JP2006210417A - Method of checking solid-state imaging device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a checking method capable of highly accurately evaluating a noise to be checked which is not caused by a defective pixel. <P>SOLUTION: An electric charge transfer device (3) is checked which is provided with a light receiver, a vertical transfer for transferring the signal charges transferred from the light receiver in a vertical direction, a horizontal transfer for transferring signal charges transferred from the vertical transfer in a vertical direction, an electric charge detector for converting the electric charges transferred from the horizontal transfer into voltages and detecting the voltages, and a power source for setting the reset voltage of the electric charge detector. A first image signal to be outputted from the electric charge detector when the power source (2) applies the first reset voltage to the electric charge detector is acquired (4A), a second image signal to be outputted from the electric charge detector when the power source (2) applies the second reset voltage to the electric charge detector is acquired (4B), and a difference voltage between the first and second image signals is compared with a reference signal (7). Thus, the presence or absence of a noise is checked in an image signal to be acquired from the electric charge detector. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a linear noise inspection method for a solid-state imaging device.

  Conventionally, low-cycle flaws and horizontal linear noise that appear due to defects in a CCD (Charge Coupled Device) are evaluated based on the flaw level and flaw deviation value of an image obtained in a dark state.

  FIG. 10 shows a configuration of a general CCD type solid-state imaging device 10. Pixels having a light receiving unit 11 that performs photoelectric conversion of incident light are arranged in a matrix, and a vertical transfer unit 12 that transfers signal charges photoelectrically converted by the light receiving unit 11 in a vertical direction, and is transferred from the vertical transfer unit 12. A horizontal transfer unit 13 for transferring the signal charge in the horizontal direction, and an output unit 14 for converting the signal charge transferred from the horizontal transfer unit 13 into a voltage or current signal and outputting it. A horizontal dummy transfer unit 17 is provided at an end of the horizontal transfer unit 13. The light receiving area composed of a plurality of pixels includes an effective pixel portion 15 and a light-shielded horizontal OB (Optical Black) portion 16.

  Although not shown, for example, the output unit 14 detects, for example, a charge detection unit that detects the charge transferred from the horizontal transfer unit 13 by converting it into a voltage, a power supply unit that sets a reset voltage of the charge detection unit, and a charge detection by a reset signal. A reset switch unit that controls connection between the unit and the power source unit. The charge transfer in the vertical and horizontal directions is performed by applying a drive signal to the transfer electrodes provided in the vertical transfer unit 12 and the horizontal transfer unit 13. The horizontal dummy transfer unit 17 is not always necessary.

  A conventional image noise discrimination method disclosed in Patent Document 1 will be described with reference to FIGS. FIG. 11 is a block diagram showing a conventional example, and FIG. 12 is an explanatory diagram showing a difference detection operation. In FIG. 11, the luminance conversion circuit 21 generates a luminance signal from the video signal, and the luminance signal from the luminance conversion circuit 21 is stored in the frame memories 22 to 26. The changeover switch 27 selects any one of the frame memories 22 to 26 and connects it to the difference detection circuit 28.

  The difference detection circuit 28 determines the luminance of the pixel at the coordinates (i, j) of the still image stored in the frame memory 24 and the luminance of the pixel group in a predetermined range of two lines before and after the pixel at the coordinates (i, j). And the difference between the luminance of the pixel group at the coordinates (i, j) of the still images in the frame memories 22 and 23 and the frame memories 25 and 26 and the brightness of the pixel group around the pixel is calculated. Compare the difference with the reference value. The determination circuit 29 determines that the pixel signal at the coordinates (i, j) is noise when all the differences are larger than the reference value as a result of the comparison by the difference detection circuit 28.

  Next, the operation of the above circuit will be described. The video signal is converted into a luminance signal by the luminance conversion circuit 21. For example, the video signal is subjected to color signal processing, and a luminance signal is synthesized from the obtained R, G, and B. The luminance signal of the (n-2) th frame from the luminance conversion circuit 21 shown in FIG. 12 is stored in the frame memory 22, and the luminance signal of the (n-1) th frame is stored in the frame memory 23 and the luminance of the nth frame. The signal is stored in the frame memory 24, the luminance signal of the (n + 1) th frame is stored in the frame memory 25, and the luminance signal of the (n + 2) th frame is stored in the frame memory 26.

  First, the changeover switch 27 is switched to the frame memory 24, and the difference detection circuit 28 calculates the difference between the luminance of the pixel at the coordinates (i, j) of the nth frame and the luminance of the pixel group in a predetermined range of two lines above and below. The calculated difference is compared with the reference value.

  Next, the changeover switch 27 is switched to the frame memories 22, 23, 25, and 26, respectively, and the difference detection circuit 28 causes the luminance of the pixel at the coordinates (i, j) of the nth frame and the (n-2) th frame. The luminance difference between the pixel at the coordinates (i, j) of the (n−1) th frame, the (n + 1) th frame, and the (n + 2) th frame, and the pixel group around the pixel is calculated and calculated. The difference and the reference value are compared.

If all the differences are larger than the reference value as a result of the comparison by the difference detection circuit 28, the determination circuit 29 determines the pixel at the coordinates (i, j) of the nth frame as noise. Thus, image noise that can be sensed only in a moving image can be detected from two pieces of information, that is, spatial difference information and temporal difference information.
JP-A-5-191834

  However, when the solid-state imaging device is evaluated as described above, a plurality of noises are mixed in the image signal output from the solid-state imaging device, and noise generated due to reset drain leakage or generated in the horizontal transfer unit. It has been difficult to separate and evaluate noise such as noise that does not originate from defects in the pixel portion.

  The image signal output from the solid-state imaging device includes components such as white scratches and roughness caused by the dark current from the photodiode. On the other hand, in the evaluation by the image comparison as in the conventional example, it is a comparison target. Basically, no noise is generated at the same address between the frame and the frame to be compared. Therefore, the detection result and the noise level of the actual image may not always match because of including an uncertain element due to a probabilistic problem.

  An object of the present invention is to solve such problems and to provide an inspection method capable of highly accurate evaluation of noise that is a detection target not caused by a defect in a pixel portion.

  The inspection method for a solid-state imaging device according to the present invention includes a light receiving unit in which pixels having photodiodes are two-dimensionally arranged to photoelectrically convert incident light into a signal charge, and a signal read from the light receiving unit by a vertical transfer pulse signal A vertical transfer unit that transfers charges in the vertical direction, a horizontal transfer unit that transfers the signal charges transferred from the vertical transfer unit in a horizontal direction by a horizontal transfer pulse signal, and a voltage transferred from the horizontal transfer unit This is a method for inspecting a charge transfer device, comprising: a charge detection unit that converts and detects the voltage and a power supply unit that sets a reset voltage of the charge detection unit.

  In order to solve the above-described problem, the inspection method for a solid-state imaging device according to the present invention provides a first image output from the charge detection unit when a first reset voltage is applied to the charge detection unit by the power supply unit. A second image signal output from the charge detection unit when a second reset voltage is applied to the charge detection unit by the power supply unit, and the first and second images are acquired. By comparing the difference voltage of the signal with a reference voltage, the presence or absence of noise in the image signal obtained from the charge detection unit is inspected.

  According to this configuration, by changing the reset power supply voltage, noise that is a detection target that is not caused by a defect in the pixel portion, such as noise that occurs due to reset drain leakage or noise that occurs in the horizontal transfer portion, changes. . Therefore, the difference evaluation is performed between a plurality of images obtained under the respective conditions of the high power supply voltage state for enhancing the detection target noise and the low power supply voltage state that does not generate the detection target noise. As a result, desired noise evaluation becomes possible.

  In order to solve the above problem, in another inspection method of the present invention, the horizontal transfer pulse signal includes empty transfer pulses exceeding the number of pulses corresponding to the number of stages of an existing horizontal transfer unit, and the empty transfer pulse The presence or absence of noise in the image signal is inspected by comparing the signal voltage of the image signal output from the charge detection unit with the reference voltage by the transfer according to the above.

  According to this configuration, in the driving pattern for driving the horizontal transfer unit, a predetermined number of empty transfers are added after the final horizontal transfer stage including the horizontal OB (Optical Black) unit and the horizontal dummy unit, and the driving pattern is obtained by driving the horizontal transfer unit. By using the image of the horizontal sky transfer unit, the influence of the dark current component from the photodiode can be avoided.

  By using an inspection method in which the above two methods are combined, a more remarkable effect can be obtained.

  In the above method, preferably, 10 or more idle transfer pulses are applied.

  Preferably, in the above method, the image signal is subdivided into a plurality of image blocks, a signal voltage deviation value is calculated for each block of the image signal, and the calculated deviation value or a difference voltage thereof is calculated. Compare with the reference value.

  According to the inspection method for a solid-state imaging device of the present invention, noise that is a detection target that is not caused by a defect in a pixel portion, such as noise that is generated due to reset drain leakage or noise that occurs in a horizontal transfer portion, changes. Therefore, it is necessary to perform a difference evaluation between a plurality of images obtained under each condition of a high power supply voltage state for emphasizing detection target noise and a low power supply voltage state that does not generate detection target noise. Thus, a desired noise evaluation can be performed.

  In addition, it is possible to avoid the influence of the dark current component from the photodiode and perform high-accuracy evaluation by using the image of the horizontal empty transfer section obtained by adding the empty transfer of the predetermined number of stages and driving it. It is.

  Hereinafter, embodiments of the present invention will be specifically described with reference to the drawings.

(Embodiment 1)
The inspection method for the charge transfer device according to the first embodiment of the present invention will be described in detail with reference to FIGS. The configuration of the CCD type solid-state imaging device that is the object of the inspection method of the present embodiment is the same as that of the conventional example shown in FIG. 9, and will be described with reference to FIG. FIG. 1 shows drive timings during normal driving of the solid-state imaging device. φV1, φV2, φV3, and φV4 represent drive signals for the vertical transfer unit 12, and φH1 and φH2 represent drive signals for the horizontal transfer unit 13. SIG indicates the content of the signal voltage output from the output unit 14. An outline of an image obtained under this driving condition is shown in FIG.

  FIG. 3 is a block diagram illustrating the inspection method for the charge transfer device according to the present embodiment. A drive pulse signal as shown in FIG. 1 is supplied from the drive pulse generation circuit 1 to the CCD 3 constituting the solid-state imaging device, which is necessary for driving the vertical transfer unit 12, the horizontal transfer unit 13, and the like. The power supply unit 2 supplies a power supply voltage necessary for driving the CCD 3.

  As will be described later, the CCD 3 is driven under two different driving conditions, and image signals obtained under the respective driving conditions are stored in the memories 4A and 4B, respectively. The image signals stored in the memories 4A and 4B are supplied to the image block processing circuit 5, and are subjected to statistical processing by subdividing the blocks, and the processing results are stored in the data storage memories 6A and 6B.

  The processing result data read from the data storage memories 6A and 6B is supplied to the difference processing circuit 7. The difference processing circuit 7 performs difference processing on the standard deviation value for each block related to images obtained under two different driving conditions. Based on the output of the comparison result by the difference processing circuit 7, the determination circuit 8 determines that noise is generated in the corresponding block when the difference value is larger than the reference value.

  Next, different driving conditions set for inspection will be described. FIG. 4 shows a cross section of a floating diffusion (FD) portion constituting the charge detection portion in the output portion 14 of the CCD. A reset transistor is composed of a floating diffusion (FD), a reset drain (RD) for discharging signal charges, and a reset gate (RG) provided between the floating diffusion (FD) and the reset drain (RD). ing. The signal charge accumulated in the floating diffusion (FD) is converted into a signal voltage and output from Vout.

  In this floating diffusion portion, when the electric field strength at both ends of the oxide film under the reset gate (RG) becomes strong, hot electrons are generated in the oxide film. The hot electrons are output through the reset gate (RG) and the signal detection capacitor CR, thereby causing noise and image quality degradation. Here, the generation of hot electrons and the generation of noise can be controlled by changing the voltage of the reset drain (RD). In the present embodiment, this phenomenon is actively used to inspect the CCD.

  Next, the operation of the inspection method for the charge transfer device configured as described above will be described. First, the drive pulse generation circuit 1 (see FIG. 3) generates a drive pulse signal as shown in FIG. 1, and the power supply unit 2 emphasizes noise in the power supply voltage supplied to the reset drain (RD). A first drive pattern that is set to a high voltage is generated. The CCD 3 is driven using the generated drive voltage, and the obtained image signal is stored in the frame memory 4A.

  The image signal stored in the frame memory 4A is subdivided into the number of X1 × Y blocks as shown in FIG. 2B by the image block processing circuit 5, and the signal voltage is statistically processed for each block to obtain the standard deviation. The value σ is calculated. Then, the calculated value is stored in the data storage memory 6A. Here, by dividing the block and obtaining the standard deviation value σ for each block, data variation within each block can be obtained. As described above, by dividing the block, the base noise can be removed and the local noise can be detected with high sensitivity.

  Next, a drive pulse signal as shown in FIG. 1 is generated by the drive pulse generation circuit 1, and the power supply voltage supplied to the reset drain (RD) by the power supply unit 2 is set to a low voltage so as to reduce noise. A second drive pattern to be set is generated. The CCD 3 is driven using the generated drive signal, and the image signal is stored in the frame memory 4B.

  The image signal stored in the frame memory 4B is also subdivided into the number of X1 × Y blocks as shown in FIG. 2B by the image block processing circuit 5, and the statistical processing of the signal voltage is performed for each block. The deviation value σ is calculated. The measured value is stored in the data storage memory 6B. As shown in FIG. 5, the standard deviation value σ indicates that the larger the value, the greater the variation within the block. The value increases when abnormal data exists.

  The data stored in the data storage memories 6A and 6B is subjected to differential processing for each block. If the difference is larger than the reference value, the determination circuit 8 determines that the area of the block coordinates (i, j) is present and is larger than the reference value. If not, the area of the block coordinates (i, j) is determined as having no noise.

(Embodiment 2)
The inspection method for the charge transfer device according to the second embodiment of the present invention will be described in detail with reference to FIGS. Although the basic operation is the same as that of the first embodiment, a highly accurate inspection can be performed by using the following driving method together.

  The driving timing during normal driving is the same as that shown in FIG. 1, and the output image obtained under this driving condition is the same as that shown in FIG. When carrying out the inspection method of the present embodiment, using the solid-state imaging device driving method shown in FIG. 6, a pattern of an image signal is created by adding a predetermined horizontal empty transfer section after the horizontal OB section and dummy section. To do. Here, the empty transfer unit is a signal obtained from the output unit 14 when the transfer is performed by the number of stages of the actual horizontal transfer unit 13 or more. An image of an output image obtained under this driving condition is shown in FIG. An image A in FIG. 7A is an image obtained by transfer corresponding to the number of stages of the horizontal transfer unit 13, and an image B is an image corresponding to the empty transfer unit. FIGS. 7B and 7C show the state of the image signal along the lines A-A ′ and B-B ′ in the images A and B, respectively.

  When an image output by adding a horizontal empty transfer unit after the OB unit is viewed in the horizontal direction, the output of the left OB unit + effective pixel unit + right OB unit is obtained from the normal transfer unit, and the empty transfer unit A dark signal without a pixel signal can be obtained.

  FIG. 8 is a block diagram illustrating an inspection method for the charge transfer device according to the present embodiment. A drive pulse signal necessary for driving the vertical transfer unit 12 and the horizontal transfer unit 13 is supplied from the drive pulse generation circuit 1 to the CCD 3. The power supply unit 2 supplies a power supply voltage necessary for driving the CCD 3.

  The image signal output from the CCD 3 is stored in the frame memory 4. The image signal stored in the frame memory 4 is supplied to the image block processing circuit 5, and the signal voltage is statistically processed after being divided into blocks, and the processing result is stored in the data storage memory 6. When the data of the processing result is supplied from the data storage memory 6 to the determination circuit 8 and the value of the data is larger than the reference value, it is determined that noise is generated in the corresponding block.

  Next, an operation based on this inspection method will be described. The drive pulse generation circuit 1 adds a signal for obtaining a horizontal empty transfer unit as shown in FIG. 6 to generate a drive pulse signal for obtaining an image signal with only noise that is not affected by the pixel signal. The power supply unit 2 generates a power supply voltage necessary for driving the CCD 3. The CCD 3 is driven by the generated drive signal, and the image signal is stored in the frame memory 4.

  FIG. 9 shows the state of the image signal stored in the frame memory 4. The area of the X1 block corresponds to an effective pixel part, and the area of the X2 block corresponds to an empty transfer part. In the present embodiment, the image signal of the empty transfer unit is subdivided into the number of X2 × Y blocks by the image block processing circuit 5, and the statistical processing is performed on the signal voltage of the image signal for each block. The standard deviation value σ corresponding to the degree of signal voltage variation is calculated. The calculated standard deviation value σ is stored in the data storage memory 6.

  When the standard deviation value σ stored in the data storage memory 6 is larger than the reference value, the determination circuit 8 determines that the area of the block coordinates (i, j) has noise. On the other hand, when the standard deviation value σ is not larger than the reference value, the area of the block coordinate (i, j) is determined as having no noise.

  In order to improve the measurement accuracy, it is desirable to provide 10 or more idle transfer pulses.

(Embodiment 3)
The inspection method for the charge transfer device according to the third embodiment has a configuration in which the first and second embodiments are combined. That is, in the second embodiment, the power supply voltage supplied to the reset drain (RD) is fixed, but in this embodiment, an empty transfer unit is provided and the reset voltage is reset as in the first embodiment. The power supply voltage supplied to the drain (RD) is set to a high voltage in order to emphasize noise, and is set to a low voltage to reduce noise, and image signals are captured in two cases. By performing arithmetic processing such as standard deviation calculation based on the image signal obtained by the idle transfer, it is possible to perform more accurate noise evaluation.

  This will be described below with reference to the drawings. As a driving condition, as in FIG. 6, a pattern in which a predetermined empty transfer unit is added after the horizontal OB unit and the dummy unit is created, and an image signal as shown in FIG. 7 is obtained. The inspection method for the charge transfer device in the present embodiment is represented by the block diagram of FIG. 3 as in the first embodiment.

  Next, an operation based on this inspection method will be described. The drive pulse generation circuit 1 adds a signal for obtaining a horizontal empty transfer unit as shown in FIG. 6 to generate a drive pulse signal that obtains an image signal with only noise and no pixel signal. At the same time, the power supply unit 2 generates a first drive pattern that sets the power supply voltage supplied to the reset drain (RD) to a high voltage for emphasizing noise. The CCD 3 is driven using the generated drive signal, and the obtained image signal is stored in the frame memory 4A.

  Then, the image block processing circuit 5 subdivides the image signal obtained by horizontal empty transfer among the image signals stored in the frame memory 4A into X2 × Y blocks as shown in FIG. Processing is performed to calculate a standard deviation value σ. The calculated standard deviation value σ is stored in the data storage memory 6A.

  Next, the drive pulse generation circuit 1 generates a drive pulse signal for obtaining a noise-only image signal by providing a horizontal empty transfer section as described above, and the power supply section 2 supplies power to the reset drain (RD). A second drive pattern is generated for setting the voltage to a low voltage for reducing noise. The CCD 3 is driven using the generated drive signal, and the obtained image signal is stored in the frame memory 4B.

  Then, in the image block processing circuit 5, the image signal obtained by horizontal empty transfer among the image signals stored in the frame memory 4B is subdivided into X2 × Y blocks as shown in FIG. 9, and statistical processing is performed on each block. To calculate the standard deviation value σ. The calculated standard deviation value σ is stored in the data storage memory 6B.

  The difference processing circuit 7 performs the difference processing of the data stored in the data storage memories 6A and 6B for each block. If the difference processing circuit 7 is larger than the reference value, the determination circuit 8 determines that the area of the block coordinate (i, j) has noise. judge. When the difference is not larger than the reference value, the area of the block coordinate (i, j) is determined as having no noise.

  This is useful for inspection and evaluation of low-period horizontal noise caused by an amplifier generated mainly in a floating diffusion portion in a CCD.

Timing waveform chart showing normal drive pattern specifications of CCD (A) is a figure which shows the image by the image signal taken in by the normal drive pattern specification, (b) is a figure which shows the block process of the image signal in Embodiment 1. FIG. FIG. 2 is a block diagram showing an inspection method for a charge transfer device according to Embodiment 1 of the present invention. Sectional view showing floating diffusion in solid-state image sensor Graph showing examples of standard deviation values calculated for each image block Timing waveform diagram showing drive pattern specifications in Embodiment 2 of the present invention (A) is a figure which shows the image by the image signal taken in by the drive pattern specification, (b), (c) is a figure which shows the state of the image signal in the same image The block diagram which shows the inspection method of the charge transfer apparatus in Embodiment 2 of this invention FIG. 10 is a diagram showing block processing of an image signal in the second embodiment The figure which shows the structure of a general CCD type solid-state image sensor Block diagram showing a conventional image noise discrimination method Explanatory drawing of the difference detection operation of the conventional example

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Drive pulse generation circuit 2 Power supply part 3 CCD (solid-state imaging device)
4, 4A, 4B Frame memory 5 Image block processing circuit 6, 6A, 6B Data storage memory 7 Difference processing circuit 8 Determination circuit 10 Solid-state imaging device 11 Light receiving unit 12 Vertical transfer unit 13 Horizontal transfer unit 14 Output unit 15 Effective pixel unit 16 Horizontal OB unit 17 Horizontal dummy transfer unit 21 Luminance conversion circuit 22 Frame memory 23 Frame memory 24 Frame memory 25 Frame memory 26 Frame memory 27 Changeover switch 28 Difference detection circuit 29 Determination circuit

Claims (6)

A light receiving unit in which pixels having photodiodes are two-dimensionally arranged and photoelectrically convert incident light into a signal charge; a vertical transfer unit that transfers a signal charge read from the light receiving unit in a vertical direction by a vertical transfer pulse signal; A horizontal transfer unit that horizontally transfers the signal charge transferred from the vertical transfer unit by a horizontal transfer pulse signal; a charge detection unit that converts the charge transferred from the horizontal transfer unit into a voltage and detects the voltage; A charge transfer device inspection method comprising a power supply unit for setting a reset voltage of the charge detection unit,
A first image signal output from the charge detection unit when a first reset voltage is applied to the charge detection unit by the power supply unit, and a second reset voltage is obtained by the power supply unit. A second image signal output from the charge detection unit when applied to the first and second image signals is obtained from the charge detection unit by comparing a differential voltage between the first and second image signals with a reference voltage. A method for inspecting a solid-state imaging device, wherein the presence or absence of noise in an image signal is inspected. A light receiving unit in which pixels having photodiodes are two-dimensionally arranged and photoelectrically convert incident light into a signal charge; a vertical transfer unit that transfers a signal charge read from the light receiving unit in a vertical direction by a vertical transfer pulse signal; A horizontal transfer unit that horizontally transfers the signal charge transferred from the vertical transfer unit by a horizontal transfer pulse signal; a charge detection unit that converts the charge transferred from the horizontal transfer unit into a voltage and detects the voltage; A charge transfer device inspection method comprising a power supply unit for setting a reset voltage of the charge detection unit,
The horizontal transfer pulse signal includes empty transfer pulses exceeding the number of pulses corresponding to the number of stages of the existing horizontal transfer unit,
An inspection method for a solid-state imaging device, wherein the presence or absence of noise in the image signal is inspected by comparing a signal voltage of the image signal output from the charge detection unit by transfer using the idle transfer pulse with a reference voltage . The horizontal transfer pulse signal includes empty transfer pulses exceeding the number of pulses corresponding to the number of stages of the existing horizontal transfer unit,
The presence or absence of noise in the image signal output from the charge detection unit is inspected by comparing a differential voltage between the first and second image signals acquired by the transfer by the empty transfer pulse with the reference voltage. An inspection method for a solid-state imaging device according to Item 1.   4. The inspection method for a solid-state imaging device according to claim 2, wherein 10 or more of the empty transfer pulses are applied.   The first and second image signals are subdivided into a plurality of image blocks, and a deviation value of the signal voltage is calculated for each of the image blocks for the first and second image signals, and the calculated deviation is calculated. The solid-state imaging device inspection method according to claim 1, wherein a difference voltage between values is compared with the reference voltage.   The solid image according to claim 2, wherein the image signal is subdivided into a plurality of image blocks, a deviation value of a signal voltage is calculated for each image block of the image signal, and the calculated deviation value voltage is compared with the reference value. Inspection method for imaging apparatus.

Images