Sensor device and method for determining correction factors

Abstract

To effectively reduce the noise of a measurement signal in a sensor device.SOLUTION: A sensor device includes multiple pixels, and a control device that corrects measurement signals from the multiple pixels. Each of the multiple pixels includes a photodetector, and a pixel circuit that outputs the signal from the photodetector, and the control device acquires an unknown measurement signal of a first pixel of the plurality of pixels. The control device obtains an unknown measurement signal of one pixel of the plurality of pixels, and corrects the unknown measurement signal due to one pixel by using a correction factor based on the ratio of the statistical value of the measurement signal of multiple pixels to the value obtained from the measurement signal of one pixel, each with a plurality of reference beams with different intensities.SELECTED DRAWING: Figure 7

Description

[Technical field]

[0001] The present disclosure relates to techniques for reducing noise in signals from photodetectors. [Background technology]

[0002] Noise contained in the output signal from an optical imaging element is classified into random noise, which varies over time, and fixed pattern noise (FPN), which does not vary over time. Furthermore, there are two types of FPN: dark FPN, which is observed when there is no incident light, and bright FPN, whose magnitude changes depending on the intensity of the incident light. Dark FPN is caused by variations in the input / output characteristics of the pixel circuits and readout circuits of the imaging element. On the other hand, bright FPN is caused by variations in the characteristics of the photoelectric conversion element, and it is known that the noise intensity increases as the incident light intensity (number of photons) increases.

[0003] Since FPN does not vary over time and is spatially fixed as a characteristic variation for each imaging pixel, it can be suppressed by appropriate signal processing. For example, a method is known to reduce dark FPN by storing a signal in a dark state in memory and subtracting it from the detection signal of the photodetector.

[0004] As a method for suppressing the bright FPN, for example, a method is known in which a reference signal in a uniform reference light irradiation state of 1 or 2 is prepared and subtracted from the detection signal. Furthermore, Patent Document 1 proposes a method for suppressing the bright FPN by dividing the detection signal by the reference signal in the uniform reference light irradiation state. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2015-100099 A Summary of the Invention [Problem to be solved by the invention]

[0006] However, conventional methods have not been able to sufficiently reduce the light-state FPN. Therefore, a technology that can more effectively reduce the light-state FPN of photodetectors is desired. [Means for solving the problem]

[0007] A sensor device according to an embodiment of the present disclosure includes a plurality of pixels and a control device that corrects a measurement signal from the plurality of pixels. Each pixel of the plurality of pixels includes a photodetector and a pixel circuit that outputs a signal from the photodetector. The control device acquires an unknown measurement signal from one pixel among the plurality of pixels, and corrects the unknown measurement signal from the one pixel using a correction coefficient based on a ratio between a statistical value of the measurement signals from the plurality of pixels and a value obtained from the measurement signal from the one pixel, for each of a plurality of reference lights having different intensities.

[0008] One aspect of the present disclosure is a method for determining a correction coefficient used in correcting a measurement signal of a pixel by a control device of a sensor device, wherein each pixel of a plurality of pixels of the sensor device includes a photodetector and a pixel circuit that outputs a signal from the photodetector, the method includes acquiring a measurement signal of each pixel of the plurality of pixels for each of the plurality of reference beams having different intensities, determining a statistical value of the measurement signals of the plurality of pixels for each of the plurality of reference beams having different intensities, and determining a correction coefficient for each pixel of the plurality of pixels based on the statistical value of the measurement signals of the plurality of pixels for each of the plurality of reference beams having different intensities and the measurement signal of each pixel of the plurality of pixels. Effect of the Invention

[0009] According to one aspect of the present disclosure, it is possible to effectively reduce noise contained in a signal from a photodetector. [Brief description of the drawings]

[0010] [Figure 1] 1 is a block diagram showing an example of the configuration of an image sensor according to a first embodiment. [Diagram 2]2 illustrates a schematic configuration example of a main control device. [Diagram 3] FIG. 2 is a circuit diagram showing an example of a circuit configuration of a pixel. [Figure 4] 2 is a signal flow chart showing a calculation method for noise reduction processing according to the related art; [Diagram 5] 4 shows experimental results of noise reduction processing according to the related art on a measurement signal. [Figure 6] 4 shows experimental results of noise reduction processing according to the related art on a measurement signal. [Figure 7] 1 is a signal flow diagram illustrating a method of computing a noise reduction process according to an embodiment of the present disclosure; [Figure 8] 4 is a signal flow diagram illustrating a method for determining weighting factors according to an embodiment of the present disclosure; [Figure 9] 13 shows an example of measured pixel data before noise removal. [Figure 10] 5 shows the result of the correction process according to the related art described with reference to FIG. [Figure 11] 9 illustrates the results of a correction process according to an embodiment of the present disclosure described with reference to FIGS. [Figure 12] The relationship between the irradiated light intensity and the effective noise voltage is shown for the portion surrounded by the dashed rectangle in FIG. 10 and for the portion surrounded by the dashed rectangle in FIG. [Figure 13] In the second embodiment, the relationship between the irradiated light intensity and the effective noise voltage is shown. [Figure 14] 1 shows an example of photoelectric conversion characteristics of a pixel including a photodiode. [Figure 15] 4 shows an example of a noise voltage characteristic of a pixel. [Figure 16] 1 shows a schematic cross-sectional structure of an OLED panel with a sensor. [Figure 17] 2A and 2B are schematic diagrams illustrating an example of a cross-sectional structure of an optical sensor array. [Figure 18] 1 shows a schematic cross-sectional structure of an X-ray sensor panel. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] The image sensor of the present disclosure will be described in detail below with reference to the drawings. The size and scale of each component in each drawing are appropriately changed to ensure the visibility of the drawing. Hatching in each drawing is for distinguishing each component and does not necessarily mean a cut surface. Nonlinear elements used as switching elements or amplifying elements are called transistors, but the transistor includes thin film transistors (TFTs).

[0012] In this specification, unless otherwise specified, the term "light" includes visible light and electromagnetic waves with wavelengths shorter or longer than that of visible light. For example, it includes infrared light, ultraviolet light, and X-rays, which are electromagnetic waves with shorter wavelengths. In other words, unless otherwise specified, a photodetector or photoelectric conversion element is an element that converts electromagnetic waves of any wavelength into an electrical signal.

[0013] <Embodiment 1> 1 is a block diagram showing an example of the configuration of an image sensor according to embodiment 1. An image sensor 10 according to the present disclosure includes a sensor substrate 11 and a control system. The control system includes a drive circuit 14, a signal detection circuit 16, and a main control device 18.

[0014] The sensor substrate 11 includes an insulating substrate (e.g., a glass substrate) and a pixel region 12 in which pixels 13 are arranged in a matrix on the insulating substrate. A scintillator that receives radiation as detection light and emits fluorescence may be arranged in the pixel region 12. A drive circuit 14 drives the pixels 13 for light detection by the pixels 13. A signal detection circuit 16 detects signals from each of the signal lines. A main control device 18 controls the drive circuit 14 and the signal detection circuit 16.

[0015] In this embodiment, the drive circuit 14 and the signal detection circuit 16 are formed as components separate from the sensor substrate 11. These circuits may be mounted on different IC chips, or some or all of these circuits may be mounted on the same IC chip, or one circuit may be mounted on multiple IC chips.

[0016] The main control device 18 may have, for example, a computer configuration. Fig. 2 shows a schematic configuration example of the main control device 18. In the configuration example of Fig. 2, the main control device 18 includes a processor 201, a memory (main storage device) 202, an auxiliary storage device 203, an output device 204, an input device 205, a communication interface (I / F) 207, and an AD conversion interface (ADC) 208. The above components are connected to each other by a bus. The memory 202, the auxiliary storage device 203, or a combination of these, are storage devices, and store programs and data used by the processor 201.

[0017] The memory 202 is composed of, for example, a semiconductor memory, and is mainly used to hold programs and data being executed. The processor 201 executes various processes according to the programs stored in the memory 202. This realizes various functional units.

[0018] The auxiliary storage device 203 is composed of a large-capacity storage device such as a hard disk drive or a solid-state drive, and is used to store programs and data for a long period of time.

[0019] The processor 201 may be comprised of a single processing unit or multiple processing units and may include single or multiple arithmetic units or multiple processing cores. The processor 201 may be implemented as one or more central processing units, microprocessors, microcomputers, microcontrollers, digital signal processors, state machines, logic circuits, graphic processing units, systems on a chip, and / or any device that manipulates signals based on control instructions.

[0020] The programs and data stored in the auxiliary storage device 203 are loaded into the memory 202 at the time of startup or when necessary, and the programs are executed by the processor 201 to perform various processes of the main control device 18. Therefore, in the following, processes executed by the programs are processes by the processor 201 or the main control device 18.

[0021] The input device 205 is a hardware device that allows a user to input instructions, information, and the like to the main control device 18. The output device 204 is a hardware device that presents various images for input and output, such as a display device or a printing device. The AD conversion interface 208 is an interface that converts an input analog signal into a digital signal. The communication I / F 207 is an interface for connection to a network. The input device 205 and the output device 204 may be omitted, and the main control device 18 may be accessed from a terminal via a network.

[0022] The functions of main controller 18 may be implemented in a computer system consisting of one or more computers including one or more processors and one or more storage devices including non-transitory storage media. The computers communicate with each other via a network. For example, some of the functions of main controller 18 may be implemented in one computer and other parts may be implemented in another computer.

[0023] Fig. 3 is a circuit diagram showing an example of the circuit configuration of one pixel 13. Note that Fig. 3 shows an example of the pixel circuit, and other circuit configurations may be used. One pixel 13 of the image sensor of the present disclosure includes four transistors TR1, TR2, TR3, and TR4 and a photodiode PD.

[0024] The photodiode PD is an example of a photodetector (also called a light source conversion element). In the example shown here, the anode terminal of the photodiode PD is connected to the gate terminal of the transistor TR1 and the drain terminal of the transistor TR3, and the cathode terminal is connected to the power supply line PA. The drain terminal of the transistor TR1 is connected to the power supply line PP, and the source terminal is connected to the drain terminal of the transistor TR2.

[0025] The gate terminal of transistor TR2 is connected to control line Gn, and the source terminal is connected to signal line Dm. The gate terminal of transistor TR3 is connected to control line Rn, and the source terminal is connected to power supply line PB. The gate terminal of transistor TR4 is connected to bias line BI, the drain terminal is connected to signal line Dm, and the source terminal is connected to power supply line VE. The power supply potential of power supply line VE is lower than the power supply potential of power supply line PP. The bias line BI provides a constant bias potential to the gate terminal of transistor TR4.

[0026] The photodiode PD functions to convert light into an electric charge. The transistor TR1 (amplifying transistor) functions to amplify the potential at one end of the photodiode PD. The transistor TR2 functions to control the output. The transistor TR3 functions to reset the potential of the photodiode PD. The transistor TR4 functions as a resistor. The signal line Dm transmits a detection signal Vout of light by the photodiode PD to the signal detection circuit 16. The drive circuit 14 supplies a control signal and a power supply potential to the control line and power supply line shown in FIG. 3.

[0027] The following describes the processing executed by the main control device 18. The main control device 18 corrects the measurement signal of the light detected by each pixel 13 on the sensor substrate 11, and reduces the noise. The noise contained in the measurement signal (output signal) from the pixel 13 is classified into random noise that varies with time, and fixed pattern noise (FPN) that does not vary with time.

[0028] There are two types of FPN: dark FPN, which is observed when there is no incident light, and bright FPN, whose magnitude changes depending on the intensity of the incident light. Dark FPN is caused by variations in the input / output characteristics of pixel circuits and readout circuits. On the other hand, bright FPN is caused by variations in the characteristics of photodiodes, and it is known that the noise intensity increases with increasing incident light intensity (number of photons). FPN does not vary over time, but is fixed spatially as the characteristic variations for each pixel. The process described below reduces dark FPN and bright FPN. In the following, dark FPN is also called additive FPN and bright FPN is also called multiplicative FPN.

[0029] An embodiment herein assumes the following noise model in the measurement signal by the photodiode: Sx = So + A * So + Rd (1)

[0030] Sx is the measurement signal due to the pixel including noise. Rd is the measurement signal in the dark state. This is (additive FPN). So is the true light detection signal without noise. A is the multiplicative FPN coefficient. The main controller 18 corrects the measurement signal Sx received from the pixel 13 to remove the additive FPN and multiplicative FPN from the measurement signal Sx to obtain the true light detection signal So.

[0031] Before describing the signal processing for noise reduction in the embodiment of this specification, the signal processing for noise reduction in the related art will be described. The related art prepares three measurement data. These are a measurement signal Rd in a dark state, a measurement signal Rl when uniformly irradiated with light, and a measurement signal Sx from which a true light detection signal So is to be extracted in actual use.

[0032] 4 is a signal flow chart showing a calculation method of noise reduction processing according to the related art. The related art acquires a measurement signal Rd in the dark state of all pixels 13, and then acquires a measurement signal Sx to be corrected for one target pixel 13. The dark state in the measurement is ideally a state in which light below the detection limit of the pixel 13 is present, but may be a state in which light with an intensity below a threshold value according to the design is present.

[0033] The related technique subtracts the measurement signal Rd of the target pixel 13 in the dark state from the measurement signal Sx of the target pixel 13 to obtain a signal S1 (301). S1=Sx-Rd (2) As a result, the additive FPN is removed from the measurement signal Sx of the target pixel 13.

[0034] The related art further subtracts the measurement signals Rd of all pixels 13 in the dark state from the measurement signals Rl of all pixels 13 when irradiated with uniform light (reference light) (bright state) to obtain signals R1 of all pixels 13 (302). The uniform light intensity is within the detectable range of the pixels 13 and is much higher than the intensity of light present in the dark state. R1 = Rl - Rd (3) This removes the additive FPN from the measurement signal Rl when uniform light is irradiated.

[0035] The related technique divides the signal S1 of the target pixel 13 by the signal R1 of the target pixel 13 (303). The related technique further divides the average value of the signals R1 of all pixels 13. <r1>Furthermore, the related art removes the multiplicative FPN contained in the signal S1 by the following calculation (305): So^=(S1 / R1)* <r1>(4) Sô is an estimate of the true photodetection signal So without noise.

[0036] The reason why the related art presumes S1^ as the true light detection signal So will be explained. According to the above noise model, the following relationship holds: S1=So*(1+A) (5) Similarly, the following relationship holds: R1 = Ro * (1 + A) (6) Ro is the true light detection signal of the measurement signal Rl.

[0037] Related technology is to average all pixels of signal R1 <r1>is estimated to be the true signal value Ro of the signal R1. Ro= <r1>*(1+A) (7) In the related art, a multiplicative FPN coefficient A is calculated from the formula (7) and substituted into the formula (5) to calculate an estimated value So^ of the light detection signal So of the measurement signal Sx.

[0038] 5 and 6 show experimental results of noise reduction processing by the related art for the measurement signal Sx. Fig. 5 shows the measurement signal (Raw Data) measured under exposure conditions of uniform light of four different intensities. In the graph of Fig. 5, the horizontal axis indicates the pixel position in the row or column direction, and the vertical axis indicates the output voltage of the pixel.

[0039] The irradiated light intensities of the measurement signals Sa, Sb, Sc, and Sd increase in this order. The signal Sd is the measurement signal when the light intensity is the highest, and the signal Sa is the measurement signal when the light intensity is the lowest. As shown in FIG. 5, each measurement signal shows a large noise.

[0040] Fig. 6 shows the results Soa^, Sob^, Soc^, and Sod^ of reducing the noise of the measurement signals Sa, Sb, Sc, and Sd using equation (4) with the measurement signal Sa as the reference signal Rl. As shown by the signal Sa^, the noise of the measurement signal Sa is completely removed. This is because S1=R1 in equation (4).

[0041] However, as the light irradiation level deviates from that of the measurement signal Sa, the FPN of the measurement signals Sb, Sc, and Sd increases. Therefore, for example, as shown by the signal Sd^, in the exposure conditions that deviate significantly from that of the reference light signal Sa, the multiplicative FPN cannot be completely removed in the signal processed by formula (4).

[0042] Next, a signal processing for reducing noise in a measurement signal of a photodiode according to an embodiment of the present specification will be described. In an embodiment of the present specification, the measurement signal is corrected using a correction coefficient based on a plurality of reference light signals (measurement signals of reference light) with different intensities. This makes it possible to effectively remove noise from the measurement signal. The wavelength components of the plurality of reference light beams may be the same, for example. The intensity and wavelength components of the reference light are appropriately determined according to the characteristics of the light that the image sensor 10 is intended to detect.

[0043] The main control device 18 uses a correction coefficient based on multiple reference lights with different intensities to perform a correction process on the optical measurement signal Sx from each pixel 13. This effectively reduces the noise component of the measurement signal Sx of each pixel 13, and can estimate the true optical signal So. As shown in the noise model of equation (1), the measurement signal Sx has additive FPN and multiplicative FPN.

[0044] The example described below reduces the additive FPN and the multiplicative FPN. Here, a coefficient based on multiple reference beams is used to reduce the multiplicative noise. If the additive FPN is small, the calculation for reducing the additive FPN may be omitted. The main controller 18 can correct the measurement signal Sx of each pixel 13 to obtain an estimate of the true signal So^ by the following formula: So^=(Sx-Rd)*f (8)

[0045] Rd is the measurement signal Rd of each pixel 13 in the dark state, and is a correction coefficient for reducing the additive FPN. f is a correction coefficient for reducing the multiplicative FPN, and is determined based on the measurement results with multiple reference lights having different intensities. In one embodiment of the present specification, the correction coefficient for a specific pixel is based on the ratio between the statistical value of the measurement signals of multiple pixels with each of multiple reference lights having different intensities and the value obtained from the measurement signal of the specific pixel. In one embodiment of the present specification, the correction coefficient can be expressed as a linear combination of the ratio with multiple reference lights having different intensities and a weighting coefficient.

[0046] In actual use, the main control device 18 holds the correction coefficients Rd and f for each pixel 13 in advance, corrects the measurement signal Sx of each pixel 13 using the correction coefficients Rd and f, and obtains an estimate value So^ of the true light detection signal. Note that the main control device 18 may hold other coefficients for calculating f.

[0047] The method of determining the correction coefficient f will be described below. For example, the main control device 18 or the manufacturing device of the image sensor 10 can determine the correction coefficient f. Figure 7 is a signal flow chart showing a calculation method of the noise reduction process according to one embodiment of the present specification.

[0048] The flowchart in Fig. 7 shows a method for calculating the correction coefficient f and a method for correcting the measurement signal Sx using the correction coefficients Rd and f for each target pixel 13. In the following, it is assumed that the main control device 18 executes the process shown in Fig. 7. The explanation of one reference signal with reference to Fig. 4 can be applied to the process for each reference light in Fig. 7.

[0049] The main control device 18 acquires a measurement signal Rd of all the pixels 13 in a dark state, and measurement signals Rl1 to Rln of all the pixels 13, respectively, with first to nth reference light (uniform light) of different intensities, where n is an integer equal to or greater than 2. The measurement signals Rd, Rl1 to Rln may be acquired only from a portion of the pixels 13.

[0050] Main controller 18 subtracts measurement signal Rd from measurement signal Rl1 for each pixel 13 of all pixels 13 to obtain R1 (351). This removes additive FPN from measurement signal Rl1 when uniformly irradiated. R1 = Rl1 - Rd (9)

[0051] Furthermore, the main control device 18 calculates the average value of R1 for all pixels 13. <r1>Then, the main controller 18 calculates the average value (352). <r1>is divided by the signal R1 of the correction target pixel 13 (353). Note that a statistical value other than the average value, for example, the median value, may be used. <r1> / R1 (10)

[0052] Main controller 18 also executes the processes executed for signal R1 for the other reference light signals R2 to Rn. For example, main controller 18 executes processes 354, 355, and 356 for measurement signal R2. Main controller 18 also executes processes 357, 358, and 359 for measurement signal Rn.

[0053] Next, the main control device 18 uses the weighting coefficients k1 to kn that have been calculated and stored in advance to <r1> / R1~ <rn>A weighted average of / Rn is calculated (360). The weighting coefficients k1 to kn are normalized and their sum is 1. Σkm* <rm> / Rm (11)

[0054] m is any one of 1 to n. This weighted average is used as the correction coefficient f. In this way, by using measurement signals of reference light with different intensities, the measurement signal Sx, which can have various intensities, can be corrected more appropriately. In addition, by giving a weighting coefficient to each reference light, the measurement signal Sx can be corrected more appropriately. The value of the weighting coefficient depends on the image sensor 10, but typically at least some of the weighting coefficients have different values.

[0055] The master controller 18 subtracts the measurement signal Rd of the pixel 13 in the dark state from the measurement signal Sx of the pixel 13 of interest to obtain a signal S1 (365). S1=Sx-Rd (12) This removes the additive FPN from the measurement signal Sx of interest.

[0056] Next, main control device 18 multiplies coefficient f by signal S1 to calculate estimate value So^ of the true light detection signal of target pixel 13 (366). So^=S1*f (13)

[0057] As described above, one embodiment of the present specification derives a unique multiplicative noise factor for each pixel from a plurality of uniform light illumination reference measurement signals with different intensities, and uses the factor to correct the unknown measurement signal for each pixel, thereby effectively reducing noise over a wide range of illumination intensities.

[0058] A method for determining the weighting coefficients k1 to kn will be described below. In one embodiment of the present specification, the weighting coefficients k1 to kn are determined so that the noise effective voltage of all pixel signals when uniformly irradiated with reference light is minimized. The weighting coefficients may be appropriately set depending on the design of the image sensor 10. The main control device 18 or a manufacturing device for the image sensor 10 may determine the weighting coefficients k1 to kn. In the following, it is assumed that the main control device 18 determines the weighting coefficients k1 to kn.

[0059] 8 is a signal flow diagram illustrating a method for determining weighting factors according to an embodiment of the present disclosure. For each pixel 13 of all pixels 13, the main controller 18 subtracts its measurement signal Rd from its measurement signal Rl1 to obtain R1 (401). R1 = Rl1 - Rd (14)

[0060] Next, the main controller 18 calculates the average value of R1 for all pixels 13. <r1>Furthermore, the main control device 18 calculates the average value from the signal R1 of each pixel 13 (402). <r1>Then, the main controller 18 subtracts (403) the signal R1 and the average value for each pixel 13 of all pixels. <r1>The squared difference between <r1>) 2 Calculate (404).

[0061] The main controller 18 also executes the processes executed for the signal R1 for the other reference light signals R2 to Rn. For example, the main controller 18 executes processes 405 to 408 for the signal R2. In addition, the main controller 18 executes processes 409 to 412 for the signal Rn.

[0062] Next, main controller 18 calculates a weighted average of the squared deviations from the all-pixel average for all pixels and all reference light level signals R1 to Rn (415). More specifically, main controller 18 calculates for each pixel 13, (Rm- <rm>) 2 and the weighting coefficient km, and then calculates the sum of the calculated values ​​for all pixels 13. Here, m is an arbitrary integer from 1 to n, and the weighting coefficient is normalized. ΣΣkm*(Rm- <rm>) 2 (15)

[0063] In equation (15), the first Σ represents the SUM for all pixels, and the second Σ represents the SUM for all n reference beams at each pixel 13. Main controller 18 determines weighting coefficients k1 to kn (416) by an optimization loop that minimizes the value found by equation (15).

[0064] The effect of correction document processing of the measurement signal of pixel 13 according to an embodiment of this specification will be described below. FIG. 9 shows an example of actual measurement data of pixel 13 before noise removal. In the graph of FIG. 9, the horizontal axis indicates pixel position, and the vertical axis indicates output voltage (V) from pixel 13. FIG. 9 shows measurement signals Rl1 to Rl4 of four irradiation lights. The light source intensities (au) of these irradiation lights are 120, 160, 200, and 240.

[0065] For the measurement, a pseudo light signal formed by an optical modulation film made of one or two layers of transparent PET film was irradiated onto the pixel array. In Figure 9, the sharp peak indicates that the light intensity is reduced due to the influence of the side of the film. The smaller the output voltage, the greater the light illuminance.

[0066] Fig. 10 shows the result of the correction process according to the related art described with reference to Fig. 4. The reference light is Rl1, and its light intensity is 120. Fig. 11 shows the result of the correction process according to an embodiment of the present specification described with reference to Figs. 7 and 8. All four irradiation light beams Rl1 to Rl4 are reference light beams for determining the correction coefficients.

[0067] Fig. 12 shows the relationship between the irradiation light intensity and the noise effective voltage for the portion surrounded by the dashed rectangle 501 in Fig. 10 and the portion surrounded by the dashed rectangle 502 in Fig. 11. In the graph of Fig. 12, the horizontal axis shows the irradiation light intensity of the reference light, and the vertical axis shows the noise effective voltage (V). A dashed line 511 shows the noise after the correction process of the related art, and a solid line 512 shows the noise after the correction process of an embodiment of this specification. The intensities of the irradiation lights Rl1 to Rl4 are 120, 160, 200, and 240, respectively, as described above.

[0068] 12, the correction process result of the related art shows an effective voltage of noise that increases as the irradiation light intensity moves away from the reference light Rl1. On the other hand, the correction process result of the embodiment of this specification shows a large noise reduction effect over a wide range of irradiation light intensities. In this way, the correction process of the embodiment of this specification can effectively reduce noise over a wide range of irradiation intensities.

[0069] <Embodiment 2> In the following, a signal processing for noise reduction of a measurement signal of a photodiode according to an embodiment of the present specification will be described. The noise reduction processing described below adaptively determines a correction coefficient for reducing the multiplicative FPN based on the magnitude of the measurement signal. This makes it possible to effectively reduce noise in a wide range of irradiation intensity.

[0070] Similar to the first embodiment, the main control device 18 of this embodiment can correct the measurement signal Sx of each pixel 13 to obtain an estimate value So^ of the true signal. So^=(Sx-Rd)*f (16)

[0071] In formula (16), the measurement signal Sx and the correction coefficient Rd of the additive FPN are the same as formula (8) in embodiment 1. That is, Rd is the measurement signal of each pixel 13 in a dark state. The method of deriving the correction coefficient f of the multiplicative FPN in this embodiment is different from the derivation method in embodiment 1. Other points are the same as in embodiment 1. The method of deriving the correction coefficient f will be described below.

[0072] In the noise reduction process of this embodiment, similarly to the first embodiment, a correction coefficient is determined using a plurality of reference signals obtained by irradiating a plurality of reference beams with different intensities. The example described below uses two reference signals of uniform reference beams. It is also possible to use measurement signals of three or more reference beams with different intensities. As in the first embodiment, the following signals are obtained.

[0073] Sx: Measurement signal of one pixel to be corrected Rl1: Measurement signal in reference light 1 for each pixel Rl2: Measurement signal in reference light 2 for each pixel Rd: Measurement signal in the dark state of each pixel

[0074] The master controller 18 subtracts the measurement signal Rd of the pixel 13 in the dark state from the measurement signal Sx of the pixel 13 of interest to obtain a signal S1. S1=Sx-Rd (17) This removes the additive FPN from the measurement signal Sx of interest.

[0075] Main control device 18 subtracts measurement signal Rd from measurement signal Rl1 for each pixel 13 of all pixels 13 to obtain R1, thereby removing additive FPN from measurement signal Rl1 when uniform light is irradiated. R1 = Rl1 - Rd (18)

[0076] Furthermore, main control device 18 subtracts measurement signal Rd from measurement signal Rl2 for each pixel 13 of all pixels 13 to obtain R2. This removes additive FPN from measurement signal Rl2 when uniform light is irradiated. R2 = Rl2 - Rd (19)

[0077] The main controller 18 calculates the average of all pixels 13 of the signal R1. <r1>And the average of all 13 pixels of signal R2 <r2>As in the first embodiment, these may be the average or median of some of the pixels. Next, main control device 18 calculates a weighting coefficient for each pixel 13, which varies depending on signal S1 having unknown intensity, using the following formula. The following formula determines the coefficient by linear interpolation. <r1>but <r2>If the unknown signal S1 is larger than <r1>When the unknown signal S1 is equal to <r2>When k1 is equal to KMIN and k2 is equal to KMAX, <r1> >S1> <r2>When , k1 and k2 take values ​​between KMAX and KMIN.

[0078] k1=KMIN +(S1- <r2>)*(KMAX-KMIN) / ( <r1> - <r2>) k2=KMAX -(S1- <r2>)*(KMAX-KMIN) / ( <r1> - <r2>) (20) KMAX and KMIN are constants, where KMAX>KMIN.

[0079] Main control device 18 calculates a correction coefficient f for the integral VPN according to the following formula. f=(k1* <r1> / R1+k2* <r2> / R2) / (k1+k2) (21)

[0080] The main control device 18 multiplies the coefficient f by the signal S1 to calculate an estimate value So^ of the true light detection signal of the target pixel 13. So^=S1*f (22)

[0081] For example, the simplest is <r1>Maximum exposure conditions, <r2>As the minimum exposure condition, KMAX=1 and KMIN=0, so that k1 and k2 are values ​​between 0 and 1. However, the values ​​of KMAX and KMIN are not limited to these. When KMAX=1 and KMIN=0, <r2>If you set a value larger than the minimum exposure condition, S1 <r2>When the value becomes smaller, there may be a case where k1 becomes a negative value.

[0082] Since k1 and k2 must be positive values, in such a case, for example, KMIN=1 and KMAX=2 are set so that k1 does not become a negative value. Furthermore, values ​​optimized to minimize the overall noise effective voltage may be used for KMAX and KMIN.

[0083] Even when the number of uniform reference signals is n (three or more), the n coefficients k1 to kn can be determined by using the Lagrange interpolation method. As an example, consider the case where n=3. <r3>Maximum exposure conditions, <r1>is the minimum exposure condition. <r3> > <r2> > <r1>The third exposure condition that satisfies <r2>Then, we introduce constants KMAX, KMIN, and a third constant KMID that satisfies KMAX>KMID>KMIN. At this time, k1, k2, and k3 for signal strength S1 can be determined by the second-order Lagrange interpolation method as follows:

[0084] k1={(S1-R2)(S1-R3) / (R1-R2) / (R1-R3)}KMAX+ {(S1-R1)(S1-R3) / (R2-R1) / (R2-R3)}KMID+ {(S1-R1)(S1-R2) / (R3-R1) / (R3-R2)}KMIN k2={(S1-R2)(S1-R3) / (R1-R2) / (R1-R3)}KMIN+ {(S1-R1)(S1-R3) / (R2-R1) / (R2-R3)}KMAX+ {(S1-R1)(S1-R2) / (R3-R1)(R3-R2)}KMIN k3={(S1-R2)(S1-R3) / (R1-R2) / (R1-R3)}KMIN+ {(S1-R1)(S1-R3) / (R2-R1) / (R2-R3)}KMID+ {(S1-R1)(S1-R2) / (R3-R1) / (R3-R2)}KMAX

[0085] For example, the main controller 18 may <r1> 、 <r2>, R1, R2, Rd, KMAX, and KMIN may be stored in advance. Main control device 18 can calculate weighting coefficients k1 and k2 from these values ​​and measurement signal Sx from pixel 13, and can also calculate correction coefficient f. As described in the first embodiment, main control device 18 may measure the reference light and obtain Rl1 and Rl2 of each pixel 13.

[0086] Fig. 13 shows the relationship between the irradiation light intensity and the noise effective voltage. In the graph of Fig. 13, the horizontal axis shows the irradiation light intensity of the reference light, and the vertical axis shows the noise effective voltage (V). A dashed line 551 shows the noise after the correction process of the related art. A solid line 552 shows the noise after the correction process of the first embodiment. A solid line 553 shows the noise after the correction process of the second embodiment. In this correction process, KMAX is 1.52, and KMIN is 0.24.

[0087] As shown in FIG. 13, the correction processing result of the second embodiment shows a more stable noise reduction effect over a wider range of irradiation intensity compared to the correction processing result of the first embodiment.

[0088] <Other embodiments> FIG. 14 shows an example of photoelectric conversion characteristics of a pixel including a photodiode. FIG. 15 shows an example of noise voltage characteristics of the pixel. As shown in FIG. 14, the output voltage decreases with increasing incident light intensity. As shown in FIG. 15, the noise voltage increases with increasing incident light intensity. In other words, it can be confirmed that multiplicative noise exists. As described above, the signal correction method of the embodiment of this specification can effectively suppress multiplicative noise in addition to additive noise.

[0089] In the following, several devices to which the correction process of the measurement signal according to the embodiments of the present specification can be applied are described. The correction process according to the embodiments of the present specification can be applied to a device including a plurality of pixels, each including a photoelectric conversion element. The pixel layout may be one-dimensional or two-dimensional.

[0090] 16 is a schematic diagram showing a cross-sectional structure of an organic light emitting diode (OLED) panel with a sensor. The OLED panel with a sensor can detect the fingerprint of a finger 615. The OLED panel with a sensor includes a photosensor array 602 on a laminate film 601 that is a substrate. The photosensor array 602 includes a plurality of PIN diodes 603, and each PIN diode 603 is included in each pixel. The noise reduction process of the above embodiment can be applied to the measurement signal of the photosensor array 602.

[0091] A laminate film 605 is adhered onto the optical sensor array 602 by an OCA (Optical Clear Adhesive) 604. An array of pinholes 606, a TFT array 607, and a plurality of light-emitting layers 608 are laminated onto the laminate film 601. The light-emitting layer 608 is an OLED light-emitting layer, and the TFT array 607 controls the light emission of the light-emitting layer 608.

[0092] The light-emitting layer 608 is covered with a thin-film sealing layer 609, on which a polarizing plate 610 is laminated. A cover glass 612 is adhered onto the polarizing plate 610 by an OCA 611. A finger 615 is pressed against the surface of the cover glass 612.

[0093] Fig. 17 shows a schematic diagram of an example of the cross-sectional structure of an optical sensor array 602. The substrate SUB is the laminate film 601 shown in Fig. 16. Fig. 17 shows two TFTs and one PIN diode as an example.

[0094] An underlayer UC is formed on a substrate SUB, and a bottom gate electrode BG is disposed thereon. In FIG. 17, one of the bottom gate electrodes of two TFTs is indicated by the symbol BG as an example. The bottom gate electrode BG is covered with a gate insulating layer GI2. A semiconductor layer OX is formed on the gate insulating layer GI2. In the example of FIG. 17, the semiconductor layer OX is formed of an oxide semiconductor. Note that the semiconductor material is arbitrary. In FIG. 17, one of the semiconductor layers of two TFTs is indicated by the symbol OX as an example.

[0095] A gate insulating layer GI1 covers the semiconductor layer OX. A metal layer including a top gate electrode TG is formed on the gate insulating layer GI1. In FIG. 17, one of the top gate electrodes of the two TFTs is indicated by the symbol TG as an example.

[0096] The metal layer including the top gate electrode TG is covered with an interlayer insulating layer ILD. A metal layer M2 is formed on the interlayer insulating layer ILD. The metal layer M2 includes a source electrode and a drain electrode of the TFT, as well as a contact portion penetrating the insulating layer ILD or the insulating layer ILD and GI1.

[0097] The metal layer M2 is covered with a passivation layer PV1. A metal layer M3 is formed on the passivation layer PV1. The metal layer M3 includes a wiring layer and a contact portion that penetrates the insulating layer PV1.

[0098] The metal layer M3 is covered with a planarization layer PLN. A metal layer 4 is formed on the planarization layer PLN. The metal layer M4 includes a contact portion penetrating the insulating layer PLN in addition to the lower electrode of the PIN diode.

[0099] A part of the metal layer 4 is covered with a passivation layer PV2 and a passivation layer PV3 thereon. In an opening formed in the passivation layers PV2 and PV3, a semiconductor laminate PIN is formed on the lower electrode of the metal layer 4. The semiconductor laminate is composed of a P-type semiconductor layer, an N-type semiconductor layer, and an intrinsic semiconductor layer therebetween.

[0100] An upper electrode ITO is disposed on the semiconductor stack PIN. The upper electrode ITO is transparent to the light to be detected. A common electrode COM is connected to the upper electrodes ITO of the PIN diodes. The common electrode COM is connected to all the PIN diodes. A top passivation layer PV4 covers the upper electrode ITO and the common electrode COM.

[0101] FIG. 18 shows a schematic cross-sectional structure of an X-ray sensor panel. The X-ray sensor panel can be used in radiography devices in the fields of medical and industrial non-destructive testing. The X-ray sensor panel includes an optical sensor array 652 on a glass substrate 651. The optical sensor array 652 includes a plurality of PIN diodes 653, and each PIN diode 603 is included in each pixel. The noise reduction process of the above embodiment can be applied to the measurement signal of the optical sensor array 602.

[0102] The optical sensor array 652 is covered with a protective film 654. An X-ray conversion film (scintillator) 655 is laminated on the protective film. The X-ray conversion film 655 converts incident X-rays 671 into visible light 672 that can be detected by the PIN diode 653. Each pixel of the optical sensor array 652 measures the intensity of the visible light converted by the X-ray conversion film 655.

[0103] Although the embodiments of the present disclosure have been described above, the present disclosure is not limited to the above embodiments. A person skilled in the art can easily change, add, or convert each element of the above embodiments within the scope of the present disclosure. It is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. [Explanation of symbols]

[0104] 10 Image Sensor 11 Sensor board 13 pixels 14 Drive circuit 16 Signal detection circuit 18 Main Control Unit 201 Processor 202 Memory 203 Auxiliary storage device 204 Output Device 205 Input Device 207 Communication Interface 208 AD conversion interface < / r1> < / r2> < / r3> < / r1> < / r1> < / r1> < / rm> < / rm> < / rm> < / rn> < / r1>

Claims

1. A sensor device, A plurality of pixels; A control device that corrects the measurement signals from the plurality of pixels; Including, Each pixel of the plurality of pixels is A photodetector; a pixel circuit that outputs a signal from the photodetector; Including, The control device includes: obtaining an unknown measurement signal for a pixel of the plurality of pixels; correcting the unknown measurement signal from the one pixel using a correction factor based on a ratio between a statistical value of the measurement signals from the plurality of pixels and a value obtained from the measurement signal from the one pixel for each of a plurality of reference beams having different intensities. Sensor device.

2. The sensor device according to claim 1 , the correction coefficient is given by a linear combination of ratios of the reference light beams having the plurality of different intensities, with the weights assigned to the respective reference light beams having the plurality of different intensities being used as coefficients; Sensor device.

3. The sensor device according to claim 1 , The statistical value of the measurement signals of the plurality of pixels is an average value of values ​​obtained by subtracting the measurement signals of the plurality of pixels in a dark state from the measurement signals of the plurality of pixels. Sensor device.

4. The sensor device according to claim 1 , the control device determines the correction coefficient based on a statistical value of the measurement signals of the plurality of pixels when the reference beams having the different intensities are used, and a value obtained from the measurement signal of the one pixel when the reference beams having the different intensities are used. Sensor device.

5. The sensor device according to claim 4, The control device, in determining the correction coefficient, calculating the statistical value from a value obtained by subtracting a measurement signal in a dark state of each of the plurality of pixels from a measurement signal of each of the plurality of pixels for each of the plurality of reference beams having different intensities; subtracting a measurement signal of the one pixel in a dark state from a measurement signal of the one pixel in each of the plurality of reference beams having different intensities to obtain a first subtraction value; Calculating a linear combination of a value obtained by dividing the statistical value for each of the plurality of reference light beams having different intensities by the first subtraction value and a weighting coefficient for each of the plurality of reference light beams having different intensities; In the correction of the unknown measurement signal by the one pixel, calculating a second subtraction value by subtracting the measurement signal of the one pixel in the dark state from the unknown measurement signal of the one pixel; calculating the product of the second subtraction value and the linear combination; Sensor device.

6. The sensor device according to claim 5, The control device includes: determining the weighting factor for each of the plurality of reference beams having different intensities based on the statistics for each of the plurality of reference beams having different intensities and the unknown measurement signal for the one pixel; Sensor device.

7. The sensor device according to claim 1 , The control device includes: determining the correction factor based on the ratio for each of the plurality of reference beams having different intensities and the unknown measurement signal for the one pixel; Sensor device.

8. The sensor device according to claim 1 , The correction factor f can be expressed by the following formula: f=Σkm*<Rm> / Rm m represents an identifier of the plurality of reference beams having different intensities, km denotes the weight assigned to the reference beam m, Rm represents a value obtained from a measurement signal of one pixel in reference light m, <Rm> indicates the statistical value for reference light m; Sensor device.

9. The sensor device according to claim 8, The corrected value of the unknown measurement signal S^ can be expressed as: So^=(Sx-Rd)*f Sx is the unknown measurement signal; Rd is the measured signal of one pixel in the dark state; The value R obtained from the measurement signal of one pixel in the reference light m can be expressed by the following formula: Rm = Rlm - Rd Rlm is the measurement signal of the one pixel in the reference light m, The statistical value <Rm> for the reference light m is an average value of values ​​obtained by subtracting a measurement signal in a dark state from a measurement signal for the reference light m for the plurality of pixels. Sensor device.

10. 1. A method for determining correction factors to be used in the correction of measurement signals of pixels by a control device of a sensor device, comprising: Each pixel of the plurality of pixels of the sensor device is A photodetector; a pixel circuit that outputs a signal from the photodetector; Including, The method comprises: acquiring a measurement signal for each of the plurality of pixels for each of the plurality of reference beams having different intensities; determining a statistical value of the measurement signals of the plurality of pixels for each of the plurality of reference beams having different intensities; determining a correction coefficient for each of the plurality of pixels based on a statistical value of the measurement signals of the plurality of pixels and the measurement signal of each of the plurality of pixels for each of the plurality of reference beams having different intensities; method.

Images