JP2008199419A - Solid-state imaging apparatus, driving method thereof and camera - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a solid-state imaging apparatus for suppressing image quality failures by preventing occurrence of smears in a moving picture mode. <P>SOLUTION: A timing generation circuit 20 outputs a driving signal so as to generate a first mixed packet and a second mixed packet in a horizontal transfer part 14. In the first mixed packet, two or more signal packets of the same color in the same column are mixed in a prescribed stage of the horizontal transfer part, and in the second mixed packet, two or more signal packets and two or more dummy packets of the same color in the same column are mixed, in a stage other than the prescribed stage of the horizontal transfer part. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a solid-state imaging device, a solid-state imaging device driving method, and a camera that have a plurality of light receiving elements, a plurality of vertical transfer units, and a horizontal transfer unit that are two-dimensionally arranged and output as image signals. In particular, the present invention relates to a solid-state imaging device having a still image capturing mode and a moving image capturing mode, a method for driving a solid-state image sensor, and a camera.

  Conventionally, a solid-state imaging device that converts received light into an electrical signal and outputs it as a video signal is known, and a camera such as a digital still camera that displays a video signal obtained from the solid-state imaging device as a still image is known. ing. In recent years, a camera using such a solid-state imaging device has been required to further improve image quality and function, and the density of pixels has been increasing.

  In such a solid-state imaging device, a driving method for reducing the number of pixels in an output video signal by thinning out pixels from which signal charges are read out is particularly disclosed in Patent Documents 1 and 2, for example, in order to improve output speed in a monitor mode or the like. It is disclosed.

  However, in such thinning driving, noise components such as smear and dark current become a problem. This will be described with reference to FIG.

  FIG. 11 is a schematic diagram for explaining the problems of the prior art disclosed in Patent Document 3. In the schematic diagram of FIG. 11, a white circle (◯) indicates a noise component on the vertical transfer unit 13, and a black circle (●) indicates a signal component including a noise component transferred once in the horizontal blanking period. Each is shown.

  In the all-pixel readout drive used in the still image capturing mode or the like, as shown in FIG. 11A, the signal charges of all the pixels (indicated by black squares in the figure) are simultaneously transmitted to the vertical transfer unit 13 at the same time. The signal is transferred in the vertical direction without mixing in the vertical transfer unit 101, and the signal charge is transferred in the horizontal direction by the horizontal transfer unit 14 and read out through the charge detection unit 15. .

  In this all-pixel readout drive (a), the number of transfer pixels in one vertical transfer unit per horizontal blanking period is one pixel. Further, the number of signal pixels output at one time is one pixel, and the amount of noise output at one time is one vertical transfer unit (the amount of noise for one packet). Here, it is assumed that the mechanical shutter is not used for all pixel readout.

  On the other hand, in the thinning readout drive used in the monitor mode or the like, as shown in FIG. 11B, when thinning out every other pixel in the vertical direction, for example, odd rows (1, 3, 5,. Only the signal charge of the pixel of () is read out to the vertical transfer unit 13. At this time, an empty packet exists in the vertical transfer unit 13. Then, the signal packet and the empty packet are paired and transferred in the vertical direction, and the charges for the two packets are mixed in the horizontal transfer unit 14 and transferred in the horizontal direction, and read out through the charge detection unit 15. ing.

  In this thinning readout drive (b), the number of transfer pixels in one vertical transfer unit per horizontal blanking period is two pixels. Further, the number of signal pixels output at one time is one pixel, and the amount of noise output at one time is two vertical transfer units (the amount of noise for two packets).

  As is apparent from the above description of the operation, noise is promoted by the thinning readout driving method (b) because there is an empty packet from which no signal charge is read, and this empty packet is also on the vertical transfer unit 101. Therefore, by performing mixing between two vertical pixels (upper and lower two packets) in the horizontal transfer unit 14, the noise component for two packets is added to the signal component for one pixel, This is because the S / N is deteriorated twice (6 dB).

  Proposals for solving this problem have been made in the past, and one method is disclosed in Patent Document 3 above. This technique will be described below.

  FIG. 12 is a schematic configuration diagram showing a solid-state image sensor according to the prior art. Here, for example, in the all-pixel readout driving type CCD image pickup device, a case where a thinning readout mode in which pixel information is thinned out every other pixel in the vertical direction can be taken as an example in addition to the all-pixel readout mode is described as an example. It shall be.

  In FIG. 12, an imaging unit (imaging area) 11 is arranged in a matrix on a semiconductor substrate, and receives a plurality of light receiving elements such as photodiodes that convert incident light into signal charges having a charge amount corresponding to the amount of light. An element (pixel) 12 and a plurality of vertical transfer units 13 provided along the arrangement direction for each vertical column of the plurality of light receiving elements 12 are configured.

  The plurality of vertical transfer units 13 includes a set of signal packets provided with a 1: 1 correspondence with each pixel and empty packets (packets) arranged one by one on the front side in the vertical transfer direction of these signal packets. And three transfer electrodes (not shown) are arranged in the transfer direction on the transfer channel of each packet. Here, the packet refers to a unit for transferring a signal on a signal charge transfer path. The vertical transfer unit 13 is driven to transfer by, for example, three-phase vertical transfer pulses Vφ1, Vφ2, and Vφ3.

  That is, in the horizontal blanking period, the vertical transfer unit 13 transfers signal charges for one row (one line) line by line in accordance with the three-phase vertical transfer pulses Vφ1, Vφ2, and Vφ3 (hereinafter referred to as “line”). (Referred to as shift). Here, since a part of the transfer electrode of the vertical transfer unit 13 also serves as a read gate electrode, any one of the vertical transfer pulses Vφ1, Vφ2, and Vφ3 has a low level, an intermediate level, and a high level. The ternary level is taken, and the third level high level pulse becomes the readout pulse.

  However, how to apply the three-phase vertical transfer pulses Vφ1, Vφ2, and Vφ3 to each packet is different in the thinning-out reading mode from the all-pixel reading mode. That is, as a matter of course, no readout pulse is applied to a pixel from which signal charges are read out. In the vertical transfer unit 13, the line shift is performed once in the horizontal blanking period in the all-pixel readout mode, whereas the line shift is performed twice in the horizontal blanking period in the thinning readout driving for every other pixel. It will be.

  On the lower side of the image pickup unit 11 in the drawing, a horizontal transfer unit 14 is disposed adjacent to the transfer direction side ends of the plurality of vertical transfer units 13. In the thinning readout mode, the horizontal transfer unit 14 needs to transfer a signal component and a noise component as a pair. Therefore, a packet arrangement (packet string) that is twice the number of pixels in the horizontal direction (double density) is required. ). Then, the transfer is driven by the horizontal transfer pulses Hφ1 and Hφ2 having opposite phases. The horizontal transfer pulses Hφ1 and Hφ2 are set to twice the normal frequency because the horizontal transfer unit 14 is configured with double density.

  At the end of the horizontal transfer unit 14 on the transfer destination side, the signal charge transferred by the horizontal transfer unit 14 is detected, converted into a signal voltage and output, for example, a charge detection unit having a floating diffusion amplifier configuration. 15 is arranged. The charge detection unit 15 includes an FD (floating diffusion) unit 17 provided adjacent to the final output gate 16 of the horizontal transfer unit 14, an RD (reset drain) unit 18 that sweeps out charges, and charges of the FD unit 17. An RG (reset gate) unit 19 for discharging to the RD unit 18 is configured.

  In the charge detection unit 15, a predetermined drain voltage Vrd is given to the RD unit 18. Further, the horizontal transfer pulses Hφ1 and Hφ2 and, for example, the reset gate pulse RGφ having the same cycle are applied to the RG unit 19. The FD unit 17 derives a signal output Vout obtained by converting the signal charge into a signal voltage. Various timing signals including vertical transfer pulses Vφ1 to Vφ3, horizontal transfer pulses Hφ1 and Hφ2, and a reset gate pulse RGφ are generated by the timing generation circuit 20.

  FIG. 13 shows the timing relationship between the horizontal transfer pulses Hφ1, Hφ2, the reset gate pulse RGφ, and the signal output Vout. In the waveform of the signal output Vout, the P phase is a preset part, the D phase is a data part, and the horizontal transfer part 14 has a double-density configuration with two packets as one unit. The information of the packet on the rear side is output as the D phase in the P phase.

  Next, the operation in each drive mode of the all-pixel readout drive type CCD image pickup device having the above configuration will be described with reference to the schematic diagrams of FIGS. 14 and 15, white circles (◯) represent noise components on the vertical transfer unit 13, and black circles (●) represent noise transferred once in the horizontal blanking period. Each of the signal components including the components is shown.

  First, when the all-pixel readout mode is set by the mode switching signal, the timing generation circuit 20 outputs vertical transfer pulses Vφ1 to Vφ3 having timings corresponding to the mode. Thereby, in the schematic diagram of FIG. 14, the signal charges of all the pixels (indicated by black squares in the figure) are simultaneously read out to the vertical transfer unit 13 at the same time (FIG. 14A).

  Subsequently, the line transfer is performed in the vertical transfer unit 13, but before that, the horizontal transfer unit 14 uses 2 packets as one unit. In order to supply signal charges, the horizontal transfer unit 14 performs a shift of 1 bit (1 packet) in advance (hereinafter referred to as 1 bit shift). Thereafter, the line shift is performed once in the horizontal blanking period. As a result, the signal charge for one line (one row) is shifted to the horizontal transfer unit 14 (FIG. 14B).

  In this way, the signal charge for one line shifted from the vertical transfer unit 13 to the horizontal transfer unit 14 is transferred in the horizontal direction by the horizontal transfer pulses Hφ1 and Hφ2, and sequentially injected into the charge detection unit 15 in units of pixels. Is done. In the charge detection unit 15, the reset gate pulse RGφ is applied to the RG unit 19 in the same cycle as the horizontal transfer pulses Hφ <b> 1 and Hφ <b> 2, thereby performing a reset operation for discharging the residual charge in the FD unit 17 to the RD 18.

  As a result, a signal output Vout having a waveform as shown in FIG. 13 is derived from the FD unit 17. In the all-pixel readout mode, as is apparent from the above-described operation description, no information is given to the forward packet from the vertical transfer unit 13. Therefore, in the signal output Vout, no information is carried on the P phase, the P phase becomes a reference for signal processing to be described later, and signal component information is carried on the D phase.

  On the other hand, when the thinning readout mode is set by the mode switching signal, the timing generation circuit 20 outputs vertical transfer pulses Vφ1 to Vφ3 having timings corresponding to the mode. Accordingly, in the schematic diagram of FIG. 15, for example, only the signal charges (indicated by black squares in the figure) of pixels in even rows (2, 4,... In the figure) are read out to the vertical transfer unit 13 (see FIG. 15). FIG. 15 (a)).

  Subsequently, the vertical transfer unit 13 performs the first line shift within the horizontal blanking period. As a result, the charge of one line of noise component is supplied from the vertical transfer unit 13 to the packet on the front side in the horizontal transfer direction of the packet pair of the horizontal transfer unit 14 (FIG. 15B). Subsequently, the horizontal transfer unit 14 performs a 1-bit shift.

  Next, the vertical transfer unit 13 performs a second line shift within the same blanking period. As a result, the charge of the signal component for one line is supplied from the vertical transfer unit 13 to the packet on the rear side in the horizontal transfer direction of the packet pair of the horizontal transfer unit 14 (FIG. 15C). As a result, on the horizontal transfer unit 14, the noise component charge is accumulated in front of the packet pair, and the signal component charge is accumulated behind.

  In this manner, the noise component and the signal component charges shifted from the vertical transfer unit 13 to the horizontal transfer unit 14 by performing the line shift twice in the horizontal blanking period are transferred in pairs in the horizontal direction, The charges are sequentially injected into the charge detector 15. In the charge detector 15, the reset operation is performed by the reset gate pulse RGφ having the same period as the horizontal transfer pulses Hφ1 and Hφ2, and as a result, in the signal output Vout having the waveform as shown in FIG. The signal component information is carried on the D phase.

  In the charge detection unit 15, the reset operation is performed in the same cycle as the transfer cycle of the horizontal transfer unit 14, so that the reset operation is also performed between the P phase and the D phase. As a result, a noise component for one packet rides on the P phase, and a signal component + noise component for one packet rides on the D phase.

  As described above, in the CCD image pickup device of the all-pixel readout driving method, in the thinning readout mode, each charge of the signal packet and the empty packet is transferred horizontally as a pair (in this example, a pair), and this horizontally transferred charge Are sequentially converted into electrical signals, the noise component of the empty packet is put on the P phase, and the signal component of the signal packet is put on the D phase and output, so that the noise of the same pixel column as the pixel information, that is, the same address in the horizontal direction, is output. Information can be obtained in combination with the pixel information.

  Therefore, the noise component included in the signal component can be canceled by performing the process of obtaining the difference between the D-phase signal component and the P-phase noise component in the subsequent signal processing system. Here, the noise component of the empty packet means noise that rides in the same direction as vertical transfer such as smear and dark current.

Further, in Patent Document 4, a signal component + smear component read from a readout pixel and a smear component included in an empty packet are differentiated by a subsequent circuit to remove the smear component, and the final part of the vertical transfer path. It is disclosed that provision of a transfer blocking gate can accommodate not only vertical thinning but also horizontal thinning.
JP-A-9-298755 Japanese Patent Laid-Open No. 11-234688 JP 2000-299817 A Japanese Patent Laying-Open No. 2005-328212

However, the above-described methods disclosed in Patent Documents 3 and 4 have the following problems.
First, the method disclosed in Patent Document 3 cannot cope with a driving method for thinning out pixels in the horizontal direction. For example, consider a case where pixels are read every other pixel in the horizontal direction. According to the driving method described above, after transferring the signal charge from the vertical transfer unit to the horizontal transfer unit, the charge is shifted by 1 bit in the horizontal transfer unit, and then the noise component of the empty packet is transferred from the vertical transfer unit to the horizontal transfer unit. In this case, the noise component and the signal component are separated from each other by sending it to the subsequent packet. However, when pixels are thinned out in the horizontal direction, signal packets and empty packets are alternately present in the same row from which signals are read, and all thinned-out rows from which signals are not read are arranged as empty packets. Therefore, even if a bit shift is performed after the signal is transferred from the vertical transfer unit to the horizontal transfer unit as in this method, the signal packet and the empty packet are not separated properly.

  Secondly, in the method disclosed in Patent Document 3, it is necessary to configure the horizontal transfer unit to be composed of a packet sequence having a double density of the number of pixels in the horizontal direction. It becomes large and is not suitable for miniaturization of elements. In addition, since the capacitance is too large for the saturation charge amount, there is a risk of signal rounding during high-speed transfer.

  Thirdly, in the method disclosed in Patent Document 3, it is necessary to set the driving frequency of the horizontal transfer unit to twice the normal driving frequency, which not only complicates the driving circuit but also leads to an increase in power consumption.

  Fourth, in the method disclosed in Patent Document 4, it is necessary to horizontally transfer signal packets and empty packets separately, and high-speed driving by reducing the number of horizontal transfers is impossible. Moreover, the method of making the horizontal transfer packet double-density has the same problem as in Patent Document 3.

  In addition, a technique for providing a transfer blocking gate according to the number of thinning out in the horizontal direction and separating a signal packet and an empty packet in the same vertical transfer string has been disclosed, but the charge blocked by the transfer blocking gate is discharged. There is no disclosure about the technology, and if this charge cannot be discharged, horizontal thinning cannot be realized. For example, it is conceivable to provide a drain for discharge near the transfer blocking gate, or horizontally transfer the blocked charge separately and discharge it, but the former is structurally difficult to cope with miniaturization, and the latter is This is not a realistic method because it is not suitable for high speed because horizontal transfer for discharging unnecessary charges is required separately. Further, since the signal component that is not sent to the horizontal transfer unit by the transfer blocking gate is discarded as it is, the signal component cannot be effectively used for improving the image quality.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a solid-state imaging device and a driving method thereof capable of obtaining a high-quality image by greatly reducing noise such as smear even in a driving mode in which pixels are thinned out in the horizontal direction.

  In order to solve the above problems, a solid-state imaging device of the present invention is provided corresponding to a plurality of light receiving elements arranged in a matrix and a row of the light receiving elements, and performs thinning-out reading from the plurality of light receiving elements in a moving image capturing mode. A plurality of vertical transfer units that vertically transfer a plurality of signal packets including the signal charges and dummy packets other than the signal packets, and a final stage of columns other than M columns in every N columns of the plurality of vertical transfer units A plurality of holding units capable of mixing, holding and vertically transferring signal charges of the signal packet and the dummy packet independently of the vertical transfer from the upstream, and the vertical corresponding to the plurality of holding units or the M columns A horizontal transfer unit that mixes, holds, and horizontally transfers signal charges transferred from the transfer unit; and a drive unit that drives the vertical transfer unit, the holding unit, and the horizontal transfer unit, and the drive unit includes: In the moving image capturing mode, the vertical transfer from the plurality of holding units to the horizontal transfer unit is driven so as to generate the first mixed packet and the second mixed packet in the horizontal transfer unit, and the first mixed packet is transmitted to the horizontal transfer unit. Are mixed with two or more signal packets of the same color in the same column in a predetermined stage of the horizontal transfer unit, and the second mixed packet includes 2 of the same color in the same column in a stage other than the predetermined stage of the horizontal transfer unit. The above signal packet and two or more dummy packets are mixed. According to this configuration, since the second mixed packet and the first mixed packet are output from the solid-state imaging device, based on the subtraction process between the second mixed packet and the first mixed packet in the signal processing at the subsequent stage of the solid-state imaging device. The noise component can be removed from the first mixed packet. As a result, image quality can be improved by reducing noise such as smear.

  Here, the number of dummy packets mixed with the second mixed packet may be greater than or equal to the number of signal packets mixed with the second mixed packet. According to this configuration, since the number of dummy packets in the second mixed packet is equal to or greater than the number of signal packets, the noise component amount in the subtraction process is larger than the effective signal component amount. The accuracy can be improved.

  Here, the number of signal packets mixed with the second mixed packet may be the same as the number of signal packets mixed with the first mixed packet. According to this configuration, it is not necessary to multiply the first mixed packet by a coefficient, and by simply subtracting the first mixed packet from the second mixed packet, the noise component amount can be calculated, that is, the subtraction process is simplified. Can do.

  Here, the signal packet mixed with the second mixed packet and the signal packet mixed with the first mixed packet may be read from light receiving elements of the same color in the same row and in the nearest column. . According to this configuration, it is possible to reduce the error between the signal components of the first mixed packet and the second mixed packet, and it is possible to improve the accuracy of calculating the noise component amount.

  Here, the plurality of holding units are final transfer stages of the vertical transfer unit in columns other than the M columns in the N columns of the plurality of vertical transfer units, and are independent transfer electrodes for columns separated by N columns. You may make it have. According to this configuration, the above-mentioned mixing that reduces noise such as smear is provided by providing independent transfer electrodes in the final transfer stage of the vertical transfer unit other than M columns in every N columns and driving them independently. Can be realized.

  The plurality of holding units are provided between the vertical transfer unit and the horizontal transfer unit in columns other than the M columns in the N columns of the plurality of vertical transfer units, and are independent for each column separated by N columns. Thus, the signal charge may be held and transferred. According to this configuration, since the holding unit for independently holding and transferring is provided between the vertical transfer unit and the horizontal transfer unit other than the M columns in every N columns, transfer driving is simplified and the frame rate is increased. Suitable for

  The solid-state imaging device driving method of the present invention is the solid-state imaging device driving method according to claim 1, wherein the first mixed packet and the second mixed packet are sequentially output from the horizontal transfer unit, and the second A noise component amount is calculated based on a difference between the mixed packet and the first mixed packet, and a noise component present in the first mixed packet is removed using the noise component amount.

  Moreover, the camera of this invention is equipped with the said solid-state imaging device, It is characterized by the above-mentioned.

  According to the solid-state imaging device, the driving method thereof, and the camera of the present invention, it is possible to improve image quality by reducing noise such as smear.

(Embodiment 1)
The solid-state imaging device according to the present embodiment is located at the last stage of columns other than the M columns in every N columns of the plurality of vertical transfer units, and independently of the vertical transfer from the upstream, the signal packet and the dummy packet signal The horizontal transfer unit includes a plurality of holding units capable of mixing, holding, and vertically transferring electric charges, and generating the first mixed packet and the second mixed packet in the horizontal transfer unit in the moving image capturing mode. Drive vertical transfer to. Here, in the first mixed packet, two or more signal packets of the same color in the same column are mixed in a predetermined stage of the horizontal transfer unit. In the second mixed packet, two or more signal packets of the same color and two or more dummy packets in the same column are mixed in a stage other than the predetermined stage of the horizontal transfer unit. A signal packet refers to a plurality of vertical transfer stages including signal charges thinned out and read from a plurality of light receiving elements. A dummy packet refers to a plurality of vertical transfer stages that are originally empty without reading signal charges from a light receiving element.

  Further, the noise component caused by smear or dark current can be removed from the first mixed packet by the subtraction process of the second mixed packet and the first mixed packet. Thereby, the image quality can be improved.

  The holding unit in the present embodiment is the final transfer stage of the vertical transfer unit in columns other than the M columns in the N columns of the plurality of vertical transfer units, and is independent for each column separated by N columns. It has an electrode. By the drive pulse applied to this independent transfer electrode, the vertical transfer is performed independently for each column separated by N columns, the signal charges of the signal packet and the dummy packet transferred from the upstream are mixed, and mixed. It is configured to be able to hold packets.

  FIG. 1 is a block diagram showing the configuration of the solid-state imaging device according to Embodiment 1 of the present invention. The solid-state imaging device shown in the figure operates in a moving image capturing mode in addition to a still image capturing mode. In the moving image capturing mode, a moving image having a resolution obtained by thinning out M (for example, 2) pixels out of N (for example, 3) pixels in the vertical direction and the horizontal direction is captured. This moving image capturing mode includes driving for vertically transferring signal packets and dummy packets including noise alternately (not necessarily one-on-one alternately). The solid-state imaging device shown in the figure includes a plurality of light receiving elements 12, a plurality of vertical transfer units 13, a horizontal transfer unit 14, a charge detection unit 15, and a timing generation circuit 20.

  The plurality of light receiving elements 12 are arranged in a matrix, and for example, color filters are arranged in a Bayer array.

  The plurality of vertical transfer units 13 are provided corresponding to the respective columns of the plurality of light receiving elements 12, and a plurality of signal packets including signal charges thinned out and read from the plurality of light receiving elements in the moving image capturing mode, and a signal packet Vertical transfer with other dummy packets.

  Each final stage 21 in columns other than M (here, 1) in the N (here, 3) columns of the plurality of vertical transfer units functions as the holding unit, and is independent of vertical transfer from upstream. In addition, the signal charges of the signal packet and the dummy packet can be mixed, held and vertically transferred. 1 column (or R column), 2 columns (L column), 3 columns (B column), 1 column (R column), 2 columns (L column), 3 columns ( B column)... In the present embodiment, the final transfer stages of each R column and each L column have transfer electrodes that are driven independently from all the upstream transfer stages. The final transfer stage of each B column has a transfer electrode that is driven in common with all the upstream transfer stages. That is, the final transfer stages of each R column and each L column can be driven independently.

  The horizontal transfer unit 14 mixes, holds, and horizontally transfers signal charges transferred from the plurality of vertical transfer units 13.

  The charge detection unit 15 holds the signal charge from the output gate 16 and the output gate 16 that take out the signal charge from the final stage of the horizontal transfer unit 14 in order to convert the signal charge horizontally transferred from the horizontal transfer unit 14 into a voltage. An FD (floating diffusion) unit 17, an RD (reset drain) unit 18 that sweeps out the FD unit 17 during reset, and an RG (reset gate) unit 19 for applying a reset voltage to the FD unit 17 are provided.

  The timing generation circuit 20 drives a plurality of vertical transfer units 13 and horizontal transfer units 14. In particular, the timing generation circuit 20 mixes the signal packet and the plurality of dummy packets in the same column in each final stage 21 in the moving image capturing mode into one packet, and holds the signal charge of the mixed packet in each final stage, Furthermore, the horizontal transfer unit 14 and the final stage are driven so that the horizontal transfer unit 14 mixes other final stage packets holding signal charge signals corresponding to the same color.

  FIG. 2 is a block diagram illustrating an electrode configuration example of the solid-state imaging device according to the first embodiment. In the figure, six-phase transfer electrodes V1 to V6 are provided on each vertical transfer unit, and a six-phase drive signal is applied. The even-phase electrodes V2, V4, and V6 are transfer electrodes to which a drive signal for vertical transfer is applied. The odd-phase electrodes V1, V3, and V4 are transfer electrodes to which a drive signal for vertical transfer is applied, and also serve as readout electrodes from the light receiving elements.

  Six-phase common drive signals are applied to the six-phase transfer electrodes of the vertical transfer stage other than the vertical final stage and the three columns (B columns) of the vertical final stage. A drive signal independent of the electrodes V3 and V5 can be applied to the electrodes V31 and V51 in one row (L row). A drive signal independent of the electrodes V3 and V5 can also be applied to the two rows (R rows) of electrodes V32 and V52. As a result, the final transfer stages of each one column and each two columns can be driven independently.

  FIG. 3 is a block diagram illustrating a schematic configuration of the digital camera according to the first embodiment. The digital camera includes an optical system including a lens 31 for imaging incident light from a subject on the imaging surface of the solid-state imaging device 1, a control unit 32 that controls driving of the solid-state imaging device 1, and a solid-state imaging device. 1 and an image processing unit 33 that performs various signal processing on the output signal from 1.

  The image processing unit 33 includes a memory, and removes noise components caused by smear and dark current from the first mixed packet by noise subtraction processing using the second mixed packet and the first mixed packet.

  The operation of the solid-state imaging device according to the present embodiment configured as described above will be described.

  4A to 4D are diagrams illustrating specific examples of transfer and mixing of signal packets and dummy packets in the moving image capturing mode. Here, a case of thinning readout in which two vertical pixels are mixed will be described.

  In FIG. 4A, 123123 in the upper stage represents RLBRLB. Of the plurality of vertical transfer units 13, only a part up to 9 rows and 12 columns is shown. R (1, 1) represents a signal packet holding a signal charge indicating red in the first row from the bottom and the first column from the left. D (9, 1) represents a dummy packet having no effective signal charge in the ninth row and the first column. G and B represent signal packets corresponding to green and blue. 4B to 4D also show the coordinates of the original transfer packet in FIG. 4A.

  FIG. 4D shows a state where the first mixed packet and the second mixed packet are generated in the horizontal transfer unit 14 by the transfer and mixing from the state of FIG. 4A. 4B and 4C show a state in the middle.

  In the figure, R (1, 3) and R (2, 3) to which a solid line ○ is given indicate an example of a signal packet to be mixed into one first mixed packet. R (1, 5), R (2, 5), D (0, 5), D (3, 3) to which a broken-line mark is given are signal packets to be mixed into one second mixed packet or An example of a dummy packet is shown.

  As shown in FIG. 4D, each stage in the horizontal transfer unit 14 with a double circle indicates a first mixed packet, and each stage with a triangular mark indicates a second mixed packet. The stage marked with x indicates a mixed packet that is unnecessary in the moving image mode. By subtracting the signal charge amount of the first mixed packet from the signal charge amount of the second mixed packet, the noise amount for two dummy packets can be calculated. This is because the signal packet in the second mixed packet and the signal packet in the first mixed packet are the closest columns in the same row, so that the signal charge amount is approximately approximate and the error can be regarded as small.

  In FIG. 4D, the number of dummy packets mixed with the second mixed packet is two, which is the same as the number of signal packets mixed with the second mixed packet, but the former is preferably greater than or equal to the latter. As the number of dummy packets in the second mixed packet is greater than or equal to the number of signal packets, the noise component amount is larger than the effective signal component amount, and therefore the accuracy of calculation of the noise component amount can be improved. This is because if the noise component amount is larger than the effective signal component amount, the error of the effective signal component amount relative to the noise component amount can be made relatively small.

  FIG. 5 is a flowchart showing the operation of the timing generation circuit 20 that drives the transfer of FIGS. 4A to 4D.

  In step S <b> 5 </ b> A of FIG. 5, the timing generation circuit 20 performs reading from the plurality of light receiving elements 12 to the vertical transfer unit 13. As a result, the state shown in FIG. 4A is obtained. In FIG. 4A, all the stages of the horizontal transfer unit 14 are still empty. Dummy packets such as D (0, 1) and D (0, 2) corresponding to the 0th row are dummy signals read from the optical black pixels or dummy pixels.

  In step S5B, the timing generation circuit 20 drives so that the entire RLB column including the last stage is vertically transferred by one stage. As a result, the state shown in FIG. 4B is obtained. In FIG. 4B, the last-stage signal packet and the dummy packet are transferred to the horizontal transfer unit 14.

  In step S5C, the timing generation circuit 20 drives to vertically transfer the entire RLB column including the last stage by one stage. As a result, the signal packet from the vertical final stage is vertically added to each stage of the horizontal transfer unit 14. As a result, the state shown in FIG. 4C is obtained.

  In step S5D, the timing generation circuit 20 drives the horizontal transfer unit 14 to perform two-stage horizontal transfer. Further, in step S5E, the timing generation circuit 20 drives the B column to transfer one column vertically and the R and L columns to be vertically mixed in the final column. As a result, the state shown in FIG. 4D is obtained.

  Furthermore, in steps S5F and S5G, the timing generation circuit 20 drives the horizontal transfer unit 14 to sequentially transfer one row horizontally, and if there are untransferred signal packets in the plurality of vertical transfer units 13, the process proceeds to step S2B. Return and repeat the above operation.

  FIG. 6 is a flowchart showing noise subtraction processing in the image processing unit 33. The noise subtraction process in FIG. 6 is performed using the first mixed packet and the second mixed packet output by horizontal transfer for one row in step S5F in FIG.

  First, the image processing unit 33 detects the signal level of the first mixed packet (S81), and detects the signal level of the second mixed packet closest to the first mixed packet (S82).

  The image processing unit 33 detects the noise level included in the first mixed packet by subtracting the signal level of the first mixed packet and the signal level of the second mixed packet (S83). In the example of FIG. 4D, since the second mixed packet includes two signal packets and two dummy packets, the second mixed packet includes an effective signal for two packets and a noise component for four packets. Further, since the first mixed packet includes two signal packets, the first mixed packet includes an effective signal for two packets and a noise component for two packets. When the first mixed packet and the second mixed packet are the most recent packets, the effective signal level per packet of the first and second mixed packets can be regarded as substantially the same, and the noise component per packet The level can be regarded as almost the same. Therefore, the difference between the two can be regarded as representing a noise component for two packets. Further, the image processing unit 33 subtracts the noise component level from the signal level of the first mixed packet (S84).

  As described above, noise components due to smear and dark current can be removed from the first mixed packet by the noise subtraction process using the second mixed packet and the first mixed packet output from the solid-state imaging device. , Image quality can be improved.

  In FIG. 4D, the second mixed packet includes two signal packets and two dummy packets, and the second mixed packet includes two signal packets. However, the second mixed packet includes the dummy packets included in the second mixed packet. The number is preferably larger than the number of signal packets included in the second mixed packet. For example, when the second mixed packet includes three signal packets and six dummy packets, it includes three signals and nine noises. Further, when the first mixed packet is composed of three signal packets, it includes three signal packets and two noise packets. When the first mixed packet and the second mixed packet are the most recent packets, the three valid signals of the first and second mixed packets can be regarded as substantially the same. In this case, the difference between the two can be regarded as representing 6 packets of noise. In this case, since the amount of noise component in the subtraction process in S83 is larger than the amount of effective signal component, the effective amount of signal between the signal packets of the first and second mixed packets becomes relatively small, and the amount of noise component The accuracy of calculation of can be improved.

  Further, the signal packet mixed with the second mixed packet and the signal packet mixed with the first mixed packet are signals read from the light receiving elements of the same color in the same row and the nearest column in the state of FIG. 4A. Preferably it is a packet. By doing so, the error of the effective signal amount between the signal packets of the first and second mixed packets is reduced, and the accuracy of calculation of the noise component amount can be improved.

  Furthermore, it is preferable that the number of signal packets mixed with the second mixed packet is the same as the number of signal packets mixed with the first mixed packet. In this way, if the first mixed packet is simply subtracted from the second mixed packet without multiplying the first mixed packet by a coefficient, the noise component amount can be calculated, that is, the subtraction process can be simplified.

(Embodiment 2)
In the present embodiment, none of the final stages of the vertical transfer unit 13 has an independent transfer electrode, and the vertical transfer unit and the horizontal transfer unit in columns other than the M columns in every N columns of the plurality of vertical transfer units. A configuration in which a holding unit that holds and transfers signal charges independently for each column separated by N columns is provided as the above holding unit will be described.

  FIG. 7 is a block diagram showing a configuration of the solid-state imaging device according to Embodiment 2 of the present invention. The figure differs from FIG. 1 in that the final stage 21 of the all vertical transfer part 13 does not have an independent transfer electrode and that a plurality of hold parts 21a are added. Hereinafter, description of the same points will be omitted, and different points will be mainly described.

  The plurality of hold units 21a are provided between the vertical transfer unit 13 and the horizontal transfer unit 14 in columns other than the M columns in the N columns of the plurality of vertical transfer units 13, and are independently provided for each column separated by N columns. Holds and transfers signal charge. Functionally, the final stage 21 and the hold unit 21a in FIG. 1 are substantially the same.

  In addition to driving the plurality of vertical transfer units 13 and the horizontal transfer unit 14, the timing generation circuit 20 also drives a plurality of hold units 21a. In particular, the timing generation circuit 20 drives the vertical transfer from the plurality of hold units to the horizontal transfer unit so that the horizontal transfer unit 14 generates the first mixed packet and the second mixed packet in the moving image capturing mode.

  FIG. 8 is a block diagram illustrating an electrode configuration example of the solid-state imaging device according to the second embodiment. 2 is different from FIG. 2 in that the electrodes VS1 and VB2 on the hold unit 21a are provided instead of the six-phase electrodes on the final stage 21 of the vertical transfer unit 13. Hereinafter, description of the same points will be omitted, and different points will be mainly described.

  The electrode VS1 is a storage electrode, and a drive signal for controlling whether or not the signal charge is held in the hold unit is applied. The electrode VB2 is a barrier electrode, and a drive signal for controlling whether or not the signal charge is transferred from the hold unit to the horizontal transfer unit 14 is applied.

The configuration of the camera in the present embodiment is the same as that in FIG.
The operation of the solid-state imaging device according to the present embodiment configured as described above will be described.

  9A to 9E are diagrams illustrating specific examples of transfer and mixing of signal packets and dummy packets in the moving image capturing mode.

  In FIG. 9A, as in FIG. 4A, the upper 123123... Represent RLBRLB. The state of the hold unit 21a as well as the vertical transfer unit is also illustrated. 9B to 9E also show the original coordinates in FIG. 9A.

  FIG. 9E shows a state where the first mixed packet and the second mixed packet are generated in the horizontal transfer unit 14 by the transfer and mixing from the state of FIG. 4A. 9B to 9D show a state in the middle.

  In the figure, R (1, 3) and R (2, 3) to which a solid line ○ is given indicate an example of a signal packet to be mixed into one first mixed packet. R (1, 5), R (2, 5), D (0, 5), D (3, 3) to which a broken-line mark is given are signal packets to be mixed into one second mixed packet or An example of a dummy packet is shown.

  FIG. 10 is a flowchart showing the operation of the timing generation circuit 20 that drives the transfer of FIGS. 9A to 9E.

  In step S <b> 8 </ b> A of FIG. 10, the timing generation circuit 20 performs reading from the plurality of light receiving elements 12 to the vertical transfer unit 13. As a result, the state shown in FIG. 9A is obtained. In FIG. 9A, all stages of the horizontal transfer unit 14 are still empty. A dummy packet such as D (0, 1) and D (0, 2) corresponding to the 0th row in the hold unit is a dummy signal read from the optical black pixel or the dummy pixel.

  In step S8B, the timing generation circuit 20 drives to transfer the signal charge from the hold unit 21a to the horizontal transfer unit 14. As a result, the state of FIG. 9B is obtained. In FIG. 4B, the last-stage signal packet and the dummy packet are transferred to the horizontal transfer unit 14.

  In step S8C, the timing generation circuit 20 drives so that the RLB all columns are vertically transferred by one stage without being held by the hold unit. As a result, the state shown in FIG. 9C is obtained. FIG. 9C shows a result of transferring the signal packet and the dummy packet from the vertical transfer unit 13 to the horizontal transfer unit 14 via the hold unit 21a.

  In step S8D, the timing generation circuit 20 drives so that the RLB all columns are vertically transferred by one stage without being held by the hold unit. As a result, the state shown in FIG. 9D is obtained. FIG. 9D shows a result of transferring the signal packet and the dummy packet from the vertical transfer unit 13 to the horizontal transfer unit 14 via the hold unit 21a.

  In step S8E, the timing generation circuit 20 drives the horizontal transfer unit 14 to perform two-stage horizontal transfer. Further, in step S8F, the timing generation circuit 20 drives to transfer the B column in one vertical stage while holding the R and L columns. As a result, the state shown in FIG. 9E is obtained.

  Further, in steps S8F and S8G, the timing generation circuit 20 drives the horizontal transfer unit 14 to sequentially perform horizontal transfer for one row, and if untransferred signal packets remain in the plurality of vertical transfer units 13, the process proceeds to step S2B. Return and repeat the above operation.

  During the horizontal transfer period for one row in step S8F, the image processing unit 33 performs noise subtraction processing. This noise subtraction process is the same as that shown in FIG.

  As described above, the solid-state imaging device according to the present embodiment has a holding unit that holds and transfers independently between the vertical transfer unit and the horizontal transfer unit other than the M columns in every N columns. Similarly to Embodiment 1, in the moving image capturing mode, noise such as smear can be reduced, and the image quality can be improved by noise reduction processing. In addition, transfer driving is simplified, which is suitable for increasing the frame rate.

  The present invention is suitable for a solid-state imaging device having a plurality of light-receiving elements formed on a semiconductor substrate and a camera having the solid-state imaging device. For example, a CCD image sensor, a digital still camera, a mobile phone with a camera, and a surveillance camera Suitable for cameras built into notebook computers, camera units connected to information processing equipment, and the like.

It is a block diagram which shows the structure of the solid-state imaging device in Embodiment 1 of this invention. 3 is a block diagram illustrating an example of an electrode configuration of the solid-state imaging device according to Embodiment 1. FIG. 2 is a block diagram illustrating a configuration of a camera in Embodiment 1. FIG. 6 is an explanatory diagram of pixel mixture in Embodiment 1. FIG. 6 is an explanatory diagram of pixel mixture in Embodiment 1. FIG. 6 is an explanatory diagram of pixel mixture in Embodiment 1. FIG. 6 is an explanatory diagram of pixel mixture in Embodiment 1. FIG. 6 is a flowchart for driving pixel mixing in Embodiment 1. FIG. FIG. 6 is a flowchart showing noise subtraction processing in the first embodiment. It is a block diagram which shows the structure of the solid-state imaging device in Embodiment 2 of this invention. It is a block diagram which shows the electrode structural example of the solid-state imaging device in Embodiment 2 of this invention. FIG. 10 is an explanatory diagram of pixel mixing in the second embodiment. FIG. 10 is an explanatory diagram of pixel mixing in the second embodiment. FIG. 10 is an explanatory diagram of pixel mixing in the second embodiment. FIG. 10 is an explanatory diagram of pixel mixing in the second embodiment. FIG. 10 is an explanatory diagram of pixel mixing in the second embodiment. FIG. 10 is a flowchart for driving pixel mixing in the second embodiment. It is a schematic diagram for demonstrating the pixel reading drive of a prior art. It is a schematic block diagram which shows the solid-state image sensor concerning a prior art. It is a figure which shows the timing relationship of horizontal transfer pulse Hphi1, Hphi2, reset gate pulse RGphi, and signal output Vout. It is a figure which shows operation | movement in each drive mode of the CCD image pick-up element of an all pixel read drive system. It is a figure which shows operation | movement in each drive mode of the CCD image pick-up element of an all pixel read drive system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Solid-state image sensor 12 Light receiving element 13 Vertical transfer part 14 Horizontal transfer part 15 Charge detection part 16 Output gate 17 FD (floating diffusion) part 18 RD (reset drain) part 19 RG (reset gate) part 20 Timing generation circuit 21 Final stage 21a Hold unit 31 Lens 32 Control unit 33 Image processing unit

Claims (8)

A plurality of light receiving elements arranged in a matrix;
A plurality of vertical transfer units that are provided corresponding to the columns of the light receiving elements, and that vertically transfer a plurality of signal packets including signal charges read from the plurality of light receiving elements, and dummy packets other than the signal packets;
Located in the last stage of columns other than the M columns in the N columns of the plurality of vertical transfer units, the signal charges of the signal packet and the dummy packet can be mixed, held and vertically transferred independently of the vertical transfer from the upstream. A plurality of holding parts,
A horizontal transfer unit that mixes, holds, and horizontally transfers signal charges transferred from the plurality of holding units or the vertical transfer unit corresponding to the M columns;
A drive unit that drives the vertical transfer unit, the holding unit, and the horizontal transfer unit;
The driving unit drives vertical transfer from the plurality of holding units to the horizontal transfer unit so as to generate a first mixed packet and a second mixed packet in the horizontal transfer unit in the moving image capturing mode;
In the first mixed packet, two or more signal packets of the same color in the same column are mixed in a predetermined stage of the horizontal transfer unit,
In the second mixed packet, two or more signal packets of the same color in the same column and two or more dummy packets are mixed in a stage other than the predetermined stage of the horizontal transfer unit. . The solid-state imaging device according to claim 1, wherein the number of dummy packets mixed with the second mixed packet is equal to or greater than the number of signal packets mixed with the second mixed packet. The solid-state imaging device according to claim 2, wherein the number of signal packets mixed with the second mixed packet is the same as the number of signal packets mixed with the first mixed packet. The signal packet mixed with the second mixed packet and the signal packet mixed with the first mixed packet are read from the light receiving elements of the same color in the same row and in the nearest column. The solid-state imaging device described. The plurality of holding units are final transfer stages of the vertical transfer unit in columns other than the M columns in the N columns of the plurality of vertical transfer units, and have independent transfer electrodes for columns separated by N columns. The solid-state imaging device according to any one of claims 1 to 4. The plurality of holding units are provided between the vertical transfer unit and the horizontal transfer unit in columns other than the M columns in the N columns of the plurality of vertical transfer units, and independently signal for each column separated by N columns. The solid-state imaging device according to claim 1, wherein charge is held and transferred. The method for driving the solid-state imaging device according to claim 1,
The first mixed packet and the second mixed packet are sequentially output from the horizontal transfer unit,
Calculating a noise component amount based on a difference between the second mixed packet and the first mixed packet;
A method for driving a solid-state imaging device, wherein noise components inherent in the first mixed packet are removed using the noise component amount.   A camera comprising the solid-state imaging device according to claim 1.

Images