JP2020088536A - Imaging apparatus - Google Patents

Abstract

To prevent noise of a reset pulse for resetting a photoelectric conversion unit, when reading a predetermined row.SOLUTION: An imaging apparatus capable of adjusting exposure time, comprises: a plurality of photoelectric conversion units arranged in rows and columns, for generating charges by photoelectrically converting incident light; reading means that sequentially supplies a reading pulse of different phase for each row, in order to sequentially read, in row unit, charges generated by the plurality of photoelectric conversion units; reset pulse supply means that supplies the photoelectric conversion units with a reset pulse for resetting the photoelectric conversion units during a vertical period; and control means that shifts a phase of the reset pulse or a phase of the reading pulse by a predetermined time during reading of a predetermined row such that noise of the reset pulse does not overlap the reading period of the reading pulse. Thus, noise in an image signal is prevented.SELECTED DRAWING: Figure 4

Description

The present invention relates to an imaging device capable of adjusting exposure time.

In an imaging device such as a digital camera, there is known a technique of reducing the transmittance of the amount of photographing light for photographing in an extremely bright environment. This is because even in a bright environment, the aperture is opened to realize a subject with a shallow depth of field, or even if a long-second exposure is performed, saturation of the subject (for example, a waterfall) does not occur. This is for expressing a moving locus, and is conventionally performed using an optical filter such as an ND filter.

On the other hand, in the imaging device of Patent Document 1, the charge converted by the photoelectric conversion unit is transferred to the storage unit a plurality of times, and the charges transferred a plurality of times are collectively stored, so that the exposure time, the exposure amount, etc. There is described a configuration capable of rapidly changing the condition (1) at high speed. By controlling the operation of the image sensor using such a technique, it is possible to adjust the light amount without using an optical ND filter and obtain a natural moving image.
Furthermore, by providing two storage units in one pixel, while the signal stored in one storage unit is being read out, the charge stored in the photoelectric conversion unit is stored in the other storage unit. Can also As a result, the exposure/accumulation operation can be performed at the same time while the signal is being read from the pixel. Therefore, it is possible to increase the degree of freedom in setting the shooting time within one frame when shooting a moving image.

JP, 2010-157893, A

However, in the device of Patent Document 1, for example, when an attempt is made to store one of the storage units a plurality of times as described above when shooting a moving image, a pulse for resetting the photoelectric conversion unit is a signal being read. There is a problem that it is mixed in as noise.

Further, although the signals are sequentially read out row by row, there is a problem that an error occurs in the brightness data due to the influence of the noise in the row where the signal reading is performed at the same timing as the timing of resetting the photoelectric conversion unit. It was

It is an object of the present invention to provide an image pickup device capable of preventing noise of a pulse supplied during a vertical period for adjusting an exposure time from riding on an image signal.

The imaging device of the present invention is
A plurality of photoelectric conversion units arranged in rows and columns for photoelectrically converting incident light to generate charges,
Read-out means for sequentially supplying read-out pulses having different phases for each row in order to sequentially read out the charges generated by the plurality of photoelectric conversion units row by row,
Reset pulse supply means for supplying to the photoelectric conversion unit a reset pulse for resetting the photoelectric conversion unit during a vertical period,
Control means for shifting the phase of the reset pulse or the phase of the read pulse for a predetermined time when reading a predetermined row so that the noise of the reset pulse does not overlap the read period of the read pulse. Characterize.

According to the present invention, it is possible to prevent the noise of the pulse supplied during the vertical period for adjusting the exposure time from riding on the image signal.

External view of a digital camera. 3 is a block diagram of the image pickup apparatus according to the embodiment. FIG. 3 is a partial circuit diagram of the image sensor in the embodiment. 6 is a timing chart showing a drive sequence of the image sensor in the embodiment. 6 is a timing chart showing a drive sequence of the image sensor in the embodiment. 6 is a timing chart showing another example of the driving sequence of the image sensor. FIG. 5 is a diagram illustrating an example of an imaging result when a brightness error occurs. FIG. 6 is an explanatory diagram of an imaging result when the delay time amount Δt is large in the embodiment.

An image pickup apparatus according to an embodiment of the present invention will be described below.
1A and 1B are external views of, for example, a digital camera as an image pickup apparatus in the embodiment, FIG. 1A is a front view of the image pickup apparatus, and FIG. 1B is a rear view of the image pickup apparatus. is there. In the figure, reference numeral 151 is an image pickup apparatus main body having an image pickup element and a shutter device housed therein, and 152 is an image pickup optical system having an aperture as described later. Further, 153 is a movable display unit for displaying shooting information and video, 154 is a switch ST used mainly for shooting still images, and 155 is a switch for starting and stopping moving image shooting. It is MV. The display unit 153 has a display luminance range capable of displaying an image having a wide dynamic range without suppressing the luminance range.

Reference numeral 156 is a shooting mode selection lever for selecting a shooting mode, and 157 is a menu button for shifting to a function setting mode for setting functions of the image pickup apparatus. Further, 158 and 159 are up/down switches for changing various setting values, 160 is a dial for changing various setting values, and 161 is a reproduction button for shifting to the reproduction mode. When the play button 161 is pressed, the image file recorded on the recording medium in the main body of the image pickup apparatus is reproduced and displayed on the display unit 153.

FIG. 2 is a block diagram showing the schematic arrangement of the image pickup apparatus of this embodiment. In the figure, 184 is an image pickup device for converting an optical image of a subject formed through the taking optical system 152 into an electrical video signal, and 152 is a taking optical system for forming an optical image of the subject on the image pickup device 184. Is. Further, 180 is an optical axis of the photographing optical system 152, and 181 is a diaphragm for adjusting the amount of light passing through the photographing optical system 152, which is controlled by the diaphragm controller 182. An optical filter 183 limits the wavelength of light incident on the image sensor 184 and the spatial frequency transmitted to the image sensor 184. The image sensor 184 has a sufficient number of pixels, a signal reading speed, a color gamut, and a dynamic range to satisfy the Ultra High Definition Television standard.

A digital signal processing unit 187 compresses the digital video data output from the image sensor 184 after performing various corrections. Reference numeral 189 is a timing generator that outputs various timing signals to the digital signal processor 187 and the like, and reference numeral 178 is a system control CPU as a computer that controls various calculations and the entire digital still motion camera. The timing signal from the timing generator 189 is also supplied to a vertical scanning circuit 530, a horizontal scanning circuit 531 and the like in the image pickup device 184, which will be described later, and the driving is performed as shown in the timing charts of FIGS.

Reference numeral 190 is a memory unit that stores a computer program for operating the system control CPU to perform operation control as shown in the timing chart of FIG. The memory unit 190 also has a function of temporarily storing digital video data output from the digital signal processing unit 187 via the system control CPU 178. The memory unit 190 may be composed of multiple chips. Reference numeral 191 is a display interface unit for displaying the captured image on the display unit 153. Reference numeral 193 is a removable recording medium such as a semiconductor memory for recording video data or additional data, and 192 is a recording interface unit for recording or reading the recording medium 193. Reference numeral 196 is an external interface unit for communicating with the external computer 197 and the like. Reference numeral 195 is a printer such as a small inkjet printer, and 194 is a print interface unit for outputting a photographed image to the printer 195 for printing. Reference numeral 199 is a computer network such as the Internet, and 198 is a wireless interface unit for communicating with the computer network 199. A switch input unit 179 includes a switch ST154, a switch MV155, and a plurality of switches for switching various modes.

FIG. 3 is a partial circuit diagram of the image sensor 184. In the figure, among a large number of pixels arranged two-dimensionally on the light receiving surface of the image sensor 184, the pixel unit 300 in the first row and the first column (1,1) and an arbitrary m row and the first column (m, 1 ) Of the pixel portion 301. Since the pixel unit 300 and the pixel unit 301 have the same configuration, the constituent elements are given the same numbers.

Each pixel unit 300, 301, etc. of the image sensor 184 of this embodiment has two signal holding units (capacitors) for one photodiode 500, but the number is not limited to two and three or more. You may have a signal holding part. The photodiodes 500 function as photoelectric conversion units and photoelectrically convert incident lights to generate charges, and are arranged in rows and columns on the light receiving surface of the image sensor 184.

In the circuit diagram of FIG. 3, each pixel unit 300, 301, etc. includes a photodiode 500, a first signal holding unit 507A, a first transfer transistor 501A, a second transfer transistor 502A, and a second signal. The holding portion 507B, the third transfer transistor 501B, and the fourth transfer transistor 502B are included. Each pixel portion 300, 301 and the like has a fifth transfer transistor 503, a floating diffusion region 508, a reset transistor 504, an amplification transistor 505, a selection transistor 506, and a charge discharging region 510.

The first to fifth transfer transistors function as first to fifth transfer units, respectively, and the amplification transistor 505 functions as an amplification unit.

The first transfer transistor 501A and the second transfer transistor 502A are turned on when the transfer pulse φTX1A(1) and the transfer pulse φTX2A(1) are at high level, respectively. The third transfer transistor 501B and the fourth transfer transistor 502B are turned on when the transfer pulse φTX1B(1) and the transfer pulse φTX2B(1) are at high level, respectively. The reset transistor 504, the selection transistor 506, and the fifth transfer transistor 503 are turned on when the reset pulse φRES(1), the selection pulse φSEL(1), and the transfer pulse φTX3(1) are at high levels. The transfer pulse φTX3 functions as a reset pulse for resetting the photodiode (photoelectric conversion unit). Here, each control pulse is supplied from the vertical scanning circuit 530, and the vertical scanning circuit 530 also functions as a reset pulse supply means together with the timing generator 189 and the like.
Further, 520 and 521 are power supply lines, 523 is a signal output line, and 531 is a horizontal scanning circuit for sequentially reading out the signal of the signal output line 523 in the horizontal direction and outputting it to the outside of the image sensor 184.
The details of the operation of the image sensor 184 in this embodiment will be described below with reference to FIGS.

4 and 5 are timing charts showing the driving sequence of the image sensor 184 in this embodiment. Assuming that video recording is performed under the condition of 30 fps (frame per second), 1/480 seconds of charge accumulation is performed 4 times during 1/30 seconds which is one vertical period, and each charge is added. , To obtain the image signal. Timing information such as photoelectric conversion/accumulation/discharge shown in the timing charts of FIGS. 4 and 5 is stored in the memory unit 190 of FIG. The system control CPU 178 controls the operation of the image sensor 184 on the basis of the read timing information from the memory unit 190 when taking an image.

The image sensor 184 of this embodiment has a pixel array of a large number of rows in the vertical direction, and in FIGS. 4 and 5, the timings of the drive pulses of the first row, the second row, the n-th row, and the m-th row among them. Is shown. Then, these controls are repeatedly executed in the vertical direction (downward direction in the drawing) by the vertical scanning circuit 530 in synchronization with the horizontal synchronizing signal, thereby performing vertical scanning. The vertical scanning circuit 530 and the horizontal scanning circuit 531 perform an accumulation/readout operation for all pixels of the image sensor 184. The vertical scanning circuit 530, the horizontal scanning circuit 531 and the like constitute a reading means for sequentially reading out the charges generated by the plurality of photoelectric conversion units (photodiodes) row by row. The reading means sequentially supplies read pulses having different phases for each row.

In FIGS. 4 and 5, times t1, t6, t11, t13, and t14 of the vertical synchronization signal φV are vertical synchronization signals at which the vertical period starts, and 1/30 seconds, which is the time from t1 to t6, is one vertical period ( 1 frame period).
As a shooting condition, for a moving image, charge accumulation of 1/480 seconds is intermittently performed four times at equal pitches in one vertical period of 1/30 seconds, and a total of 1/120 is obtained by adding these charges. An example of obtaining an exposure time of seconds is shown.
First, the operation of the first line will be described.
First, at time t1, the timing generator 189 sets the vertical synchronizing signal φV to the high level, and at the same time sets the horizontal synchronizing signal φH to the high level.

The reset pulse φRES(1) of the first row becomes low level in synchronization with the time t1 when the vertical synchronizing signal φV and the horizontal synchronizing signal φH become high level. As a result, the reset transistor 504 of the first row is turned off, and the reset state of the floating diffusion region 508 is released. Further, at this time, φTX2B(1) becomes high level, the fourth transfer transistor 502B in the first row is turned on, and the signal accumulated in the second signal holding unit 507B is transferred to the floating diffusion region 508. Further, the selection pulse φSEL(1) of the first row becomes the high level, the selection transistor 506 of the first row is turned on, and the image signal stored in the second signal holding unit 507B of the first row is read out. Is possible.

That is, the output corresponding to the change in the potential of the floating diffusion region 508 is read to the signal output line 523 via the amplification transistor 505 and the selection transistor 506. Then, the horizontal scanning circuit 531 sequentially reads the image signals of the first row in the horizontal period, and outputs the signals of one row of the moving image to the outside.

Next, at time t2, the transfer pulse φTX2A(1) of the first row becomes high level for a short time, and the second transfer transistor 502A of the first row is turned on.
In that state, at time t2, the reset pulse φRES(1) becomes high level and the reset transistor 504 is turned on, so that the floating diffusion region 508 and the first signal holding unit 507A of the first row are reset. Note that at time t2, the selection pulse φSEL(1) for the first row becomes low level.

On the other hand, since the transfer pulse φTX3(1) of the first row is at the high level until time t3, the fifth transfer transistor 503 is turned on and the photodiode 500 of the first row is reset. However, at time t3, when the transfer pulse φTX3(1) of the first row becomes low level, the fifth transfer transistor 503 is turned off, the reset of the photodiode 500 of the first row is released, and the photodiode 500 receives the signal charge. Starts to accumulate.

Next, at time t4, when the transfer pulse φTX1A(1) of the first row becomes high level, the first transfer transistor 501A is turned on, and the signal charge accumulated in the photodiode 500 after t3 is the first signal. It is transferred to the holding unit 507A. Here, the first signal holding unit 507A functions as one of two signal holding units that alternately hold the charges photoelectrically converted in the first row every vertical period.

At time t5, when the transfer pulse φTX1A(1) of the first row becomes low level, the first transfer transistor 501A is turned off, and the signal charge accumulated in the photodiode 500 is transferred to the first signal holding unit 507A. Ends. At the same time, the transfer pulse φTX3(1) of the first row also returns to the high level, and the photodiode 500 of the first row is reset again. As a result, the charges accumulated in the photodiode 500 in the first row during the period from time t3 to time t5 are accumulated in the first signal holding unit 507A.

Here, the times t3 to t5 correspond to one accumulation time (exposure time) of 1/480 seconds in one vertical period, and are shown as the accumulation period 602-1 in the upward-sloping hatched area. Such an accumulation operation is performed four times intermittently and periodically within one vertical period, and is illustrated as accumulation periods 602-1, 602-2, 602-3, and 602-4 in the upward-sloping hatched area, respectively. There is. Then, by adding the charges accumulated in these four accumulation periods, the accumulation time (1/480 seconds×4 times=1/120 seconds) equivalent to the charges in one accumulation period of 120 seconds is obtained. You can get a charge. Note that the control operation in the accumulation periods 602-2, 602-3, and 602-4 is the same as the operation from time t3 to t5 in the accumulation period 602-1, and thus the description thereof will be omitted.

The transfer pulse φTX3 for the first row, which functions as a reset pulse for resetting the photoelectric conversion unit, is supplied a plurality of times during one vertical period from time t1 to time t6. Further, the pixel signal (charge) accumulated and added in the accumulation periods 602-1, 602-2, 602-3, 602-4 in which the transfer pulse φTX3 is at the low level is held in the signal holding unit 507A. The charge held in the signal holding unit 507A is not output as a pixel signal in the N-th vertical period (t1 to t6), but within a predetermined horizontal period (the period from t6 to t7) in the next N+1-th vertical period. ) Will be output to.

Next, at time t6, the vertical synchronizing signal φV supplied from the timing generating unit 189 becomes high level, and at the same time, the horizontal synchronizing signal φH becomes high level, and the next vertical period (N+1th cycle) is started. At this time, in synchronization with time t6 when the vertical synchronizing signal φV and the horizontal synchronizing signal φH become high level, when the reset pulse φRES(1) of the first row becomes low level, the reset transistor 504 of the first row is turned off. Thus, the reset state of the floating diffusion region 508 is released. At the same time, when the selection pulse φSEL(1) of the first row becomes high level, the selection transistor 506 of the first row is turned on, and the image signal of the first row can be read.
At this time, φTX2A(1) also becomes high level, and the second transfer transistor 502A in the first row is turned on. As a result, the charges accumulated in the first signal holding unit 507A are transferred to the floating diffusion region 508. Then, the output corresponding to the change in the potential of the floating diffusion region 508 is read out to the signal output line 523 via the amplification transistor 505 and the selection transistor 506. Then, the horizontal scanning circuit 531 sequentially reads the pixel signals of one row accumulated in the first signal holding unit 507A in the horizontal period.

At time t7, the transfer pulse φTX2B(1) of the first row becomes high level for a short time,
The fourth transfer transistor 502B in the first row is turned on. When the reset pulse φRES(1) becomes high level and the reset transistor 504 is turned on at time t7, the floating diffusion region 508 of the first row and the second signal holding unit 507B are reset. Note that at time t7, the selection pulse φSEL(1) for the first row becomes low level.
As described above, the pixel signals are accumulated and added in the first signal holding unit 507A as the moving image accumulation periods 602-1, 602-2, 602-3, and 602-4 between the times t1 and t6. Then, this pixel signal is read out by the horizontal scanning circuit 531 between times t6 and t7 and is output to the outside as the image signal of the first row.

Next, at time t8, when the transfer pulse φTX3(1) of the first row becomes low level, the fifth transfer transistor 503 is turned off, the reset of the photodiode 500 of the first row is released, and the photodiode 500 is released. The accumulation of signal charges is started.

At time t9, when the transfer pulse φTX2B(1) for the first row becomes high level, the third transfer transistor 501B is turned on. Then, the signal charges accumulated in the photodiode 500 are transferred to the second signal holding unit 507B, which is one of the two signal holding units holding the photoelectrically converted charges in the first row. In the Nth vertical period, the charges are transferred and accumulated in the first signal holding unit 507A, whereas in the N+1th vertical period, the charges are transferred and accumulated in the second signal holding unit 507B which is the other signal holding unit. is doing.

At time t10, when the transfer pulse φTX1A(1) of the first row becomes low level, the third transfer transistor 501B is turned off, and the signal charge accumulated in the photodiode 500 is transferred to the second signal holding unit 507B. Ends. At the same time, the transfer pulse φTX3(1) in the first row also returns to the high level, so that the fifth transfer transistor 503 returns to the ON state and the photodiode 500 in the first row is reset again.

Here, the time t8 to the time t10 corresponds to 1/480 seconds of one accumulation time (exposure time) of the moving image in the (N+1)th vertical period, and is illustrated as the accumulation period 702-1 in the lower right diagonally shaded area. It Such accumulation operations are performed intermittently and periodically four times, and these are illustrated as accumulation periods 702-1, 702-2, 702-3, and 702-4 in the lower right hatched area. Then, the total accumulation time (1/480 seconds×4 times=1/120 seconds) is obtained by adding these four accumulations. Note that the control operation in the accumulation periods 702-2, 702-3, and 702-4 is the same as the operation from time t8 to t10 in the accumulation period 702-1, and thus the description thereof will be omitted.

The image signal accumulated and added four times as the accumulation periods 702-1, 702-2, 702-3, and 702-4 in the vertical period from the time t6 to t11 is different from the previous vertical period in the second signal holding unit. 507B. The charge is not output as an image signal in the N+1th vertical period, and is read out within a predetermined horizontal period (the period of t11 to t12) of the next N+2th vertical period.

Next, at time t11, the vertical synchronizing signal φV becomes high level in the timing generating section 189, the horizontal synchronizing signal φH becomes high level at the same time, and the (N+2)th vertical period is started.
After time t11, the same operation as that of the Nth vertical period two before the above is performed.

That is, in synchronization with the time t11 when the vertical synchronizing signal φV and the horizontal synchronizing signal φH become high level, when the reset pulse φRES(1) of the first row becomes low level, the reset transistor 504 of the first row is turned off. The reset state of the floating diffusion region 508 is released. At the same time, when the selection pulse φSEL(1) of the first row becomes high level, the selection transistor 506 of the first row is turned on, and the image signal of the first row can be read.
At this time, similarly to the Nth vertical period, the transfer pulse φTX2B(1) of the first row becomes high level, and the fourth transfer transistor 502B of the first row is turned on. As a result, the charges accumulated in the second signal holding unit 507B are read out to the floating diffusion region 508. Then, the output corresponding to the change in the potential of the floating diffusion region 508 is read out to the signal output line 523 via the amplification transistor 505 and the selection transistor 506. Then, the horizontal scanning circuit 531 sequentially reads pixel signals in the horizontal period.

Further, at time t12, the transfer pulse φTX2A(1) of the first row becomes high level for a short time, and the second transfer transistor 502A of the first row is turned on. Then, at time t12, when the reset pulse φRES(1) becomes high level and the reset transistor 504 is turned on, the floating diffusion region 508 of the first row and the first signal holding unit 507A are reset.

After this, similarly to the N-th vertical period, the accumulation operation of 1/480 seconds per time is intermittently performed four times during the times t11 to t13. These are shown as accumulation periods 602-1', 602-2', 602-3', and 602-4' in the upward-sloping hatched area. Further, the signal charge of the photodiode at this time is similarly accumulated and held in the first signal holding unit 507A.

Then, from the time t13, the vertical synchronizing signal φV becomes the high level in the timing generating unit 189, and the horizontal synchronizing signal φH becomes the high level at the same time, and the N+3rd vertical period is started.

In the (N+3)th vertical period, the same operation as in the (N+1)th cycle is performed. First, the pixel signal accumulated and added in the first signal holding unit 507A in the (N+2)th vertical period is externally output as the image signal of the first row. Is output. Further, between times t13 and t14, a 1/480 second accumulation operation is performed four times intermittently and periodically. The signals of the photodiodes accumulated in the accumulation periods 702-1', 702-2', 702-3', and 702-4' in the region of the lower right diagonal lines are accumulated/held in the second signal holding unit 507B. ..

As described above, the signal charges generated in the photodiode 500 are accumulated and held in the first signal holding unit 507A during the Nth and N+2nd vertical periods, and the N+1th and N+3rd vertical periods, respectively. Read out during a predetermined horizontal period. On the other hand, the signal charges held in the (N+1)th and (N+3)th vertical periods in the second signal holding unit 507B are output to the outside during a predetermined horizontal period in the (N)th and (N+2)th vertical periods, respectively. ..
In the present embodiment, each pixel includes two signal holding units, and while the charge of the photodiode 500 is accumulated and added in one signal holding unit, reading can be performed from the other signal holding unit. , The degree of freedom in setting the exposure time is increased.

Next, the operation of the second and subsequent rows will be described.
In FIGS. 4 and 5, the operation of all rows is executed in synchronization with the vertical synchronizing signal and the horizontal synchronizing signal.
The drive pulse for the m-th row can be represented as φTX3(m), φTX1A(m), φTX1B(m), φSEL(m), φRES(m), φTX2A(m), φTX2B(m).
Further, in the present embodiment, basically, for φTX3(m), φTX1A(m), and φTX1B(m), signal patterns of the same timing are used for almost all rows.

As a result, the photoelectric conversion units of almost all rows are basically exposed at a common timing (exposure pattern 1), and a so-called global shutter system is realized. Moreover, for some rows (for example, the n-th row), the pulses of the exposure pattern 2 at different timings from the φTX3(m), φTX1A(m), and φTX1B(m) of the exposure pattern 1 are used. .. This makes it possible to prevent deterioration of image quality due to the influence of reset noise.

Next, drive pulses φTX3(2), φTX1A(2), φTX1B(2), φSEL(2), φRES(2), φTX2A(2), and φTX2B(2) in the second row where m=2 will be described. ..
As shown in FIG. 4, among the drive pulses in the second row, φSEL(2), φRES(2), φTX2A(2), and φTX2B(2) are φSEL(1) and φRES(1) in the first row. , ΦTX2A(1), φTX2B(1) are delayed by ΔTh.
ΔTh is set to a value that is equal to or longer than the time (time for one pulse in the horizontal synchronizing signal ΦH) required to read out the signal charges of one row by the horizontal scanning circuit.

Similarly, φSEL(m), φRES(m), φTX2A(m), and φTX2B(m) in the m-th row are delayed from the signal in the first row by the amount of time (m−1)×ΔTh.
Further, assuming that the total number of rows of pixels of the image sensor 184 is Ma rows, it takes about Ma×ΔTh or less to read all the rows. In order to read the data of all rows in one vertical period, ΔTh is set so that Ma×ΔTh<vertical period T (1 vertical period=1/30 seconds in FIGS. 4 and 5).

Then, the reading of each row is sequentially and continuously performed during one vertical period. That is, in the Nth vertical period, the signal charges accumulated and held in the second signal holding unit 507B during the (N-1)th vertical period are shifted line by line between times t1 and t6. Meanwhile, all the rows are sequentially output to the outside through the signal output line 523. In the (N+1)th vertical period, the signal charges accumulated and held in the first signal holding unit 507A during the Nth vertical period are shifted line by line between times t6 and t11. Meanwhile, all the rows are sequentially output to the outside through the signal output line 523. In the N+2nd vertical period (t11 to t13), the same operation as the Nth vertical period is performed, and in the N+3rd vertical period (t13 to t14), the same operation as the N+1th vertical period is performed and a signal is read to the outside. Similar operations are alternately performed in the subsequent vertical periods.

In FIGS. 4 and 5, the total exposure amount in one vertical period is 1/120 seconds in total, which is 1/4 of the total exposure amount for 1/30 seconds which is one vertical period. Since the exposure time is reached, the light reduction effect of ND2 steps can be obtained.
By appropriately setting the exposure (photoelectric conversion)/accumulation time 1/480 seconds per time, the total exposure amount can be controlled and the dimming effect can be arbitrarily adjusted.
For example, when the above-described exposure/accumulation/transfer operation is repeated four times during a vertical period of 1/30 seconds, the maximum exposure time per one time is 1/4 of one vertical period (1/30 seconds). It is 1/120 second.

Further, when the exposure/accumulation/transfer operation is repeated four times during the vertical period of 1/30 seconds, when the exposure time per one time is 1/240 seconds, the total exposure time becomes 1/60 seconds. .. In this case, there is an effect of one stage of ND that reduces the light amount to 1/2.
Similarly, when the exposure time per one time is 1/480 seconds, the total exposure time is 1/120 seconds, which is the effect of ND2 steps that makes the light amount 1/4.
Similarly, when the exposure time per one time is 1/960 seconds, the effect of three ND steps that the light amount is 1/8, and when the exposure time per one time is 1/1920 seconds, the light amount is 1 The effect of 4 stages of ND that is /16 is obtained.

In addition, in the present embodiment, as shown in FIGS. 4 and 5, the exposure operation is alternately performed a plurality of times in each vertical period. At this time, the timing and the number of times of the exposure operations in each vertical period are performed. Alternatively, the exposure time per exposure may be different.
As described above, since one image signal is obtained by adding the short accumulation times set at, for example, substantially equal intervals during the vertical period, a high-quality moving image without a frame-by-frame flapping feeling. Can be obtained.
Although the above embodiment has been described on the assumption that the number of times of accumulation and addition of moving images is four, it can be applied to cases of 8, 16, 32, and 64 times, for example.

Note that, as described above, when the switching operation noise for turning on/off the fifth transfer transistor 503 is superimposed during the output of the moving image signal to the outside, the luminance of the moving image fluctuates and the image quality deteriorates. is there. This will be described with reference to the timing chart of FIG.

FIG. 6 shows a timing chart when the noise is superposed. The same numbers are used for the numbers corresponding to those in FIGS. 4 and 5.
FIG. 6 differs from FIG. 5 in that the signal patterns for controlling the exposure and accumulation of φTX3, φTX1A, and φTX1B are the same in all rows.
In FIG. 6, in the Nth vertical period, the transfer pulse φTX3(n) of the first row continues to be in the high level state after the vertical period starts at time t1, and the fifth transfer transistor 503 is in the ON state. Then, at time t3, φTX3(n) becomes low level, the fifth transfer transistor 503 is turned off, and at time t5, φTX3(n) returns to high level and the fifth transfer transistor 503 is turned on. Switching back.

At this time, at time t3, φTX3(n) changes from the high level to the low level, and when the fifth transfer transistor 503 is turned off, a large switching noise occurs. In the Nth vertical period, φTX3(n) falls from the high level to the low level as described above, that is, the moment when the fifth transfer transistor 503 changes from the ON state to the OFF state, the same number of times as the number of accumulations. It exists only four times in FIG. Similarly, in the N+1th and subsequent cycles, four times, which is the same as the number of accumulations, exist in one vertical period. The timing of these four instants is indicated by a white downward arrow as noise generation timing 1 in the n-th row of FIG. The N-th, N+1-th, N+2-th, and N+3-th vertical periods are present in each of the four vertical periods, at four locations equal to the number of accumulations.

Further, while the charge of the photodiode 500 is being accumulated and added to the first signal holding unit 507A, for example, the charge held in the second signal holding unit 507B in the previous vertical period is passed through the signal output line 523 row by row. Output to the outside in order. The external output operation is sequentially performed from the first row to the final row by shifting by ΔTh which is a minute time, and the output of all rows is completed within one vertical period. At this time, at the timing of outputting the signal charges to the outside in the m-th row (the period in which φTX2A(m) or φTX2B(m) is at the high level), the same signal in the first row is output to (m-1). It is the time when ×ΔTh has passed (delayed).

Then, when the rows are different, the instant when φTX2A(m) or φTX2B(m) that outputs the signal charges to the outside becomes high level also differs. Among the rows having different external output timings, there is a row in which the timing of the external output coincides with the moment when φTX3(m) where switching noise largely occurs falls from the high level to the low level. In FIG. 6, it is shown as the nth row.

In this row, the switching noise affects the signal charge output to the outside, and an error occurs in the luminance value in all the pixels in one row. FIG. 7 shows an example of an output image when driving is performed as shown in the timing chart of FIG. 6 and the influence of switching noise is exerted. Suppose you have taken a gray image with a uniform screen. At this time, there are four rows in which the above-described edge where φTX3(n) falls from the high level to the low level coincides with the period in which φTX2A(n) or φTX2B(n) that outputs the signal charge to the high level becomes the high level ( (n1-n4 lines). In the pixel signals of the n1 to n4 rows, as shown in the figure, a luminance error is likely to occur over all of the respective rows, the luminance is lowered only in that row, and four horizontal lines are formed in the image, resulting in a poor image quality. You can see that it is getting worse. Even in other cases, noise is likely to be superposed when the edge of φTX3(n) and the timing (phase) of the high level period of φTX2A(n) or φTX2B(n) match.

In one vertical period, there is a moment when φTX3(m) in which switching noise is largely generated for the number of times of accumulation (four times in the figure) falls to a low level and then returns to a high level. Therefore, the rows in which the noise generation coincides with the external output timing and the error occurs may appear at the same number as the maximum number of times of accumulation.
In the embodiment shown in FIG. 5, the deterioration of image quality due to the luminance error caused by the noise can be eliminated.

That is, in the nth row in which an error has occurred in the example of FIG. 6, φTX3(n), φTX1A(n), and φTX1B(n) in FIG. 5 are only Δt from the patterns (exposure pattern 1) of other rows. The difference is that the delayed exposure pattern 2 is used. That is, a different operation pattern is used in which at least the operation switching time by the fifth transfer unit is delayed by a predetermined time when compared with other rows.

On the other hand, for the first, second, and m-th rows other than the n-th row, the exposure pattern 1 of φTX3(m), φTX1A(m), and φTX1B(m) at the same timing as in the example of FIG. 6 is used.
At this time, the signal patterns for controlling the video signal output to the outside of φSEL(n), φRES(n), φTX2A(n), and φTX2B(n) in FIG. 5 are the same in all rows including the nth row. , And the same timing signal as in FIG. 6 is used.

The row to which the exposure pattern 2 needs to be applied (nth row in the example of FIG. 5) can be determined from the division accumulation timing in the exposure pattern 1 and the external output timing of each row. The line number information to which the exposure pattern 2 is applied is stored in the memory unit 190, and the system control CPU 178 reads the information at the time of image pickup to control the operation of the image pickup device.
With this configuration, it is possible to prevent the influence of switching noise on the moving image output signal.

That is, in the timing chart of the nth row in FIG. 5, the moment when φTX3(n) at which switching noise largely occurs falls from the high level to the low level is indicated by a black arrow as the noise generation timing 2. As in the case of the white arrow at the noise generation timing 1 in FIG. 6, there are four pieces each, which is the same number as the number of accumulations, that is, four times in one vertical period. In the timing chart of the nth row of FIG. 5, the signal pattern of φTX3(n) is delayed by Δt compared to the example of FIG. Therefore, the black arrow at the noise generation timing 2 at which φTX3(n) falls from the high level to the low level is also delayed by Δt from the white arrow at the noise generation timing 1 in the example of FIG. In FIG. 5, the signal pattern of φTX3(m) at the same timing as that in the example of FIG. 6 is supplied to the photoelectric conversion units of a predetermined plurality of rows other than the n-th row. Therefore, as shown in the m-th row in FIG. 5 as an example, noise is generated at the timing of noise generation timing 1 indicated by the white arrow.

As an example, focusing on the (N+1)th cycle in FIG. 5, the first one of the timings indicated by the black arrows at which φTX3(n) in the timing chart of the nth row falls from the high level to the low level is time t8. It is located at a time delayed by Δt from. In the m-th row, which is the same timing as in the example of FIG. 6, the first timing at which φTX3(m) falls from the high level to the low level, which is indicated by a white arrow in the N+1th cycle, is time t8, and only Δt thereafter. Timing is late.

In FIG. 6, the timing of outputting a moving image signal to the outside in the n-th row (the timing of φTX2A(n) in the (N+1)th cycle) overlaps with the time t8. Therefore, since it is the same as the noise generation timing 1 indicated by the white arrow, the output moving image signal is affected by noise and the image quality deteriorates.
On the other hand, in the n-th row of FIG. 5, the timing of outputting the moving image signal to the outside (the timing when φTX2A(n) becomes the high level in the (N+1)th cycle) is the same time t8 as described above. However, the timing at which noise is generated is shifted by Δt from the noise generation timing 2 indicated by the black arrow. As a result, as shown in the drawing, the read period of the read pulse and the generation timing of the reset noise do not coincide with each other, so that the moving image signal on the n-th row is not affected by the noise and the deterioration of the image quality can be prevented.

That is, for photoelectric conversion units of a predetermined plurality of rows other than the n-th row of FIGS. 4 and 5, φTX3(m), φTX1A(m) of the exposure pattern 1 of the common timing similar to the example of FIG. 6, A signal pattern for controlling the exposure/accumulation of φTX1B(m) is supplied in the vertical period. However, even if noise is generated at the noise generation timing 1 indicated by the white arrow in the m-th row of FIG. 5, the noise generation timing and the external output timing do not overlap except for the n-th row, as in the example of FIG. Therefore, the moving image signal is not affected by noise and the image quality is not deteriorated. Further, φTX3(n) (exposure pattern 2) which is a reset pulse having a timing different from that of the exposure pattern 1 is supplied to the photoelectric conversion units of a predetermined row (for example, n rows) different from the predetermined plurality of rows. ing. Therefore, the moving image signal of the nth row or the like is not affected by noise, and the deterioration of image quality can be prevented.
As described above, since the two exposure patterns are switched, the error that has occurred in the nth row or the like due to noise in the example of FIG. 6 is eliminated in the embodiment of FIG. 5 and the deterioration of the moving image quality is prevented. be able to. In the present embodiment, the two exposure patterns are the same and only the phases are shifted, but the two exposure patterns may be different not only in the phase but also in the pulse width. That is, a part of the pulse width of φTX3(n) of the exposure pattern 2 may be widened and the other pulse width of φTX3(n) may be narrowed within one vertical period. In that case, for example, the total exposure time in one vertical period of φTX3(m) of the exposure pattern 1 and the total exposure time in one vertical period of φTX3(n) of the exposure pattern 2 are set to be the same. As another method of shifting the phase, for example, the number of times φTX3(n) in one vertical period may be changed, or the pulse interval of four times in one vertical period may be changed. In that case, the pulse interval may be gradually shortened or lengthened within one vertical period. When the φTX3(n) pattern can be selected from a plurality of types as described above, the number of the n-th row changes according to each pattern. Information is stored in the memory unit 190. Then, the system control CPU 178 may read the line number information for each pattern of φTX3(n) stored in the memory unit 190 and control the operation of the image sensor.

In the present embodiment, in the nth row, φTX3(n), φTX1A(n), and φTX1B(n) are delayed by Δt with respect to the same signal pattern (exposure pattern 1) in the other rows. In comparison with exposure pattern 1, the exposure/accumulation time is delayed by Δt. Therefore, when a moving subject is photographed, an image in which the subject is shifted by the subject moving speed V×the delay time Δt is output to the outside only in the row in which the exposure pattern 2 is applied to the row in which the exposure pattern 1 is applied. ..

FIG. 8 is a schematic diagram showing an example of an image pickup result when an image of the Shinkansen (subject S) moving at a speed V in the right direction of the paper is photographed using the image pickup apparatus for driving the present embodiment shown in FIG. As described above, it is assumed that the exposure pattern 1 is applied to all rows except the nth row, and the exposure pattern 2 delayed by the time Δt is applied only to the nth row.
As shown in the figure, since the exposure timing is delayed by Δt only in the image on the n-th row, the outline of the Shinkansen is shifted in the traveling direction by the speed V×the delay time Δt.
At this time, if the value of Δt is large and the amount of deviation of the contour of the subject image is large, the quality of the output moving image is degraded. In an extreme case, in the subject image in the row in which the exposure timing is delayed and the subject images in the other rows, the two subjects are too misaligned in the traveling direction on the screen, and there is no overlapping portion, so that there are two locations. Looks like.

Therefore, in the embodiment of FIG. 5, the exposure/accumulation signal pattern is configured as follows so that the subject images in different rows are at least partially overlapped. In other words, at least a part of the exposure/accumulation timing of the photodiode 500 in the exposure pattern 1 (the period during which the fifth transfer transistor 503 is in the OFF state) and the exposure/accumulation timing of the exposure pattern 2 that is deviated by the time Δt overlap. I have to. In FIG. 5, a period in which φTX3(m) in the m-th row is at a low level and a period in which φTX3(n) at the n-th row delayed by Δt is at a low level partially overlap. That is, Δt is set shorter than the non-reset period in which the reset pulse is not reset (the period in which the fifth transistor is in the OFF state).

Focusing on the (N+1)th vertical period, in the m-th row, the non-reset period in which φTX3(m) is at the low level is the period from time t8 to time t10. In the n-th row, the non-reset period in which φTX3(n) is at the low level is the period from time (t8+Δt) to time (t10+Δt). At this time, if Δt has a relationship of Δt<(t10−t8), at least a part of both time regions are common. In the present embodiment, Δt satisfies the above relationship, and a part of both periods in which the fifth transfer unit is in the OFF period (non-reset period) is common in the rows shifted by the time Δt. There is. By configuring the subject images to be exposed at the same timing as described above, it is possible to prevent the images from being separated for each row and reduce the deterioration of the image quality. That is, at least part of the period in which the fifth transfer unit in the row delayed by the predetermined time is in the non-reset state and the period in which the fifth transfer unit in the other row is in the non-reset state overlap each other I have to.
Further, if the number of lines in which the contour of the image is deviated is large, the quality of the moving image is similarly degraded.

Therefore, in the embodiment of FIG. 5, the rows to which the exposure pattern 2 for delaying the exposure/accumulation timing by Δt is applied are the rows in which the luminance value has an error due to the influence of the noise, and the number of rows in which the contour is displaced occurs. Is minimized. This keeps the quality of the video as high as possible. That is, only in the row where the timing of outputting the moving image signal to the outside and the timing of turning on the fifth transfer unit are substantially the same, the timing of turning on the fifth transfer unit is delayed by the time Δt. Image degradation is minimized.

Further, the time Δt is affected by noise only for a time of about one pulse of the horizontal synchronizing signal ΦH in the above-mentioned pulse control, so that it is the shortest time of two pulses without the effect. Is desirable. The pulse signal is generated by the timing generation unit 189 described above. When a pulse signal is emitted at a pulse generation frequency of 200 kHz, one pulse is 1/200,000 seconds, and the time for two pulses is 1/200,000 seconds, which is 1/100,000 seconds. Considering that the fastest shutter speed of a general digital camera is about 1/8000 second, it can be said that 1/100,000 second of the time Δt is a very short time. With such a delay time, the deviation of the contours of the subject in each row in the exposure pattern 1 and the exposure pattern 2 is at a level that does not pose a problem for almost all general subjects such as cars, bullet trains, and sports scenes. As described above, in the present embodiment, the delay time amount Δt is set to the shortest time period in which no noise influence occurs, that is, the time period for two pulses in the pulse control, so that the contour deviation of the subject for each row is minimized. The video quality is maintained as much as possible.

That is, the predetermined time for delaying the operation switching time is preferably one pulse signal or longer (1/200,000 seconds or longer) and two pulse signals or shorter (1/100,000 seconds or shorter).
It should be noted that the magnitude of the motion vector of the subject is detected from the image, and it is determined when the magnitude of the motion vector is larger than a predetermined value, that is, when the motion of the subject is large, and in that case, the delay time amount Δt is reduced. May be controlled as follows.

Further, in the above embodiment, the exposure/accumulation is repeated a plurality of times, but the present invention can be applied to the case where the exposure/accumulation is performed only once in one vertical period.
In the above embodiment, the transfer pulses φTX3(n), φTX1A(n), and φTX1B(n) on the n-th row as reset pulses have the phases of, for example, φTX3(n−1) and φTX1A(in the previous row). n-1) and φTX1B(n-1) were delayed by a predetermined time. For that purpose, the system control CPU 178 as the control means controls the phase change of the reset pulse.
However, instead of delaying the phase of the transfer pulse φTX3(n), the phase of the pulse for reading may be delayed. That is, the phase of φSEL(n), φRES(n), φTX2A(n) or φTX2B(n) is immediately before the phase of φSEL(n-1), φRES(n-1), φTX2A(n-1) or φTX2B(n-). You may delay from 1) for a predetermined time. In that case, the read signal may be AD-converted and then stored in a digital memory, and the delay may be returned to the original when the signal is read from the memory. That way, the screen is not adversely affected. In that case, the system control CPU 178 controls the phase change of the pulse for reading.
With the configuration of the embodiment as described above, in a device that shoots a moving image while adjusting the exposure amount by repeating exposure/accumulation a plurality of times, the occurrence of a brightness error due to noise due to the signal switching operation at the time of repetition is prevented. It is possible to provide an image pickup apparatus that can be suppressed.
Although the preferred embodiments of the present invention have been described above, the present invention is not limited to these embodiments, and various modifications and changes can be made within the scope of the gist thereof.

Moreover, you may make it supply a computer program which implement|achieves a part or all of control in a present Example with the function of the Example mentioned above to an imaging device or an information processing apparatus via a network or various storage media. Then, the computer (or the CPU, MPU, or the like) in the imaging device or the information processing device may read out and execute the program. In that case, the program and the storage medium storing the program constitute the present invention.

500 Photodiode 507A as photoelectric conversion unit First signal holding unit 507B Second signal holding unit 300, 301 Pixel unit 184 Image sensor 178 System control CPU as control means
151 Imaging device body

Claims (15)

A plurality of photoelectric conversion units arranged in rows and columns for photoelectrically converting incident light to generate charges,
In order to sequentially read out the charges generated by the plurality of photoelectric conversion units row by row,
Read-out means for sequentially supplying read-out pulses having different phases for each row;
Reset pulse supply means for supplying to the photoelectric conversion unit a reset pulse for resetting the photoelectric conversion unit during a vertical period,
Control means for shifting the phase of the reset pulse or the phase of the read pulse for a predetermined time when reading a predetermined row so that the noise of the reset pulse does not overlap the read period of the read pulse. Characteristic imaging device. The image pickup apparatus according to claim 1, wherein the control unit delays the phase of the reset pulse by a predetermined time when reading a predetermined row. The image pickup apparatus according to claim 1, wherein the control unit delays the phase of the read pulse by a predetermined time when reading a predetermined row. The image pickup apparatus according to claim 1, further comprising a signal holding unit for accumulating charges generated by the photoelectric conversion unit. The first signal holding unit and the second signal holding unit are provided for each photoelectric conversion unit in order to accumulate the charges generated by the photoelectric conversion unit. Imaging device. The read-out means reads out the charges accumulated in the second signal holding unit while the charges generated in the photoelectric conversion unit are accumulated in the first signal holding unit. The imaging device according to item 5. The image pickup device according to claim 4, wherein charges are transferred from the photoelectric conversion unit to the signal holding unit and accumulated in a period during which the photoelectric conversion unit is not reset by the reset pulse in the vertical period. apparatus. The image pickup apparatus according to claim 1, wherein the reset pulse supply unit supplies a reset pulse for resetting the plurality of photoelectric conversion units a plurality of times during the vertical period. The imaging device according to claim 7, further comprising a transfer unit that transfers charges from the photoelectric conversion unit to the signal holding unit while the photoelectric conversion unit is not reset. The image pickup apparatus according to claim 1, further comprising a reset period in which the photoelectric conversion unit is reset by the reset pulse and a non-reset period in which the photoelectric conversion unit is not reset, and the predetermined time is shorter than the non-reset period. The image pickup apparatus according to claim 1, wherein the predetermined time is 1/200,000 seconds or more and 1/100,000 seconds or less. The image pickup apparatus according to claim 1, wherein when the magnitude of the motion vector of the subject is larger than a predetermined value, the predetermined time is controlled to be shortened. The image pickup apparatus according to claim 1, further comprising a memory that can select a plurality of types of patterns as the reset pulse and that stores the number information of the predetermined row for each pattern. The image pickup apparatus according to claim 1, wherein a transistor driven by the read pulse and a reset transistor driven based on the reset pulse are provided for each photoelectric conversion unit. A first transfer transistor provided between the photoelectric conversion unit and the first signal holding unit; and a second transfer transistor provided between the photoelectric conversion unit and the first signal holding unit. The imaging device according to claim 5, wherein

Images