Solid-state imaging device, and imaging apparatus including the same

Abstract

A solid-state imaging device in which a plurality of readouts are performed with different readout gains for the same exposure signal, and correlated multiple sampling is performed in the readout with the lowest readout gain in the plurality of readouts.

Description

Technical Field

[0001] The present invention relates to a solid-state imaging device and an imaging apparatus including the solid-state imaging device.

Background Art

[0002] Conventionally, among solid-state imaging devices such as CCD (Charge Coupled Device) and CMOS (Complementary Metal Oxide Semiconductor) image sensors used in imaging apparatuses having an imaging function such as digital cameras, digital video cameras, and smartphones, for signals of the same exposure, a so-called dual conversion gain type solid-state imaging device that realizes a high dynamic range by performing reading at a high readout gain and reading at a low readout gain is known (US7075049B2).

[0003] In recent years, since a wider dynamic range (high dynamic range) has been demanded, in the above-described dual conversion gain type solid-state imaging device, it is conceivable to increase the ratio (gain ratio) of the high readout gain to the low readout gain.

[0004] However, when the ratio of the high readout gain to the low readout gain is increased, at the boundary between the signal obtained by reading at the high readout gain and the signal obtained by reading at the low readout gain (the boundary of the illuminance from the subject), an SNR step (SNR dip) occurs (see FIG. 5). For example, when imaging a subject whose brightness changes gently (such as a colorless wall or sky), due to the difference in the amount of dispersion of the signal, a false boundary occurs in the region corresponding to the SNR step in the captured image, and the quality of the captured image deteriorates.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] Therefore, an object of the present invention is to provide a solid-state imaging device in which a plurality of readings are performed with different readout gains for signals of the same exposure, and which can realize a wider dynamic range while suppressing image degradation, and an imaging device including the solid-state imaging device.

Means for Solving the Problems

[0007] The solid-state imaging device according to the present invention is

[0008] a solid-state imaging device in which a plurality of readings are performed with different readout gains for signals of the same exposure, correlated multiple sampling is performed in the reading at the lowest readout gain in the plurality of readings.

[0009] Further, in the solid-state imaging device, the different readout gains include the lowest readout gain and a high readout gain that is higher than the lowest readout gain, the correlated multiple sampling may be performed in the reading at the high readout gain.

[0010] Further, in the solid-state imaging device, the reading in which the correlated multiple sampling is performed may be configured to be switchable between the reading at the lowest readout gain and the reading at a high readout gain that is higher than the lowest readout gain.

[0011] Further, in the solid-state imaging device, the reading at the same gain includes the reading of the reset voltage and the reading of the signal voltage, In the correlated multiple sampling, the number of sampling times for reading the reset voltage and the number of sampling times for reading the signal voltage may be the same.

[0012] Further, in the solid-state imaging device, The reading at the same gain includes the reading of the reset voltage and the reading of the signal voltage. In the correlated multiple sampling, the number of sampling times for reading the signal voltage may be more than the number of sampling times for reading the reset voltage.

[0013] Further, the solid-state imaging device according to the present invention is A solid-state imaging device in which a signal of the same exposure is read a plurality of times with different read gains, Correlated multiple sampling is performed in at least one of the readings excluding the reading at the highest read gain among the plurality of readings.

[0014] Further, in the solid-state imaging device, The correlated multiple sampling may be performed in each of the readings excluding the reading at the highest read gain among the plurality of readings.

[0015] Further, the solid-state imaging device according to the present invention is A solid-state imaging device in which a signal of the same exposure is read a plurality of times with different read gains, Correlated multiple sampling is performed in at least one of the plurality of readings.

[0016] Further, the imaging device according to the present invention is An optical system,

[0017] And any one of the above solid-state imaging devices disposed at the imaging position of the optical system.

[0018] Further, the imaging device is It includes an exposure control unit connected to the solid-state imaging device, In the solid-state imaging device, a plurality of combinations of readouts in which the correlation multiple sampling is performed are set, The exposure control unit may instruct the solid-state imaging device to select the optimal combination from among the plurality of combinations during imaging.

Brief Description of the Drawings

[0019]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

[0020] The solid-state imaging device according to the present invention is a solid-state imaging device in which a signal of the same exposure is read multiple times with different read gains, and correlated multiple sampling is performed in the reading at the lowest read gain among the multiple readings.

[0021] According to such a configuration, since correlated multiple sampling is performed at least for the reading at the lowest read gain, the SNR step at the boundary (the boundary of the illuminance from the subject) between the signal obtained by the reading at the lowest read gain and the signal obtained by the reading at the next lowest read gain is suppressed (see, for example, FIGS. 6 and 9). Thereby, it is possible to suppress the deterioration of the image quality of the captured image while securing a wider dynamic range.

[0022] Further, in the solid-state imaging device, the different read gains include the lowest read gain and a high read gain that is higher than the lowest read gain, and the correlated multiple sampling may be performed in the reading at the high read gain.

[0023] According to such a configuration, the influence of noise (such as random noise) in the signal read at the high read gain is suppressed, and the image quality of the captured image is improved.

[0024] Further, in the solid-state imaging device, the reading in which the correlated multiple sampling is performed may be configured to be switchable between the reading at the lowest read gain and the reading at a high read gain that is higher than the lowest read gain.

[0025] According to such a configuration, in a scene where a high dynamic range is required for imaging, by being switched so that correlated multiple sampling is performed during reading at the lowest readout gain, the SNR difference at the boundary (the boundary of the illuminance from the subject) between the signal obtained by reading at the lowest readout gain and the signal obtained by reading at a high readout gain is suppressed. As a result, it is possible to suppress the degradation of the image quality of the captured image while securing a wider dynamic range.

[0026] On the other hand, in a scene where a high dynamic range is not required for imaging, by being switched so that correlated multiple sampling is performed during reading at a high readout gain, the influence of noise (such as random noise) in the signal read at the high readout gain is suppressed. As a result, the image quality (image quality on the low illuminance side) of the captured image is improved.

[0027] Moreover, by adopting a configuration in which the reading at which correlated multiple sampling is performed is switched, regardless of at which gain the correlated multiple sampling is performed, the total reading time of the reading at the lowest readout gain and the reading at the high readout gain is the same or substantially the same. Therefore, the frame rate during video imaging before and after the switching can be made constant or substantially constant.

[0028] Also, in the solid-state imaging device, The reading at the same gain includes the reading of the reset voltage and the reading of the signal voltage. In the correlated multiple sampling, the number of sampling times of the reading of the reset voltage and the number of sampling times of the reading of the signal voltage may be the same.

[0029] In this way, by making the number of sampling times of the reading of the reset voltage and the number of sampling times of the reading of the signal voltage the same in the correlated multiple sampling, the influence of noise (such as random noise) in the read signal is preferably suppressed. As a result, the image quality of the captured image is further improved.

[0030] In the solid-state imaging device, the readout at the same gain includes the readout of the reset voltage and the readout of the signal voltage. In the correlated multiple sampling, the number of sampling times of the readout of the signal voltage may be larger than the number of sampling times of the readout of the reset voltage.

[0031] In this way, in the correlated multiple sampling, by increasing the number of sampling times of the readout of the signal voltage compared to the number of sampling times of the readout of the reset voltage, it is possible to suppress the readout time while improving the image quality of the captured image.

[0032] Also, the solid-state imaging device according to the present invention is a solid-state imaging device in which multiple readouts are performed on signals of the same exposure with different readout gains, and correlated multiple sampling is performed in at least one of the readouts excluding the readout at the highest readout gain among the multiple readouts.

[0033] According to such a configuration, at least one SNR step is suppressed, and thereby, it is possible to suppress the deterioration of the image quality of the captured image while ensuring a wider dynamic range.

[0034] Also, the solid-state imaging device according to the present invention is a solid-state imaging device in which multiple readouts are performed on signals of the same exposure with different readout gains, and correlated multiple sampling is performed in at least one of the multiple readouts.

[0035] According to such a configuration, at least one SNR step is suppressed, or the influence of noise (such as random noise) in the signal read out at the highest readout gain is suppressed. Therefore, it is possible to suppress the deterioration of the image quality of the captured image while ensuring a wider dynamic range.

[0036] Also, the imaging device according to the present invention is an optical system, and any one of the solid-state imaging devices disposed at the imaging position of the optical system.

[0037] According to such a configuration, in the obtained captured image, it is possible to suppress deterioration of image quality while securing a wider dynamic range.

[0038] Further, the imaging device includes an exposure control unit connected to the solid-state imaging device, in the solid-state imaging device, a plurality of combinations of readouts in which the correlation multiple sampling is performed are set, and the exposure control unit may instruct the solid-state imaging device to select the optimal combination from among the plurality of combinations at the time of imaging.

[0039] According to such a configuration, the readout in which the correlation multiple sampling is performed by the exposure control unit is automatically selected.

[0040] According to the above solid-state imaging device and the imaging device including the solid-state imaging device, in a solid-state imaging device in which a signal of the same exposure is read out a plurality of times with different readout gains, it is possible to suppress deterioration of an image while realizing a wider dynamic range.

[0041] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

[0042] As shown in FIG. 1, the imaging device includes an optical system 101a and a solid-state imaging device 1 disposed at the imaging position of the optical system 101a, and is capable of imaging an object such as a digital camera, a smartphone, or a tablet device. The imaging device 100 of the present embodiment is, for example, a smartphone.

[0043] The imaging device 100 includes an imaging unit 101 and a control unit 102. The imaging device 100 also includes a non-volatile memory 103, a working memory 104, an operation unit 105, a display unit 106, a recording medium 107, a connection unit 108, a short-range wireless communication unit 109, a public network connection unit 110, a microphone 111, and a speaker 112.

[0044] Under the control of the control unit 102, the imaging unit 101 converts the image of the subject formed by the optical system 101a included in the imaging unit 101 into an electrical signal, then performs noise reduction processing, etc., and outputs the digital data as the image data of the output image. Specifically, the imaging unit 101 includes an optical system 101a composed of at least one optical element such as a lens, and a solid-state imaging device 1 that converts the image of the subject (imaging object) formed through the optical system 101a into an electrical signal.

[0045] The solid-state imaging device 1 is, for example, incorporated in a smartphone, a digital camera, etc., and is an element for imaging a subject. The solid-state imaging device 1 of the present embodiment is, for example, a CMOS image sensor. In this solid-state imaging device 1, multiple readings are performed on the signal of the same exposure with different readout gains (voltage conversion gains), and correlated multiple sampling (CMS) is performed in the reading at the lowest readout gain in the multiple readings.

[0046] The different readout gains include the lowest readout gain and a high readout gain that is higher than the lowest readout gain. The solid-state imaging device 1 of the present embodiment is configured to be able to switch the reading for which correlated multiple sampling is performed between the reading at the lowest readout gain and the reading at the high readout gain.

[0047] Specifically, as shown in FIG. 2, the solid-state imaging device 1 includes a plurality of pixels (picture elements) 2 arranged in a matrix, a plurality of row signal lines 3 arranged for each row with respect to the matrix-shaped pixel array and each extending in the row direction (left-right direction in FIG. 2), a plurality of column signal lines 4 arranged for each column and each extending in the column direction (up-down direction in FIG. 2), and a signal processing unit 5 to which the plurality of row signal lines 3 and the plurality of column signal lines 4 are connected.

[0048] As shown in FIG. 3, each of the plurality of pixels 2 arranged in a matrix includes a photoelectric conversion element 21 such as a photodiode, a floating node 22 to which the signal charge generated by the photoelectric conversion element 21 is transferred, a holding capacitor 23 capable of holding the signal charge, a transfer transistor 24 connecting the photoelectric conversion element 21 and the floating node 22, a holding switch transistor 25 connecting the floating node 22 and the holding capacitor 23, a reset transistor 26 connecting the holding capacitor 23 and the reset power supply VDD1, a source follower SF having an amplification transistor 27 and amplifying the voltage signal of the floating node 22, and a selection transistor 28 connecting the amplification transistor 27 and the column signal line 4.

[0049] In the solid-state imaging device 1 of this embodiment, the read gain (voltage conversion gain / conversion gain) when the signal charge transferred from the photoelectric conversion element 21 to the floating node 22 in each pixel 2 is read out as a voltage signal through the source follower SF can be switched between a minimum read gain and a high read gain which is higher than the minimum read gain by switching the ON / OFF of the holding switch transistor 25. It is a solid-state imaging device 1 of a so-called dual conversion gain (DCG) method. The ratio (gain ratio) of the minimum read gain (Low conversion gain: LCG) to the high read gain (High conversion gain: HCG) in the solid-state imaging device 1 of this embodiment is, for example, 1:64. Note that the "readout" in the solid-state imaging device 1 of this embodiment may be either analog readout or readout after AD conversion.

[0050] In this solid-state imaging device 1, during readout at the same gain (LCG or HCG), reset voltage readout and signal voltage readout are respectively performed.

[0051] Specifically, for one exposure (imaging) of the solid-state imaging device 1, in each pixel 2, as readout at the same gain (LCG or HCG), readout of the reset voltage as a voltage signal in a state where the floating node 22 is reset, and readout of the signal voltage as a voltage signal in a state where the floating node 22 has signal charges transferred from the photoelectric conversion element 21 after the readout of the reset voltage are performed, and difference signals between these signal voltages and the reset voltage are respectively generated. That is, in the solid-state imaging device 1, so-called correlated double sampling (CDS) is performed in the readout from each pixel 2.

[0052] In the above solid-state imaging device 1, readout of the voltage signal at HCG and readout of the voltage signal at LCG in each pixel 2 arranged in a matrix are performed for each row. And in the readout of the voltage signal for each row, based on an instruction from the control unit 102, correlated multiple sampling (CMS) is performed either in the readout at HCG or in the readout at LCG. This CMS is one of the signal sampling methods, and by sampling, adding and averaging the reset level voltage and the signal level voltage a plurality of times respectively and taking the difference, it is possible to remove correlated noise components, reduce low-frequency noise and random noise due to the averaging effect.

[0053] Specifically, when the control unit 102 performs automatic exposure during imaging, for example, when it is determined that a certain area on the captured image is blacked out or overexposed, and it is judged that a high dynamic range is required for the scene, in the solid-state imaging device 1, the target of correlated multiple sampling is switched to readout in the LCG according to an instruction from the control unit (exposure control unit) 102. At this time, in the solid-state imaging device 1, the number of sampling times for reading the reset voltage in the LCG in correlated multiple sampling is the same as the number of sampling times for reading the signal voltage in the LCG. In the solid-state imaging device 1 of the present embodiment, the number of sampling times for reading the reset voltage in the LCG in correlated multiple sampling and the number of sampling times for reading the signal voltage in the LCG are both two times each.

[0054] More specifically, for the signals of the same exposure generated by the photoelectric conversion elements 21 of each pixel 2, in a plurality of pixels 2 arranged in a matrix, for each row, the readout of the reset voltage in the LCG (LCG Ref ADC1, 2 in FIG. 4) is performed two times (that is, the sampling of the reset voltage in correlated multiple sampling is two times), the readout of the reset voltage in the HCG (HCG Ref ADC in FIG. 4) is performed one time, the readout of the signal voltage in the HCG (HCG Sig ADC in FIG. 4) is performed one time, and the readout of the signal voltage in the LCG (LCG Sig ADC1, 2 in FIG. 4) is performed two times (that is, the sampling of the signal voltage in correlated multiple sampling is two times). Note that FIG. 4 is a diagram schematically showing the readout steps for each row, and the length in the left-right direction in each step corresponds to the length of the readout time. The same applies to FIGS. 7, 8, 10, and 11.

[0055] Then, in the solid-state imaging device 1, after the reset voltage and the signal voltage of the LCG that have been correlated multiple sampled, that is, sampled multiple times (two times in the example of the present embodiment), are averaged respectively, the generation of the differential signal between the averaged reset voltage and the averaged signal voltage (that is, correlated double sampling) is performed, and at the same time, the generation of the differential signal between the reset voltage and the signal voltage of the HCG (correlated double sampling) is performed.

[0056] At this time, in the readout by the LCG, since the readout gain is low, the influence of random noise and the like becomes large. However, by generating an image signal using a signal obtained by sampling a plurality of times by correlation multiple sampling and averaging, compared with the case where correlation multiple sampling is not performed in the readout by the LCG, the influence of noise such as random noise in the readout by the LCG can be suppressed. As a result, the SNR difference (SNR dip) α (see FIG. 6) at the boundary between the signal obtained by the readout by the LCG and the signal obtained by the readout by the HCG (the boundary of the illuminance from the subject) can be reduced. In FIG. 6, the graph on the right side of the SNR difference α shows the readout signal by the LCG, and the graph on the left side of the SNR difference α shows the readout signal by the HCG.

[0057] For example, in the solid-state imaging device 1 of the present embodiment (when the number of sampling times of the reset voltage and the signal voltage in the LCG is 2 times each: see FIG. 6), compared with the case where correlation multiple sampling is not performed (see FIG. 5), the SNR difference α is suppressed from 6 dB to 4 dB. Note that FIGS. 5 and 6 are graphs with the vertical axis being SNR [dB] and the horizontal axis being the illuminance [lux] from the subject. The same applies to FIG. 9.

[0058] On the other hand, when the control unit 102 determines that a high dynamic range is not required in imaging, in the solid-state imaging device 1, the target of correlation multiple sampling is switched to the readout by the HCG according to an instruction from the control unit (exposure control unit) 102. At this time, in the solid-state imaging device 1, the number of sampling times of the readout of the reset voltage in the HCG in the correlation multiple sampling is the same as the number of sampling times of the readout of the signal voltage in the HCG in the correlation multiple sampling, similar to the case where correlation multiple sampling is performed in the readout by the LCG. In the solid-state imaging device 1 of the present embodiment, the number of sampling times of the readout of the reset voltage in the HCG in the correlation multiple sampling and the number of sampling times of the readout of the signal voltage in the HCG in the correlation multiple sampling are both 2 times.

[0059] More specifically, for the signals of the same exposure generated by the photoelectric conversion element 21 of each pixel 2, in a plurality of pixels 2 arranged in a matrix, for each row, the reset voltage is read once with the LCG, and the reset voltage is read twice with the HCG (that is, the reset voltage is sampled twice in correlated multiple sampling), the signal voltage is read twice with the HCG (that is, the signal voltage is sampled twice in correlated multiple sampling), and the signal voltage is read once with the LCG (see FIG. 7).

[0060] In the solid-state imaging device 1, correlated double sampling of the reset voltage and the signal voltage of the LCG is performed, and after the reset voltage and the signal voltage of the HCG that have been correlated multiple sampled, that is, sampled a plurality of times (in the example of this embodiment, twice), are averaged respectively, correlated double sampling of the averaged reset voltage and the averaged signal voltage is performed.

[0061] At this time, by generating an image signal using the signal that has been sampled a plurality of times and averaged by correlated multiple sampling in the reading with the HCG, compared with the case where correlated multiple sampling is not performed in the reading with the HCG, the influence of noise such as random noise in the reading with the HCG is suppressed, and thereby, the image quality on the side with low illuminance in the reading signal with the HCG in FIG. 6 is improved.

[0062] Returning to FIG. 1, the control unit 102 controls each part of the imaging device 100 according to the input signals and programs. Further, the control unit 102 generates an imaging image (image data) from the signals output from the solid-state imaging device 1 and outputs it to the display unit 106. Further, the control unit 102 of this embodiment has an exposure control unit 102A that instructs whether to perform correlated multiple sampling on the solid-state imaging device 1 in either the reading with the LCG or the reading with the HCG.

[0063] The non-volatile memory 103 is a non-volatile memory that can be electrically erased and recorded. The non-volatile memory 103 of the present embodiment records an OS (operating system), which is basic software executed by the control unit 102, and an application that realizes application functions in cooperation with this OS.

[0064] The working memory 104 is used as an image display memory for the display unit 106 and a working area of the control unit 102 and the like.

[0065] The operation unit 105 is used for a user or the like to input an instruction to the imaging device 100. The operation unit 105 of the present embodiment includes a power button for instructing ON / OFF of the power supply of the imaging device 100 and a touch panel formed on the display unit 106.

[0066] The display unit 106 performs display (output to the outside) of the captured image (image data) and character display for operation.

[0067] The recording medium 107 records the image data output from the imaging unit 101.

[0068] The connection unit 108 is an interface for connecting to an external device. Through this connection unit 108, the imaging device 100 exchanges data with the external device.

[0069] The short-range wireless communication unit 109 is a communication unit for performing short-range wireless communication. The short-range wireless communication unit 109 is composed of an antenna for wireless communication and a modulation / demodulation circuit and a communication controller for processing wireless signals.

[0070] The public network connection unit 110 is an interface for performing public wireless communication. Through this public network connection unit 110, the imaging device 100 conducts communication for making calls with other devices. At this time, the control unit 102 realizes the call by inputting and outputting audio signals via the microphone 111 and the speaker 112. The public network connection unit 110 of the present embodiment is an antenna, and the control unit 102 connects to the public network via this antenna.

[0071] The solid-state imaging device 1 of the imaging device 100 described above is a solid-state imaging device 1 in which a signal of the same exposure is read multiple times with different readout gains, and correlated multiple sampling (CMS) is performed in the readout with the lowest readout gain (LCG) in the multiple readouts.

[0072] Thus, since correlated multiple sampling is performed at least for the readout with LCG, the SNR step α at the boundary (the boundary of the illuminance from the subject) between the signal obtained by the readout with LCG and the signal obtained by the readout with the readout gain lower than the LCG (HCG in the example of the present embodiment) is suppressed (see the symbol α in FIGS. 5 and 6). As a result, it is possible to suppress the deterioration of the image quality of the captured image while ensuring a wider dynamic range (HDR).

[0073] In addition, in the solid-state imaging device 1 of the present embodiment, the different readout gains include the lowest readout gain (LCG) and a high readout gain (HCG) that is higher than the LCG, and correlated multiple sampling (CMS) is performed in the readout with HCG.

[0074] According to such a configuration, the influence of noise (such as random noise) in the signal read with HCG is suppressed, and the image quality is improved.

[0075] In addition, in the solid-state imaging device 1 of the present embodiment, the readout in which correlated multiple sampling (CMS) is performed can be switched between the readout at the lowest readout gain (LCG) and the readout at a high readout gain (HCG) that is higher than the LCG.

[0076] According to such a configuration, in a scene where a high dynamic range is required in imaging, by switching so that correlated multiple sampling is performed in the readout at the LCG, the SNR difference α at the boundary (the boundary of the illuminance from the subject) between the signal obtained by the readout at the LCG and the signal obtained by the readout at the HCG can be suppressed. As a result, it is possible to suppress the degradation of the image quality of the captured image while ensuring a wider dynamic range.

[0077] On the other hand, in a scene where a high dynamic range is not required in imaging, by switching so that correlated multiple sampling is performed in the readout at the HCG, the influence of noise (such as random noise) in the signal read out at the HCG can be suppressed. As a result, the image quality (image quality on the side with low illuminance) of the captured image is improved.

[0078] Moreover, by adopting a configuration in which the readout in which correlated multiple sampling is performed is switched, no matter at which readout gain (voltage conversion gain) the correlated multiple sampling is performed, the total readout time of the readout at the LCG and the readout at the HCG is the same or substantially the same. Therefore, the frame rate during video imaging before and after the switching can be made constant or substantially constant.

[0079] In addition, in the solid-state imaging device 1 of the present embodiment, the readout at the same gain (the same readout gain) includes the readout of the reset voltage and the readout of the signal voltage, and in correlated multiple sampling (CMS), the number of sampling times of the readout of the reset voltage and the number of sampling times of the readout of the signal voltage are the same.

[0080] Thus, in correlated multiple sampling, by making the number of sampling times for reading the reset voltage the same as the number of sampling times for reading the signal voltage, the influence of noise (such as random noise) in the read signal is preferably suppressed, and thereby, the image quality of the captured image is further improved.

[0081] Further, the imaging device 100 of the present embodiment includes an optical system 101a and a solid-state imaging device 1 disposed at the imaging position of the optical system 101a.

[0082] According to this imaging device 100, in the obtained captured image, it is possible to suppress the deterioration of the image quality while ensuring a wider dynamic range.

[0083] Note that the solid-state imaging device 1 and the imaging device 100 including the solid-state imaging device 1 of the present invention are not limited to the above-described embodiment, and it goes without saying that various changes can be made without departing from the gist of the present invention. For example, the configuration of another embodiment can be added to the configuration of a certain embodiment, and a part of the configuration of a certain embodiment can be replaced with the configuration of another embodiment. Further, a part of the configuration of a certain embodiment can be deleted.

[0084] The number of sampling times of the correlated multiple sampling in the solid-state imaging device 1 of the above embodiment is 2 times each for reading the reset voltage and reading the signal voltage, but is not limited to this configuration. The number of sampling times for reading the reset voltage in the LCG and the number of sampling times for reading the signal voltage in the LCG may be 3 times or more.

[0085] The larger the number of sampling times for reading the reset voltage in the LCG and the number of sampling times for reading the signal voltage in the LCG, the smaller the SNR difference α at the boundary between the signal obtained by reading in the LCG and the signal obtained by reading in the HCG (the boundary of the illuminance from the subject). For example, as shown in FIG. 8, when the number of sampling times for reading the reset voltage in the LCG and the number of sampling times for reading the signal voltage in the LCG are each 4 times, as shown in FIG. 9, the SNR difference α at the boundary between the signal obtained by reading in the LCG and the signal obtained by reading in the HCG is suppressed to 2 dB.

[0086] In addition, in the solid-state imaging device 1 of the above embodiment, in correlated multiple sampling, the number of sampling times for reading the reset voltage with the same gain and the number of sampling times for reading the signal voltage are the same, but the configuration is not limited to this. In correlated multiple sampling, the number of sampling times for reading the reset voltage with the same gain and the number of sampling times for reading the signal voltage may be different.

[0087] In this case, as shown in FIG. 10, in correlated multiple sampling, it is preferable that the number of sampling times for reading the signal voltage is larger than the number of sampling times for reading the reset voltage with the same gain. In this way, in correlated multiple sampling, by increasing the number of sampling times for reading the signal voltage compared to the number of sampling times for reading the reset voltage, it is possible to suppress the readout time in the solid-state imaging device 1 while improving the image quality of the captured image (to a certain extent).

[0088] In addition, in the solid-state imaging device 1 of the above embodiment, correlated multiple sampling is performed in reading with either the LCG or the HCG gain, but the configuration is not limited to this. Correlated multiple sampling may be performed in reading with each gain. For example, as shown in FIG. 11, correlated multiple sampling may be performed in each of the reading in the LCG and the reading in the HCG.

[0089] In addition, in the solid-state imaging device 1 of the above-described embodiment, the readout in which correlated multiple sampling is performed is switched between readouts with different gains, but the configuration is not limited to this. The readout in which correlated multiple sampling is performed may be fixed to, for example, the readout in LCG or the like.

[0090] In addition, in the solid-state imaging device 1 of the above-described embodiment, the number of readout gains when reading the voltage signal is two (LCG and HCG), but it may be three or more.

[0091] In this case, correlated multiple sampling may be performed in the readout at the lowest readout gain in multiple readouts. According to such a configuration, since correlated multiple sampling is performed at least for the readout at the lowest readout gain, the SNR step at the boundary portion (the boundary portion of the illuminance from the subject) between the signal obtained by the readout at the lowest readout gain and the signal obtained by the readout at the readout gain second lowest to the lowest readout gain is suppressed, and thereby, it is possible to suppress the deterioration of the image quality while securing a wider dynamic range.

[0092] In addition, correlated multiple sampling may be performed in at least one readout excluding the readout at the highest readout gain in multiple readouts. According to such a configuration, at least one SNR step is suppressed, and thereby, it is possible to suppress the deterioration of the image quality while securing a wider dynamic range.

[0093] In addition, correlated multiple sampling may be performed in at least one readout in multiple readouts. According to such a configuration, at least one SNR step is suppressed, or the influence of noise (such as random noise) in the signal read out at the highest readout gain is suppressed, and thereby, it is possible to suppress the deterioration of the image quality while securing a wider dynamic range.

[0094] In addition, the ratio (gain ratio) of the LCG to the HCG in the solid-state imaging device 1 of the above embodiment is, for example, 1:64, but is not limited to this configuration. For example, the gain ratio may be other gain ratios such as 1:4 or 1:8. Note that since the SNR step α increases as the gain ratio increases, significant effects can be obtained in suppressing the SNR step α by performing correlated multiple sampling.

[0095] In addition, the imaging device 100 includes an exposure control unit 102A connected to the solid-state imaging device 1. In the solid-state imaging device 1, a plurality of readout combinations for performing correlated multiple sampling are set, and the exposure control unit 102A may be configured to instruct the solid-state imaging device 1 to select an optimal combination from among the plurality of combinations during imaging.

[0096] As an example of switching the readout combination, when the SNR step is large, such as when the absolute value of the gain of the LCG is small and the gain ratio of the LCG to the HCG is large (for example, 1:32 or more), a combination in which the number of sampling times of the correlated multiple sampling of the LCG is larger than that of the HCG, and when high dynamic range imaging is performed in a dark scene where the absolute value of the gain of the LCG is large (for example, 4 times or more) and it is desired to improve the SNR in a dark area, a combination in which the number of sampling times of the correlated multiple sampling of the HCG is larger than that of the LCG, and the above two types of combinations are switched based on an instruction. That is, the exposure control unit 102A selects, as the optimal combination, the combination that provides the best image quality from among the plurality of set combinations from the viewpoint of suppressing deterioration of the image due to a wide dynamic range and the SNR step.

[0097] According to such a configuration, the readout for performing correlated multiple sampling is automatically selected by the exposure control unit 102A.

[0098] In order to represent the present invention, the present invention has been appropriately and fully described through embodiments with reference to the drawings above. However, those skilled in the art should recognize that it is easy to make changes and / or improvements to the above-described embodiments. Therefore, as long as the changes or improvements made by those skilled in the art do not depart from the scope of the claims described in the claims, such changes or improvements are construed as being included in the scope of the claims of the claims.

Explanation of Reference Numerals

[0099] 1... Solid-state imaging device, 2... Pixel, 21... Photoelectric conversion element, 22... Floating node, 23... Holding capacitor, 24... Transfer transistor, 25... Holding switch transistor, 26... Reset transistor, 27... Amplification transistor, 28... Selection transistor, 3... Row signal line, 4... Column signal line, 5... Signal processing unit, 100... Imaging device, 101... Imaging unit, 101a... Optical system, 102... Control unit, 102A... Exposure control unit, 103... Non-volatile memory, 104... Working memory, 105... Operation unit, 106... Display unit, 107... Recording medium, 108... Connection unit, 109... Short-range wireless communication unit, 110... Public network connection unit, 111... Microphone, 112... Speaker, SF... Source follower, α... SNR step (SNR dip)

Claims

1. A solid-state imaging device in which multiple readouts are performed with different readout gains for the same exposure signal, wherein correlated multiple sampling is performed in the readout with the lowest readout gain among the multiple readouts.

2. The different readout gains include the lowest readout gain and a high readout gain that is higher than the lowest readout gain, and the solid-state imaging device according to claim 1, wherein the correlated multiple sampling is performed in the readout with the high readout gain.

3. The solid-state imaging device according to claim 1 or 2, wherein the readout in which the correlated multiple sampling is performed is configured to be switchable between the readout with the lowest readout gain and the readout with a high readout gain that is higher than the lowest readout gain.

4. The readout with the same gain includes the readout of the reset voltage and the readout of the signal voltage, and in the correlated multiple sampling, the number of sampling times of the readout of the reset voltage is the same as the number of sampling times of the readout of the signal voltage, for the solid-state imaging device according to any one of claims 1 to 3.

5. The readout with the same gain includes the readout of the reset voltage and the readout of the signal voltage, and in the correlated multiple sampling, the number of sampling times of the readout of the signal voltage is larger than the number of sampling times of the readout of the reset voltage, for the solid-state imaging device according to any one of claims 1 to 3.

6. A solid-state imaging device in which multiple readouts are performed with different readout gains for the same exposure signal, wherein correlated multiple sampling is performed in at least one readout excluding the readout with the highest readout gain among the multiple readouts.

7. The solid-state imaging device according to claim 6, wherein the correlated multiple sampling is performed in each readout excluding the readout with the highest readout gain among the multiple readouts.

8. A solid-state imaging device in which multiple readouts are performed with different readout gains for the same exposure signal, wherein correlated multiple sampling is performed in at least one readout among the multiple readouts.

9. An optical system, An imaging device comprising: the solid-state imaging device according to any one of claims 1 to 8, disposed at an imaging position of the optical system.

10. Comprising an exposure control unit connected to the solid-state imaging device, In the solid-state imaging device, a plurality of combinations of readouts in which the correlated multiple sampling is performed are set, The exposure control unit instructs the solid-state imaging device to select an optimal combination from among the plurality of combinations during imaging. The imaging device according to claim 9.

Images