demonstrated. The operation from (timing T0) to (timing T6) until the (2) th row is read out is omitted.

  (Timing T7) Similar to timing T4, the timing control unit 107 outputs a comparison timing signal to the comparison unit 302 (the gate of the switch transistor 372), latches the output of the comparator 371 at that time in the capacitor 373, and then compares the comparison unit. The output of 302 is updated. However, in the example of FIG. 5B, the level difference between the (3) th row signal being read from the imaging unit 102 and the (2) th row signal held in the holding unit 301 is equal to or greater than a threshold (NG). Therefore, unlike the case of FIG. 5A, the comparison unit 302 outputs logic “0”. Therefore, the holding timing signal output from the timing control unit 107 at the timing T8 in FIG. 5A does not appear in the output of the AND 351, and therefore is not given to the holding unit 301. As a result, the output of the holding unit 301 is not updated from the data on the (2) line to the data on the (3) line, and continues to hold the data on the (2) line.

  As described above, since the above processing is performed for each column, if the comparison result is normal in columns other than the (m) th column (for example, the (2) th column), the (2) th column The column data acquisition unit 151 (2) of the column data acquisition unit 151 stores the data of the (3) th row.

  (Timing T <b> 9) As in FIG. 5A, the output of the imaging unit 102 ((4) row signal) is read out to the vertical signal line 103.

  (Timing T10) As in FIG. 5A, the output of the comparison unit 302 is updated.

  (Timing T11) The operation is the same as in FIG. 5A, but in the example of FIG. 5B, the signal in the (4) th row being read from the imaging unit 102 and held in the holding unit 301 (2 ) The signal in the row is compared. That is, it is not necessarily compared with the signal of the previous row, but is compared with the signal of the previous row held in the holding unit 301 (normal) and is NG (abnormal). The line is skipped.

  Here, in the example of FIG. 5B, the level difference between the (4) th row signal being read from the imaging unit 102 and the (2) th row signal held in the holding unit 301 is less than the threshold value. In this example, logic “1” is output from the comparison unit 302, and the signal in the (4) th row is held in the holding unit 301. For example, if it is equal to or greater than the threshold value, logic “0” is output, so the holding unit 301 continues to hold the data on the (2) th row. Subsequent processing is the same as that shown in FIG.

  In the above description, the column data acquisition unit 151 (m) in the (m) th column has been described. However, the column data acquisition unit 151 (1) from the (1) th column to the (M) th column receives the column data. The acquisition unit 151 (M) operates in the same manner. However, as described above, the above-described processing is performed for each column, so that the comparison result may be OK or NG even in the same row.

  In this way, the FPN data read for each row is output to the horizontal output unit 106 described with reference to FIG. 1 and then read to the outside of the solid-state imaging device 101 in the column order. As a result, as described with reference to FIG. 3B, when there is a large level of noise 260 when a signal in a certain row of a certain column is read, the past normal signal is output as the output signal of the column in which the noise 260 exists. Since the row signal is output, the FPN data 256b in which the noise 261 does not remain as shown in FIG. 3B can be acquired. As a result, FPN correction with high accuracy can be performed at the time of image capturing, and a high-quality image can be captured.

(Second Embodiment)
Next, an FPN data acquisition apparatus and a solid-state image sensor according to the second embodiment will be described. FIG. 6 is a block diagram illustrating a configuration of the column data acquisition unit 151b (m) of the (m) column of the FPN data acquisition unit 105b of the solid-state imaging device 101b according to the second embodiment. In FIG. 6, the column data acquisition unit 151 b (m) includes a holding unit 301, a comparison unit 302, an AND (logical product gate) 351, and an input changeover switch 401. In FIG. 6, the same reference numerals as those in FIG. 4 of the first embodiment denote the same components. In particular, the difference between FIG. 6 and FIG. 4 is that an input changeover switch 401 is provided in the column data acquisition unit 151b (m). The input selector switch 401 selects either the signal read from the imaging unit 102 or a preset reference signal as the input signal of the holding unit 301. Also, the timing control unit 107b in FIG. 6 is slightly different from the timing control unit 107 in FIG. 4 and outputs an input switching timing signal for controlling the input switching switch 401. For example, the timing control unit 107 outputs the logic “1” of the input switching timing signal to switch the input selector switch 401 to the a side, and outputs the logic “0” to switch the input selector switch 401 to the b side. Points other than the above are the same as in FIG. 4 of the first embodiment.

  Next, the operation when the FPN data acquisition mode is designated for the solid-state imaging device 101b will be described in detail with reference to FIG. FIG. 7A is a timing chart when the FPN data from the (1) line to the (N) line is acquired by the circuit of FIG. 6, and corresponds to FIG. 5B of the first embodiment. . Accordingly, FIG. 7A is a timing chart when the comparison unit 302 detects a row (third row) having noise equal to or higher than the threshold. In FIG. 7A, the same reference numerals as those in FIG. 5B indicate the same timing. Only operations different from those in FIG. 5B will be described below.

  (Timing T21) The timing control unit 107b outputs a holding timing signal at a timing T21 before the timing T1 at which the signal of the (1) -th row is read from the imaging unit 102. The timing controller 107b sets the initial value of the input switching timing signal to logic “0” before the timing T21, the input selector switch 401 is connected to the b side, and the reference signal is input to the holding unit 301. Yes. For this reason, when the holding timing signal is supplied to the holding unit 301 via the AND 351, the holding unit 301 holds the reference signal. Thereafter, the timing control unit 107b sets the input switching timing signal to the logic “1” at the timing T1, and thereafter holds the logic “1” until the reading of the (N) th row is completed. As a result, the input selector switch 401 is connected to the a side, and a signal read from the imaging unit 102 is input to the holding unit 301. That is, the input changeover switch 401 is switched to the reference signal side only before (1) reading of the row.

  (Timing T <b> 22) The timing control unit 107 b outputs a short pulse comparison timing signal to the comparison unit 302. As a result, the reference signal held in the holding unit 301 is compared with the (1) -th row signal read from the imaging unit 102. The example of FIG. 7A shows a case where the level difference between the signal in the (1) row being read from the imaging unit 102 and the reference signal held in the holding unit 301 is less than the threshold value, and logic “1”. Is output from the comparison unit 302.

  The operation after the next timing T2 is the same as that shown in FIG.

  Thus, as in the first embodiment, the FPN data read for each row is output to the horizontal output unit 106 described with reference to FIG. . As a result, as described with reference to FIG. 3B, when there is a large level of noise 260 when a signal in a certain row of a certain column is read, the past normal signal is output as the output signal of the column in which the noise 260 exists. Since the row signal is output, the FPN data 256b in which the noise 261 does not remain as shown in FIG. 3B can be acquired. As a result, FPN correction with high accuracy can be performed at the time of image capturing, and a high-quality image can be captured. In the first embodiment, (1) the signal in the row cannot be removed if there is an abnormality. In this case, it is determined that the signal in the (2) row and thereafter is abnormal. There was a fear. On the other hand, in the present embodiment, (1) since the signal in the first row is compared with the reference signal, the reference signal is set to a standard level predicted as a signal value by (1) ) If there is an abnormality in the signal in the row, it can be removed.

(Third embodiment)
Next, an FPN data acquisition apparatus and a solid-state image sensor according to a third embodiment will be described. FIG. 8 is a block diagram illustrating a configuration of the column data acquisition unit 151c (m) of the (m) column of the FPN data acquisition unit 105c of the solid-state imaging device 101c according to the third embodiment. In FIG. 8, the column data acquisition unit 151 c (m) includes a holding unit 301, a comparison unit 302, an input changeover switch 401, and an OR (logical sum gate) 402. In FIG. 8, the same reference numerals as those in FIG. 6 denote the same components. In particular, the difference between FIG. 8 and FIG. 6 is in the input switching timing signal of the input selector switch 401 of the column data acquisition unit 151c (m). In FIG. 6, the input switching timing signal output from the timing control unit 107 c is directly supplied to the input switching switch 401. However, in FIG. 8, the input switching timing signal output from the timing control unit 107 c is input via the OR 402. The changeover switch 401 is provided. Further, since the output of the comparison unit 302 is also given to the OR 402, when the comparison result of the comparison unit 302 is OK (in the case of logic “1”), the input switching timing signal is not given to the input changeover switch 401. The input selector switch 401 is on the a side. Conversely, when the comparison result is NG (logic “0”), the input switch timing signal is given to the input switch 401, and when the input switch timing signal is logic “0”, the input switch 401 is on the b side. . Also, the input switching timing signal output by the timing control unit 107c is always output at the same timing according to the readout signal from the imaging unit 102, as shown in FIG. 7B, unlike the timing control unit 107b. Also, the holding timing signal output from the timing control unit 107c is directly given to the holding unit 301, unlike the timing control unit 107b of FIG. The points other than the above are the same as in FIG.

  Next, the operation when the FPN data acquisition mode of the solid-state image sensor 101c is selected will be described in detail with reference to FIG. FIG. 7B is a timing chart when the FPN data from the (1) line to the (N) line is acquired by the circuit of FIG. 8, and corresponds to FIG. 7A of the second embodiment. . Accordingly, FIG. 7B is a timing chart when the comparison unit 302 detects a row (third row) having noise equal to or higher than the threshold. In FIG. 7B, the same reference numerals as those in FIG. 7A indicate the same timing. Hereinafter, FIG. 7B will be described.

  (Timing T21) Similar to FIG. 7A, the timing control unit 107c outputs a holding timing signal at a timing T21 before the timing T1 at which the signal in the (1) -th row is read from the imaging unit 102. Since the timing control unit 107c initializes the input switching timing signal to be “0” before the timing T21, the input switching switch 401 is connected to the b side, and the holding unit 301 receives the reference signal. Have been entered. When the holding timing signal is given to the holding unit 301 at timing T21, the holding unit 301 holds the reference signal. Thereafter, the timing control unit 107c sets the input switching timing signal to logic “1” at the timing T1, and thereafter outputs the input switching timing signal at the same cycle as the cycle for reading the signal of each row from the imaging unit 102.

  (Timing T <b> 22) As in FIG. 7A, the timing control unit 107 c outputs a short pulse-shaped comparison timing signal to the comparison unit 302. As a result, the reference signal held in the holding unit 301 is compared with the (1) -th row signal read from the imaging unit 102.

  From the next timing T2 to timing T6, since the output of the comparison unit 302 is logic "1", the operation is the same as in FIG. 7A, and the signal in the (2) th row is held in the holding unit 301 at the timing T7. Then, the signal in the (3) th row is read from the imaging unit 102.

  (Timing T7) Similar to timing T4, the timing control unit 107c outputs a comparison timing signal to the comparison unit 302 and updates the output of the comparison unit 302. However, in the example of FIG. 7B, the level difference between the (3) th row signal being read from the imaging unit 102 and the (2) th row signal held in the holding unit 301 is equal to or greater than a threshold (NG). Therefore, the comparison unit 302 outputs logic “0”. Therefore, during the period up to timing T9, the logic “0” of the input switching timing signal is given to the input switching switch 401 via the OR 402, and the input switching switch 401 is connected to the b side.

  (Timing T8) At timing T8 before timing T9, the timing control unit 107c outputs a holding timing signal. At this time, since the input selector switch 401 is connected to the b side, the holding unit 301 inputs a reference signal, and the output of the holding unit 301 is updated from the data in the (2) row to the reference value based on the reference signal. The

  (Timing T9) At this time, since the input switching timing signal changes from logic “0” to logic “1”, it is given to the input selector switch 401 via the OR 402, and the input selector switch 401 is connected to the a side. .

  (Timing T10) Similar to the timing T4, the timing control unit 107c outputs a comparison timing signal to the comparison unit 302 and updates the output of the comparison unit 302. In the example of FIG. 7B, the signal on the (4) th row being read from the imaging unit 102 is compared with the reference value held in the holding unit 301.

  Here, in the example of FIG. 7B, when the level difference between the signal in the (4) th row being read from the imaging unit 102 and the reference value held in the holding unit 301 is less than the threshold (OK). ), The logic “1” is the output of the comparison unit 302. Accordingly, the input selector switch 401 is connected to the a side, and the signal in the (4) th row being read from the imaging unit 102 is input to the holding unit 301.

  When the level difference between the signal in the (4) th row being read from the imaging unit 102 and the reference value held in the holding unit 301 is equal to or greater than the threshold value (NG), logic “0” is set to the comparison unit. Since the input switching timing signal is also logic “0”, the input switching switch 401 is connected to the b-side reference signal, and the holding unit 301 continues to hold the reference value.

  (Timing T11) Since the input changeover switch 401 is connected to the a side in the example of FIG. 7B, the signal in the (4) th row is held in the holding unit 301 by the holding timing signal of the timing control unit 107c. The

  Thereafter, the signal in the (N) -th row is read out, and the same operation is performed until timing T12 when the FPN data reading period ends.

  The reference signal may be given a different value for each column. For example, if the same reference signal is given to all the columns by setting the center value of the fluctuation range of the shading characteristics, noise with extremely different values can be obtained. The effect which removes can be acquired.

  In this way, the FPN data read for each row is output to the horizontal output unit 106 described with reference to FIG. 1 and then read to the outside of the solid-state imaging device 101 in the column order. As a result, as described with reference to FIG. 3B, when there is a large level of noise 260 when a signal in a certain row in a certain column is read, a reference signal is output as an output signal of the column in which the noise 260 exists. Therefore, as shown in FIG. 3B, the FPN data 256b in which the influence of the noise 261 does not remain can be acquired. As a result, FPN correction with high accuracy can be performed at the time of image capturing, and a high-quality image can be captured.

(Fourth embodiment)
Next, an FPN data acquisition apparatus and a solid-state image sensor according to a fourth embodiment will be described. FIG. 9 is a block diagram illustrating a configuration of the column data acquisition unit 151d (m) of the (m) column of the FPN data acquisition unit 105d of the solid-state imaging device 101d according to the fourth embodiment. In FIG. 9, the column data acquisition unit 151 d (m) includes a holding unit 301, a comparison unit 302, an AND (logical product gate) 351, an input changeover switch 401, and a threshold changeover switch 403. In FIG. 9, the same reference numerals as those in FIG. 6 of the second embodiment denote the same components. In particular, the difference between FIG. 9 and FIG. 6 is that a threshold value changeover switch 403 is provided in the column data acquisition unit 151d (m). Then, the threshold value switch 403 selects either the first threshold value (Va) or the second threshold value (Vb) as the threshold value given to the comparison unit 302. Further, the timing control unit 107d in FIG. 9 is slightly different from the timing control unit 107b in FIG. 6 and outputs a threshold value switching signal for controlling the threshold value switching switch 403. For example, the timing control unit 107b outputs the logic “1” of the threshold switching signal to switch the threshold switching switch 403 to the a side (first threshold), outputs the logic “0”, and sets the threshold switching switch 403 to the b side. Switch to (second threshold). Points other than the above are the same as in FIG. 6 of the second embodiment.

  The relationship between the first threshold value (Va) and the second threshold value (Vb) is such that Va <Vb, as depicted by the dotted line (a) in FIG. Therefore, the range in which the comparison unit 302 is OK (logic “1”) when the first threshold value (Va) is given is narrow, and the comparison unit 302 is OK (when the second threshold value (Vb) is given). The range of logic “1”) becomes wider. Then, the timing control unit 107d selects the second threshold value by setting the threshold value switching signal to logic “0” before the timing T3, and selects the first threshold value by setting the threshold value switching signal to logic “1” after the timing T3. Control.

  Accordingly, in the timing chart of FIG. 7A, the second threshold value is obtained at the timing T22 when the signal in the (1) row being read from the imaging unit 102 and the reference value held in the holding unit 301 are compared. Is selected, the signal on the (1) line being read from the imaging unit 102 is compared with the reference value held in the holding unit 301 using a wider threshold considering the variation range of the shading characteristics. be able to. On the other hand, since the first threshold value in a narrow range is selected at timing T4 after timing T3, the (2) row signal being read from the imaging unit 102 and the holding unit 301 hold the (1) row. The eye signal can be compared with high accuracy.

  In this way, as in the second embodiment, the FPN data read for each row is output to the horizontal output unit 106 described with reference to FIG. . As a result, as described with reference to FIG. 3B, when there is a large level of noise 260 when a signal in a certain row of a certain column is read, the past normal signal is output as the output signal of the column in which the noise 260 exists. Since the row signal is output, the FPN data 256d in which the noise 261 does not remain as shown in FIG. 3B can be acquired. As a result, FPN correction with high accuracy can be performed at the time of image capturing, and a high-quality image can be captured.

  In particular, in this embodiment, a wide threshold value is used when comparing with the reference signal, and a narrow threshold value is used when comparing signals in other rows, so that high-precision comparison can be performed and smaller noise is removed. It becomes possible.

  The FPN data acquisition apparatus and the solid-state imaging device according to the first to fourth embodiments have been described above, but an embodiment of an electronic camera equipped with these solid-state imaging devices will be described.

(Embodiment 1 of electronic camera)
FIG. 10 is a block diagram illustrating a configuration of the electronic camera 500. In FIG. 10, an electronic camera 500 includes a lens optical system 501, an imaging mechanism 502, a solid-state imaging device 503, an AFE (analog front end) 504, an image buffer 505, an image processing unit 506, and a control unit 507. , A memory card I / F (interface) 508, a display unit 509, an operation member 510, and a memory 511.

  In the electronic camera 500, the subject light incident from the lens optical system 501 is imaged on the light receiving surface of the solid-state imaging device 503 through the photographing mechanism 502 including a diaphragm and a shutter, and is converted into an electric signal according to the amount of light. The Here, the solid-state imaging device 503 corresponds to any one of the solid-state imaging devices 101, 101b, 101c, and 101d described in the first to fourth embodiments. The electric signal output from the solid-state imaging device 503 is subjected to noise removal and amplification by AFE, A / D converted, and taken into the image buffer 505. Thereafter, white balance processing, color interpolation processing, JPEG image compression processing, and the like are performed by the image processing unit 506 and the control unit 507, and finally stored as a captured image in the memory card 508a attached to the memory card I / F 508. Is done. Note that the captured image is displayed on the screen of the display unit 509 as necessary. The photographer uses the electronic camera 500 by operating a power button, a mode switching button, a release button, or the like of the operation member 510. The operation information of the operation member 510 is input to the control unit 507, and the control unit 507 controls the operation of each unit of the electronic camera 500 according to the operation information.

  The basic operation of the electronic camera 500 has been described above. Next, the FPN correction process that is a feature of the present embodiment will be described. The electronic camera 500 appropriately acquires FPN data as described with reference to FIG. 3 when the power is turned on or before shooting, and performs FPN correction processing using the acquired FPN data as described with reference to FIG. 2 when shooting an image. It has become. Therefore, in order to perform the FPN correction processing in the electronic camera 500 by the control unit 507, the control unit 507 is provided with an FPN data acquisition unit 551, an FPN correction data generation unit 552, and an FPN correction processing unit 553.

  The FPN data acquisition unit 551 designates the FPN data acquisition mode for the solid-state image sensor 503 with the imaging mechanism 502 in a light-shielding state. As described in the first to fourth embodiments, the FPN data read from the solid-state image sensor 503 is taken into the image buffer 505 via the AFE 504.

  As described with reference to FIG. 3B, the FPN correction data generation unit 552 calculates the average value for each row by adding the signals in the same column of the FPN data fetched into the image buffer 505 over all the rows, and stores the memory 511. Is stored as FPN correction data.

  The FPN correction processing unit 553 reads the FPN correction data stored in the memory 511 and subtracts the FPN correction data for each row of the captured image data at the time of image shooting, thereby correcting shading characteristics and pulse-like noise. Etc. are removed. The actual processing for subtracting the FPN data from the captured image is performed by the image processing unit 506 so that the FPN correction data is passed from the control unit 507 to the image processing unit 506 and the image processing unit 508 is instructed to perform FPN correction processing. It doesn't matter if you do.

  The FPN corrected image data is stored in the memory card 508a from the control unit 507 via the memory card I / F 508.

  In this manner, the electronic camera 500 uses any one of the solid-state imaging devices 101, 101b, 101c, and 101d described in the first to fourth embodiments as the solid-state imaging device 503. FIG. As described in FIG. 3B, when there is a large level of noise 260 in a certain column when a signal in a certain row is read, it is not output as a signal in a column having the noise 260, so the FPN in FIG. FPN data that is not affected by the noise 260 like the data 256d can be acquired. As a result, FPN correction with high accuracy can be performed at the time of image capturing, and a high-quality image can be captured.

(Embodiment 2 of electronic camera)
Next, a second embodiment of the electronic camera will be described. The electronic camera of this embodiment has the same configuration as the electronic camera 500 shown in FIG. However, the solid-state image sensor 503 is not one of the solid-state image sensors 101, 101b, 101c, and 101d described in the first to fourth embodiments, but a normal solid-state image sensor is used. This is different from the first embodiment of the camera. For this reason, in the electronic camera 500 according to the present embodiment, the operation of the FPN data acquisition unit 551 is also different and operates as follows.

  The FPN data acquisition unit 551 reads out image data from the solid-state imaging device 503 in units of rows with the imaging mechanism 502 in a light-shielding state. Then, as described with reference to FIG. 3B, signals in the same column from the (1) line to the (N) line are added to obtain an average value. The M average values for one row thus obtained are acquired as FPN data and stored in the memory 511 as FPN correction data. At this time, the FPN data acquisition unit 551 performs the following processing.

  (Step 1) Read M data in the (n) th row from the solid-state image sensor 503 and temporarily store it in the image buffer 505.

  (Step 2) The M data before the (n−1) th row temporarily stored in the image buffer 505 and the M data of the (n) th row read out in Step 1 for each column. Compare. This corresponds to, for example, the processing of the comparison unit 302 in FIG. 4. For each column, the absolute value of the difference between the data on the (n) th row and the data before the (n−1) th row is obtained, and the memory 511 is obtained in advance. This is a process of comparing each threshold value with a threshold value stored in the table and determining whether or not the threshold value is greater than or equal to the threshold value.

  (Step 3) If the comparison result in Step 2 is NG (when the comparison result is equal to or greater than the threshold), the data in the (n) th row of the column is replaced with the data before the (n-1) th row in the same column, and the image buffer 505 Remember again.

  (Step 4) The processing from Step 1 to Step 3 is repeated until the processing from the (1) line to the (N) line is completed. Note that when reading the data on the (1) line, an appropriate initial value or a past FPN data value is used as the reference value as data before the (n-1) line.

  In this manner, the FPN data acquisition unit 551 can acquire FPN data even when the solid-state imaging devices 101, 101b, 101c, and 101d described in the first to fourth embodiments are not used.

  As described with reference to FIG. 3B, the FPN correction data generation unit 552 adds the signals in the same column of the FPN data fetched into the image buffer 505 over all the rows to obtain an average value for each row, and In 511, it is stored as FPN correction data.

  As in the first embodiment of the electronic camera, the FPN correction processing unit 553 reads out the FPN correction data stored in the memory 511 and subtracts the FPN correction data for each row of the captured image data at the time of image shooting. , Shading characteristics correction and pulse noise etc. are removed.

  The FPN corrected image data is stored in the memory card 508a from the control unit 507 via the memory card I / F 508.

  As described above, the electronic camera 500 according to the second embodiment of the electronic camera is any of the solid-state imaging devices 101, 101b, 101c, and 101d described in the first to fourth embodiments as the solid-state imaging device 503. The same effect can be obtained without using it. As described with reference to FIG. 3B, when there is a large level of noise 260 in a certain column when a signal in a certain row is read, the column in which the noise 260 is present. Therefore, it is possible to obtain FPN data that does not remain affected by the noise 260, such as the FPN data 256d in FIG. As a result, FPN correction with high accuracy can be performed at the time of image capturing, and a high-quality image can be captured.

  Note that the processing from step 1 to step 4 requires a comparison process for each column, and there is a problem that it takes a processing time. For example, when FPN data is acquired for each shooting, high-speed processing is required, and it is necessary to mount an expensive LSI in the electronic camera. Further, since the noise is determined for each column after being taken into the image buffer 505, there is a possibility that the solid-state image pickup device itself does not necessarily acquire shading or noise FPN data. On the other hand, when the solid-state imaging devices 101, 101b, 101c, and 101d described in the first to fourth embodiments are used, since they are processed in an analog manner inside the solid-state imaging device, the obtained FPN is obtained. Data accuracy is increased, and the processing burden on the electronic camera can be reduced.

  The embodiments of the FPN data acquisition device, the solid-state imaging device, and the electronic camera according to the present invention have been described above, but can be implemented in various other forms without departing from the spirit or main features thereof. . Therefore, the above-described embodiment is merely an example in all respects and should not be interpreted in a limited manner. The present invention is defined by the claims, and the present invention is not limited to the text of the specification. Further, all modifications and changes belonging to the equivalent scope of the claims are within the scope of the present invention.

101, 101 b, 101 c, 101 d,... Solid imaging device; 102, imaging unit; 103, vertical signal line; 104, correlated double processing unit, 105, 105 b, 105 c, 105 d,. FPN data acquisition unit; 106 ... horizontal output unit; 107, 107b, 107c, 107d ... timing control unit; 151 ... column data acquisition unit; 152 ... mode switch; 302: Comparison unit; 351: AND (logical product gate); 361 ... Buffer; 362 ... Switch transistor; 363 ... Capacitor; 371 ... Comparator; 373 ... capacitor; 351 ... AND; 401 ... input changeover switch; 402 ... OR (logical sum gate); ... Threshold switch; 500... Electronic camera; 501... Lens optical system; 502 ... Imaging mechanism; 503 ... Solid-state image sensor; 504 ... AFE (analog front end); Image buffer; 506, image processing unit; 507, control unit; 508, memory card I / F (interface); 509, display unit; 510, operation member; Memory: 508a ... Memory card; 551 ... FPN data acquisition unit; 552 ... FPN correction data generation unit; 553 ... FPN correction processing unit

Claims (7)

An imaging unit in which pixels that output signals according to the amount of light incident through the optical system are arranged in a matrix;
A holding unit for temporarily holding signals read out in the row order from the pixels in the same column;
The signal held in the holding unit is compared with the signal being read. If the comparison result is less than the threshold value, the signal held in the holding unit is updated to the signal being read, and the comparison result is equal to or greater than the threshold value. In the case of the FPN data acquisition apparatus, the FPN data acquisition apparatus includes: a comparison unit that controls the signal held in the holding unit not to be updated. In the FPN data acquisition device according to claim 1,
A reference signal generator for generating a reference signal is further provided,
The FPN data acquisition apparatus, wherein the holding unit holds the reference signal as an initial value before reading a signal from a first pixel. In the FPN data acquisition device according to claim 1 or 2,
The threshold value of the comparison unit has a range in which a signal held in the holding unit is a reference value, (the reference value + the threshold value) is an upper limit value, and (the reference value−the threshold value) is a lower limit value, When the comparison result is within the range, the signal held in the holding unit is updated to the signal being read, and when the comparison result is outside the range, the signal held in the holding unit is updated. An FPN data acquisition device characterized in that control is performed so as not to occur. In the FPN data acquisition device according to claim 3,
The comparison unit divides the threshold value into two, a first threshold value and a second threshold value that is relatively larger than the first threshold value, and sets a threshold value when a signal read from a pixel is held in the holding unit. An FPN data acquisition apparatus, wherein a first threshold value is set as a second threshold value when the reference signal is held in the holding unit. A solid-state imaging device equipped with the FPN data acquisition device according to any one of claims 1 to 4,
A selection unit for selecting a normal shooting mode and an FPN data acquisition mode;
When the normal photographing mode is selected by the selection unit, a signal read from the imaging unit is output to the outside, and when the FPN data creation mode is selected by the selection unit, a signal held by the holding unit is output. A solid-state imaging device comprising: an output unit that outputs to the outside. An electronic camera equipped with the FPN data acquisition device according to any one of claims 1 to 4,
A selection unit for selecting a normal shooting mode and an FPN data acquisition mode;
An FPN correction data creation unit that creates FPN correction data from a signal output by the FPN data acquisition device when an FPN data acquisition mode is selected in the selection unit;
An FPN correction processing unit that corrects a signal read from the imaging unit using the FPN correction data created by the FPN correction data creation unit when the normal photographing mode is selected by the selection unit;
An electronic camera comprising: a storage unit that stores photographed image data corrected by the FPN correction processing unit. An electronic camera equipped with the solid-state imaging device according to claim 5,
A control unit for controlling the selection unit of the solid-state imaging device;
An FPN correction data creation unit that creates FPN correction data from a signal output by the FPN data acquisition device when an FPN data acquisition mode is selected in the selection unit;
An FPN correction processing unit that corrects a signal read from the imaging unit using the FPN correction data created by the FPN correction data creation unit when the normal photographing mode is selected by the selection unit;
An electronic camera comprising: a storage unit that stores photographed image data corrected by the FPN correction processing unit.

Images