JP2000165752A - Signal processing method for solid-state image pickup device, and solid-state image pickup device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To effectively correct bright and dark horizontal stripes generated on a picture image picked-up under a fluorescent lighting by a CMOS sensor. SOLUTION: A signal from an image sensor part 10 is connected through an amplifier 15 to an A/D converter 18 for converting an analog signal into a digital signal. A line memory A20 and a line memory B21 are connected through a selector A19 to the A/D converter 18. A frame memory 23, gain correction coefficient calculator 24, multiplier 26, and selector C27 are connected through a selector B with the line memories A20 and B21. Also, a coefficient memory 25 is connected with a gain correction coefficient calculator 24. A camera signal processing part 28 is connected to a selector C27.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a signal processing method for a solid-state imaging device using a CMOS image sensor and a solid-state imaging device.

[0002]

2. Description of the Related Art In recent years, new markets for image sensors such as digital still cameras and cameras for personal computer TV conferences have been opened. Conventionally, a CCD has been used as an image sensor of a TV camera. However, when a camera is mounted on a portable device driven by a battery, the power consumption of the CCD is largely unsuitable, and a low power consumption MOS image sensor is used. It is getting attention.

When a MOS type image sensor is operated under illumination of a fluorescent lamp, the operation is different from that of a CCD,
A problem occurs in which flicker of light and dark horizontal stripes occurs. Hereinafter, this phenomenon will be described. As shown in FIG. 13, it is known that the light emission intensity of the fluorescent lamp changes in a cycle twice the power supply frequency. Therefore, the blinking period is 50 Hz.
100 Hz in the power supply area of 1 Hz and 1 in the area of 60 Hz.
20 Hz. When the MOS image sensor is driven at a frame rate of 30 frames / sec (that is, a vertical frequency of 30 Hz) under such a fluorescent lamp, bright and dark horizontal stripes as shown in FIG. 14 are generated (so-called flicker). FIG.
, The area A is dark and the area B is bright.

[0004] The mechanism of the occurrence of light and dark will be described with reference to the schematic diagram of FIG. The vertical period (33 ms) shown in FIG.
ec) is the time for reading the pixel cells from top to bottom, and the signal reading time in the area A in FIG. At this time, if the exposure time Δt shown in FIG. 15 is set, the exposure start time is AS in the area A and BS in the area B, and the shaded area in FIG. 15 represents the total exposure amount of the areas A and B. It can be seen that the area becomes smaller in the region A. Moreover, when the vertical period is not an integral multiple of the flicker cycle of the fluorescent lamp,
There is a problem that bright and dark horizontal stripes flow in the vertical direction and image quality degradation is conspicuous.

In order to make light and dark horizontal stripes invisible, in principle, the signal output level may be adjusted by a variable amplifier along the vertical direction of the screen. However, concrete measures for solving the following problems have not been devised yet.

(1) Normally, bright and dark horizontal stripes flow in the vertical direction, so that the amount of gain adjustment is not constant even if the horizontal line is at the same position. If the reading time of one screen is set to an integral multiple of the flickering cycle of the fluorescent lamp, the position of the bright and dark horizontal stripes does not flow. However, it is not necessarily synchronized with the power supply frequency. In this case also, position detection is required.

(2) When horizontal stripe components of a picture itself and horizontal stripes caused by flickering of a fluorescent lamp are mixed, it is difficult to accurately detect a change in signal level of only a bright and dark horizontal stripe portion to be corrected.

[0008]

As described above, the MO
When the S-type image sensor is used under a fluorescent lamp, so-called fluorescent lamp flicker (light and dark horizontal stripes) occurs, which causes a problem that image quality is impaired, and a problem that there is no effective correction means.

An object of the present invention is to provide a signal processing method for a solid-state imaging device and a solid-state imaging device capable of suppressing the occurrence of bright and dark horizontal stripes generated in the vertical direction of an image and improving the image quality.

[0010]

Means for Solving the Problems [Configuration] The present invention is configured as follows to achieve the above object. (1) A signal processing method for a solid-state imaging device according to the present invention (claim 1) includes a plurality of photoelectric conversion units that are two-dimensionally arranged in a row direction and a column direction to generate an electric signal according to an amount of incident light; A vertical signal line from which an electric signal from the conversion unit is sequentially read for each column; a vertical control unit for controlling reading of an electric signal generated by the photoelectric conversion unit to the vertical signal line; and the vertical control unit A signal processing method for a solid-state imaging device comprising: a horizontal control unit that controls a horizontal transfer of an electric signal read to a signal line; and a first electric signal read from the photoelectric conversion unit; A gain correction for a change in luminance of the amount of light incident on the photoelectric conversion unit provided in the row direction from the second electric signal read from the photoelectric conversion unit at a time Δt shifted from the reading time of the first electric signal. Calculate coefficient A step, based on the gain correction coefficient, characterized in that it comprises a step of correcting the second electric signal.

The time Δt is equal to the first signal S read from the photoelectric conversion unit and the photoelectric conversion at which the first signal S is read at a time Δt ′ from the read time of the first signal S. ΔS from the second signal S ″ read from the unit
Are obtained by changing the time Δt ′, and a plurality of ΔS obtained by changing the time Δt ′
And determining a cycle of the luminance change by comparing

(2) A solid-state image pickup device according to the present invention (claim 3), a plurality of photoelectric conversion units arranged two-dimensionally in a row direction and a column direction to generate an electric signal according to the amount of incident light, and the photoelectric conversion unit A vertical signal line from which an electric signal from the unit is sequentially read out for each column, a vertical control unit for controlling reading of an electric signal generated by the photoelectric conversion unit to the vertical signal line, and the vertical signal unit A first signal read from the photoelectric conversion unit and a first signal read from the photoelectric conversion unit, the solid-state imaging device including: a horizontal control unit that controls a horizontal transfer of an electric signal read to a line; Time Δt from time
From the second signal read from the photoelectric conversion unit from which the first signal has been read, the gain correction coefficient for the luminance change of the amount of light incident on the photoelectric conversion unit provided in the row direction is calculated. It is characterized by comprising gain correction coefficient calculation means and correction means for correcting the second signal based on the calculated gain correction coefficient.

(3) The solid-state imaging device according to the present invention (claim 4) includes a plurality of photoelectric conversion units arranged two-dimensionally in a row direction and a column direction to generate an electric signal according to the amount of incident light; A vertical signal line from which an electric signal from the unit is sequentially read out for each column, a vertical control unit for controlling reading of an electric signal generated by the photoelectric conversion unit to the vertical signal line, and the vertical signal unit A solid-state imaging device including a horizontal control unit that controls the transfer of an electric signal read to a line in a horizontal direction, wherein a first signal S read from the photoelectric conversion unit and a first signal S Means for obtaining a plurality of differences ΔS from the second signal S ″ read from the photoelectric conversion unit from which the first signal S has been read at a time Δt ′ shifted from the read time by changing the time Δt ′. And the time Δ
means for determining a period of a luminance change by comparing a plurality of ΔS obtained by changing t ′.

The first signal S read from the photoelectric conversion unit and the first signal S read from the photoelectric conversion unit from which the first signal S was read are shifted by t time from the reading time of the first signal S. By comparing a plurality of differences ΔS with the second signal S ″ by changing the time Δt ′ and a plurality of ΔS obtained by changing the time Δt ′, the cycle of the luminance change can be determined. And a means for determining the time Δt from the determined cycle of the luminance change, wherein the time Δt is (2N + 1) / 2 with respect to the blink cycle _Tf .
(N is an integer).

[Operation] The present invention has the following operation and effects by the above configuration. From the first electric signal and the second electric signal read from the photoelectric conversion unit shifted by a predetermined time from the read time of the first electric signal, the amount of incident light to the photoelectric conversion unit provided in the row direction from the first electric signal By calculating a gain correction coefficient corresponding to the change in luminance and correcting the second electric signal based on the calculated gain correction coefficient, it is possible to suppress a change in luminance of the incident light amount.

The first signal S read from the photoelectric conversion unit
And the second signal S ″ read from the photoelectric conversion unit from which the first signal S was read at a time Δt ′ shifted from the read time of the first signal S, and the difference ΔS between the predetermined time Δt '
Is changed and the plurality of ΔS obtained by changing the predetermined time Δt ′ are compared to detect the cycle of the luminance change.

[0017]

Embodiments of the present invention will be described below with reference to the drawings. [First Embodiment] FIG. 1 is a block diagram showing a schematic configuration of a MOS type imaging device according to a first embodiment of the present invention. In the image sensor unit 10 in which a plurality of pixel cells 11 described later are arranged in a row direction and a column direction, a group of pixel cells 11 connected in a column direction (horizontal line) is sequentially selected, and a pixel selected as a vertical signal line described later is selected. A vertical register (vertical control unit) 12 for controlling transfer from the cell 11 to an electric signal is connected. Pixel cell 1 selected in vertical register 12
Scanning unit 13 that samples and holds a group of image signals
Are connected to the image sensor unit 10. A horizontal register (horizontal control unit) 14 for sequentially reading the image signals of the pixel cells 11 of the horizontal line held by the scanning unit 13
Is provided. An amplifier 15 for amplifying the image signal is connected to the horizontal register 14.

The vertical register 12 and the horizontal register 1
4 is driven by a pulse supplied from the timing generator 16. Image sensor unit 10, pixel cell 11, vertical register 12, scan unit 13, horizontal register 14, amplifier 15, timing generator 16
Are all formed on one semiconductor substrate.

The timing generator 16 has T
A clock CLK, a horizontal read reference pulse HP, a vertical read reference pulse VP, and an electronic shutter pulse ESR are supplied from the G / SG 17.

Here, the electronic shutter pulse ESR is
This pulse sets the exposure time of each pixel cell 11 so that the pixel signal does not saturate when the light to the image sensor unit 10 is strong. This exposure time is set to a common value for each pixel.

An A / D converter 18 for converting the signal amplified by the amplifier 15 from an analog signal to a digital signal is connected to the amplifier 15. The line memory A20 is connected to the A / D converter 18 via the selector A19.
And a line memory B21. A frame memory 23, a gain correction coefficient calculator 24, a multiplier 26,
And the selector C27 are connected to each other. Further, a coefficient memory 25 is connected to the gain correction coefficient calculator 24. The camera signal processing unit 28 is connected to the selector C27.

FIG. 2 is a diagram showing an equivalent circuit of a typical pixel cell portion. The pixel cell 11 includes the photodiode 3
1, a transfer gate (read gate) 32, an amplification transistor 33, an address transistor 34, a reset transistor 35, and a detection node 36.

The read operation will be described briefly. Light condensed by a lens (not shown) is applied to the photodiode 31, and is photoelectrically converted into electronic charges (or hole charges, here, electronic charges) in accordance with the amount of light applied. The electronic charge Q photoelectrically converted and stored for a predetermined storage time (shutter time) is read from the photodiode 31 to the detection node 36 via the transfer gate 32 by applying a read pulse Φread to the transfer gate 32. And converted into a voltage signal Vsig by the parasitic capacitance C of the detection node 36. The signal Vsig is applied to the address transistor 37 by applying an address selection pulse Φaddr to the address line 37 to turn on the address transistor 34.
3 to the vertical signal line 38. The signal charges are accumulated for one frame period, and the signal is read out during the horizontal blanking period. The potential of the detection node 36 is reset to an external voltage VDD by a reset transistor 35 before or after reading in order to generate a predetermined reference voltage.

First, before describing the operation of the present apparatus, the principle of the signal processing method of the present apparatus will be described with reference to FIG. FIG.
FIG. 3 is a diagram showing a read timing of the MOS imaging device according to the first embodiment of the present invention.

[0025] In Figure 3, VD is a vertical synchronizing signal, representing the period in T _v. Further, VP is a vertical read reference pulse of the sensor, and is set to be active within the width of the vertical synchronization period VD in normal read.

The sensor output under fluorescent lamp illumination has a sinusoidal amplitude variation as shown in FIG. And
The flicker cycle of the fluorescent lamp is represented by _Tf , and _Tf = 10 ms for a 50 Hz power supply and _Tf = about 8.3 ms for a 60 Hz power supply.
It is.

A feature of the present invention is that a vertical synchronization period is provided between adjacent normal vertical read reference pulses V _p to start sensor reading by a vertical read reference pulse V _{p ′} for comparison shifted in position. Is to be. The time of the vertical read reference pulse V _{p ′} for comparison is such that the period T _p between the subsequent normal vertical read reference pulse V _p and T _p = (2N + 1) × T _f / 2 (N: integer) (1) Determine that the relationship is substantially satisfied.

FIG. 3 shows the vertical read reference pulse V _{p ′} for comparison when N = 1, and the read line of the sensor is 1, 2 starting from the position of the vertical read reference pulse V _{p ′} for comparison. , 3, ... Here, an output X _i of the i-th line read by the vertical read reference pulse V _{p 'for} comparison, the output Y _i of the i-th line read by the regular vertical read reference pulse Vp is, Tp The reading time is shifted by time. Therefore, the sensor output change component due to the fluorescent lamp illuminance change between the output X _i and the output Y _i has the same absolute value and a different sign.

[0029] Therefore, when calculating the _{Zi = (Y i + X i} ) / 2, the sensor output variation value by the fluorescent lamp illumination changes are canceled, the net of the subject signal is obtained when the subject is stationary.

Next, a specific calculation will be described below. The signal level S _real (m, n) of the mth pixel in the horizontal direction and the nth pixel in the vertical direction on the sensor screen under steady illumination
Is determined as follows.

[0031]

(Equation 1)

Here, Δt is the shutter time, and P is the amount of energy (constant) incident on the pixel (m, n) from the subject per unit time.

The signal level received by a pixel when a sinusoidal vertical illuminance change is given will be described. When there is a sinusoidal change in the illuminance, the energy of light is proportional to the square of the amplitude of the light. Therefore, the signal incident on the pixel from the subject at time t is P · sin ² (2πt / T _f ) (2) Become. Accordingly, the sensor output signal S (m, n) output from the pixel (m, n) in a state where there is a sinusoidal change in illuminance is as follows.

[0034]

(Equation 2)

The pixels in the horizontal direction are read out at the same time, so that there is no influence of the change in illuminance by the fluorescent lamp.

By the way, from time t _n , (2N + 1) T _f
/ 2 (= t _p) shifted by a time t 'pixels in the _n (m,
n), the comparison sensor output signal T (m, n) output from
It looks like this:

[0037]

(Equation 3)

Then, when the arithmetic mean S _A of S (m, n) and T (m, n) is calculated, S _A (m, n) = (S (m, n) + T (m, n) ) / 2 = S _real (m, n) (6) Therefore, the arithmetic mean S _{A of} S (m, n) and T (m, n)
It can be seen from the calculation that the signal S _real (m, n) which is not affected by the illuminance change due to the fluorescent lamp can be obtained.

By the way, from the equation (6), it can be seen that a signal having no change in illuminance can be obtained by simply taking the average. There is. Further, the average value can be calculated only for the period _Tf for one cycle of the illuminance change.

Therefore, in a period _Tf of one cycle of the illuminance change of the fluorescent lamp, a gain correction coefficient averaged in the horizontal pixel is obtained, and the gain of each horizontal line is corrected in the cycle of the illuminance change.

The gain correction coefficient G (n) of the n-th line is

[0042]

(Equation 4)

Is defined as Here, _me is the number of pixels in the horizontal direction. Then, a corrected output S _comp (m, n) is obtained by calculating S _comp (m, n) = S (m, n) · G (n) (8).

When S (m, n) and T (m, n) are constant irrespective of m, naturally, G (n) = S _A (m, n) / S (m, n) ) (9), and S _comp (n) = S _A (m, n) (10)

Next, the operation of the present apparatus will be described with reference to FIGS. FIG. 4 is a timing chart in units of vertical scanning, and FIGS. 5 and 6 are timing charts in units of horizontal scanning. Note that FIG. 5, normal prior to the vertical read reference pulse V _p, the horizontal scanning period of the vertical scanning period in VD for reading a comparative sensor output signal T described above by the vertical read reference pulse V _{p 'for} comparison HD a timing diagram of the unit, FIG. 6 is a diagram showing a timing of the horizontal scanning period HD units within a vertical scanning period VD for reading the sensor output signal S by the vertical read reference pulse V _p legitimate.

[0046] In the following description, the comparison sensor output signal T described above (m, n), expressed as comparative sensor output signal T _n together the signals of all the pixel cells of the n-th line. Also, in the aforementioned sensor output signal S ′ (m, n),
Together signals of all the pixel cells of the n-th line is represented as S _'n. Further, the above-mentioned signal S _{A (m, n)} is represented as Scomp _{(m, n),} and the corrected signals of all the pixel cells on the n-th line are collectively expressed as S comp _{(m, n).}
expressed as comp _n .

In the timing charts of FIGS. 4, 5, and 6, the write operation and the read operation for the signal S _x to the line memories A20, B21 or the frame memory 23 are denoted by S _x (W) and S _x (R), respectively. ing.

[0048] First, TG / SG17 from normal read time, to generate a vertical read reference pulse V _{p 'for} comparison shift time T _p, to start reading of the comparison sensor output signal T. The comparison sensor output signal T from the image sensor unit 10 is sequentially converted into a digital signal. Then, the comparison sensor output signal T converted into a digital signal
However, signals for one horizontal scan are alternately recorded in the line memory A20 or the line memory B21 via the selector A19. In the timing chart of FIG.
The portions indicated by R are R (read) / W (write) of the input / output of the comparison sensor output signal T to the line memories A20 and B21.
Period.

FIG. 5 shows the section E of FIG. 4 in units of horizontal synchronization, and the capture of the comparison sensor output signal T will be described in more detail with reference to the timing chart of FIG.
First, in the horizontal synchronization period HD ₁ by the horizontal synchronizing signal HD,
By connecting the A / D converter 18 and the line memory A20 with the selector A19, the A / D-converted comparison sensor output signal T (m, 1) is transferred to the line memory A2.
0 Yuki sequentially writes, the writing comparative sensor output signal T ₁ of the the pixel sensors of the first line in the line memory A20.

[0050] In addition, in the second half of the horizontal synchronization period HD _1,
By connecting the line memory B21 and the frame memory 23 by the selector B22, the line memory B
The comparison sensor output T ₀ written in 21 is written in the frame memory 23.

[0051] Then, in the horizontal synchronization period HD ₂ by the next horizontal synchronizing signal HD, by connecting an A / D converter 18 and the line memory B21 by the selector A19, similarly to the case of the comparative sensor output signal T ₁ , Second
Line A comparative sensor output signal T ₂ of the line memory B2
Write to 1.

[0052] In addition, in the second half of the horizontal synchronization period HD _2,
By connecting the line memory A20 and the frame memory 23 by the selector B22, the line memory A
The comparison sensor output T ₁ written in 20 is written in the frame memory 23.

[0053] That is, in the horizontal scanning period HD _n, the selector A19 is a line memory A20 when n is an odd number
Is connected to the A / D converter 18, and when n is an even number, the line memory B21 is connected to the A / D converter 18 so that the comparison sensor output signal T converted into a digital signal is output.
Write _n to the line memory A20 or the line memory B21. Then, the line memory A20 or the line memory B
In parallel to the writing operation of the comparative sensor output signal T _n to 21, the selector B22 is, n is connected to the line memory B21 and the frame memory 23 in the case of odd, a line memory A20 when n is an even number By connecting to the frame memory 23, the comparison sensor output signal T _n-1 written in the line memory A20 or the line memory B21 is written in the frame memory 23.

Then, the same operation is sequentially performed, and the first to Nth operations are performed.
_{The f-} line comparison sensor output _Tn is stored in the frame memory 23.
Write to. Note that the N _f, the number of lines in the vertical direction are read out during the blink cycle T _f time.

[0055] Then, into the vertical synchronization period in which the output gain correction in the vertical read reference pulse V _p of the following regular. The operation in the vertical synchronization period for performing the output gain correction will be described with reference to FIG. In the output correction period of FIG. 4, S (1), S (2), S (3), and S (4) are shown for each blink cycle _Tf .

[0056] First of all, in the horizontal synchronization period HD _1, selector A
19, the line memory A20 and the A / D converter 1
By connecting the 8, A / D-converted sensor output signal S (m, 1) Yuki sequentially writes the line memory A20, and writes the sensor output signals S ₁ from the first line to the line memory A20.

In the _first half of the horizontal synchronizing period HD1, the selector B22 connects the gain correction coefficient calculator 24 to the frame memory 23 and the line memory B21, so that the sensor output signal written in the frame memory 23 is obtained. S ₀ and the comparison sensor output signal T ₀ written in the line memory B 21 are transferred to the gain correction coefficient calculator 24 in synchronization with the horizontal pixel position (the read clock is doubled).

The gain correction coefficient calculator 24 calculates T (m, 0) from the comparison sensor output signal T (m, 0) and the sensor output signal S (m, 0) read out in synchronization with the horizontal pixel position. 0)
/ S (m, 0) is calculated. Then, after calculating T (m, 0) / S (m, 0) for all the horizontal pixel sensors, the gain correction coefficient G ₍₀₎ is calculated based on the above equation (7), and the calculated gain correction is performed. The coefficient G ₍₀₎ is stored in the coefficient memory 25.

Next, in the period of the second half of the horizontal scanning period HD _1, by connecting the line memory B21 and the gain correction coefficient calculator 24 by the selector B22, reads the sensor output signal S _0, the coefficient memory 25 , The gain correction coefficient G ₍₀₎ is read out, and the multiplier 26 multiplies the sensor output signal S (m, 0) by the gain correction coefficient G _{(0 )} to obtain a corrected sensor output signal Scomp _{(m, 0)} . Then, the correction sensor output signal Scomp _{(m, 0)} is sequentially transferred from the multiplier 26 to the camera signal processing unit 28 via the selector C27. At the same time, the correction sensor output signal Scomp
_{(m, 0)} is sequentially written from the multiplier 26 to the frame memory 23 via the selector B22.

[0060] Then, in the next horizontal synchronization period HD _2, by connecting the line memory B21 and A / D converter 18 by the selector A19, the sensor output signal S
As with the _first write, writing the sensor output signal S ₂ from the second line to the line memory B21.

[0061] Further, in the first half of the horizontal scanning period HD _2, by connecting the gain correction coefficient calculator 24 and the frame memory 23 and the line memory A20 by the selector B22, the sensor output signal S (m, 1) and comparative The sensor output signal T (m, 1) is transferred to the gain correction coefficient calculator 24 while synchronizing the horizontal pixel position.

The gain correction coefficient calculator 24 calculates T (m, 1) / S (m, 1) from the comparison sensor output signal T _{(m , 1)} and the sensor output signal S _{(m, 1)} of the same pixel. 1) Calculate Then, after calculating T (m, 1) / S (m, 1) for all the horizontal pixels, the gain correction coefficient G ₍₁₎ is calculated based on the above equation (7), and the calculated gain correction coefficient is calculated. G ₍₁₎ is stored in the coefficient memory 25.

[0063] Further, in the period of the second half of the horizontal scanning period HD _2, by connecting the line memory A20 and the gain correction coefficient calculator 24 by the selector B22, it reads the sensor output signal S _1, a coefficient memory 25 gain correction coefficient read out G _(1), the multiplication of the sensor output signals S ₁ and the gain correction coefficient G ₍₁₎ by the multiplier 26 obtains a corrected signal output Scomp _{(m, 1).} And
The corrected signal output Scomp _{(m, 1)} is transferred from the multiplier 26 to the camera signal processing unit 28 via the selector C27. At the same time, the corrected signal output Scomp _{(m, 1)} is output to the multiplier 2
6 to the frame memory 23 sequentially through the selector B22.

Then, the same operation is performed to obtain the corrected output signal Scomp _{(m, n)} . That is, in the horizontal scanning period HD _n, the selector A19 is, n is connected to an A / D converter 18 and the line memory A20 in the case of an odd number,
In the case of an even number, the A / D converter 18 and the line memory B2
By connecting the 1 and, A / D-converted sensor output signal S _n of the line memory A20 or the line memory B
Write to 21.

[0065] and, in the front half of the horizontal scanning period HD _n,
When n is an odd number, the selector B22 controls the gain correction coefficient calculator 24, the frame memory 23, and the line memory B21.
And when n is an even number, the gain correction coefficient calculator 2
4 and the frame memory 23 and the line memory A20, the horizontal pixel position is synchronized, and the sensor output signal S _n-1 and the comparison sensor output signal T _n-1 are sent to the gain correction coefficient calculator 24. Forward. Then, the gain correction coefficient calculator 24 calculates a gain correction coefficient G _(n-1), and stores the calculated gain correction coefficient G _(n-1) to the coefficient memory.

Next, in the period after a half of the horizontal scanning period HD _n, selector B22 is, n is connected to the line memory B21 and the gain correction coefficient calculator 24 for odd line if n is an even number By connecting the memory A20 and the gain correction coefficient calculator 24, the sensor output signal S
_{At the same time} as reading _n-1 , the gain correction coefficient G _(n-1) is read from the coefficient memory 25, and the multiplier 26 multiplies the sensor output signal S _n-1 by the gain correction coefficient G _(n-1) .
The corrected signal output Scomp _n-1 is obtained.

The correction coefficient G _(n) is obtained by reading the comparison sensor output signal T _n in the frame memory only for one period T _{f of the} illuminance change, that is, in the section S ₍₁₎ .

[0068] Subsequent _{S (2), S (3} ), in S ₍₄₎ section, speed in the second half period of S ₍₁₎ segment as well as the horizontal scanning period of the sensor output signal S _n from the line memory A20, B21 And the corresponding correction coefficient G _(n) is read from the coefficient memory 25, and the correction signal Scomp
_{Find n} . Then, the obtained gain correction signal Scomp
_n is transferred to the camera signal processing unit 28 via the selector C27 and, at the same time, to the frame memory 23 via the selector B22.

The gain correction signal Scomp _n written in the frame memory 23 is supplied to the signal FR ₁ during the timing of the frame memory in FIG. ', F
R ₂ ′, FR ₃ ′, FR ₄ ′ are read in the first half of each horizontal synchronization period HD. With this operation, the gain correction signal in the previous vertical synchronization period can be extracted even in the gain correction comparison output capturing period.

In the above operation, writing to and reading from the frame memory 23 is performed at the reference data rate of 2
However, when there is a limitation on the cycle time of the frame memory 23, if the data of two pixels is accessed in parallel (for example, an 8-bit width is changed to a 16-bit width), it is not necessary to double the speed. .

[0071] According to the solid-state imaging device of the first embodiment as described above, one frame memory, in a configuration comprising two line memories, comparative sensor output by changing the oscillation position of the vertical read reference pulse V _p vertical synchronization period to obtain the vertical read reference pulse V _p by the vertical synchronizing period of the normal, in any, can be taken out of the sensor output obtained by correcting the fluorescent lamp illumination changes.

Since the line memory can be easily built in the IC, the configuration of one frame memory is advantageous in cost. However, in this case, output is performed at a double data rate only in the first half or the second half of the horizontal period. However, in most cases, since a line memory for color separation is provided in the camera signal processing unit, it is easy to return the data rate to the original and make the final video output continuous.

In a case where a moving image of a camera is compressed and transmitted, a frame memory is used as a rate buffer due to a restriction of a transmission rate. However, the above-mentioned frame memory can also be used.

[Second Embodiment] With a configuration having two frame memories, a write operation and a read operation can be performed at the same time, so that a line memory becomes unnecessary, and a write / read operation to the frame memory requires a double speed. Sex disappears. Therefore, in the present embodiment, an example of a MOS-type imaging device including two frame memories will be described with reference to FIGS.

FIG. 7 is a block diagram showing a schematic configuration of a MOS type imaging device having two frame memories.
7, the same parts as those in FIG. 1 are denoted by the same reference numerals, and detailed description thereof will be omitted. FIG. 8 is a diagram showing a timing chart for explaining the operation of the solid-state imaging device of FIG.

The A / D converter 1 is connected to the selector D61.
8, a frame memory A62, a frame memory B63, a gain correction coefficient calculator 24, and a multiplier 26 are connected to each other. The A / D converter 1 is connected to the frame memory A62 and the frame memory B63 by the selector D61.
8 is switched.

In the timing chart at the vertical synchronization interval shown in FIG. 8, during the period of taking in the comparison output for gain correction, the output T of the comparison sensor from the A / D converter 18 is obtained.
_n is input to the frame memory A62 via the selector D61.
Is written to. Then, the output gain correction period, the sensor output signal S _n, at the same time written to the frame memory B63 via the selector D61, are transferred to the gain correction coefficient calculator 24.

[0078] gain correction coefficient in synchronism with the transfer of the sensor output signal S _n to the calculator 24, the transfer from the corresponding comparative sensor output signal T _n to the frame memory A62 via the selector D61 to the gain correction coefficient calculator 24 Then, the gain correction coefficient G _(n) is calculated and stored in the coefficient memory 25.

In the next period of taking in the comparison output for gain correction, the sensor output signal S written in the frame memory B63 is read.
_n is transferred to the multiplier 26 via the selector D61, and the corresponding gain correction coefficient G _(n) is read out and multiplied by the multiplier 26 to correct the output fluctuation due to the flickering of the fluorescent lamp. A gain correction signal Scomp _n is obtained. Then, the obtained gain correction signal Scomp _n is transferred to the camera signal processing unit 28.

When two frame memories A62 and B63 are provided, as shown in the timing chart, writing and reading are performed alternately in the frame memory B63 every vertical scanning period, so that the gain correction signal Scomp
_n is an intermittent output every vertical scanning period. If intermittent output is assumed in the compression of a moving image, such intermittent output may be used. However, if continuous output is desired, a dual port product may be used as a frame memory or another frame memory may be added.

[Third Embodiment] A system which does not need to output a video output every frame, for example, an application for transmitting a moving image at a low rate of 15 frames / sec or less instead of 30 frames / sec in real time. The form is shown below.

FIG. 9 is a block diagram showing a main configuration of a solid-state imaging device according to the third embodiment of the present invention. 9, the same parts as those in FIG. 1 are denoted by the same reference numerals, and the detailed description thereof will be omitted.

As shown in FIG. 10, the operation of the present apparatus is as follows. As shown in FIG. 10, the signals from the image sensor unit are read out one by one in a frame unit, such as S ₁ and S _2. T is read and the frame memory 23 is read out.
Write to. When the comparison sensor output signal T is read, signals from all lines are read instead of reading from some lines as in the above embodiment.

[0084] Further, with reference to the timing chart of FIG. 11, illustrating a signal behavior for the sensor output signal S _2. At the same time as writing the sensor output signal S ₂ to the line memory B 21 via the selector E 81, the sensor output S ₂ and the frame memory 29 are output by the gain correction coefficient calculator 24.
Performing the calculation of the correction coefficient from a comparison sensor output T ₂ Metropolitan of the same pixel read from.

When a color filter is provided on the surface of a pixel cell, the signal level change due to the fluorescent lamp flicker is caused for the different color filters on one horizontal line due to the bias of the spectral distribution of the fluorescent lamp and the like. Therefore, it is necessary to obtain a gain correction coefficient for each color filter pixel. Therefore, in the present apparatus, a gain correction coefficient is obtained for each color filter pixel, and the result is written to each of the other coefficient registers A82 or B83. In the RGB Bayer array, which is a color filter array generally used in general, a line in which R (red) and G (green) are alternately arranged in one horizontal line, and G (green) and B (blue) Are two types of arrangement with alternately arranged lines, so that the coefficient register only needs to have two systems.

Next, the sensor output signal S ₂ is read from the line memory B 21, and the gain coefficient written in the coefficient registers A 83 and B 84 is selected by the selector F 84 in accordance with the color of the pixel, and the gain is calculated by the multiplier 26. A correction signal is required.

The above embodiment deals with a general case where the flicker cycle of the fluorescent lamp does not have a multiple relationship with the vertical synchronizing cycle. Therefore, for each scene of the vertical synchronization period to be taken out, before the regular normal vertical synchronization period for taking in the output from the pixel cell, the output of the comparison sensor is taken in every time by the vertical reading reference pulse _{Vp '} for comparison. I was However, when the flicker cycle of the fluorescent lamp and the vertical scanning cycle are set to a multiple relationship (for example, a flicker cycle of 10
ms, the vertical synchronization period 30 ms), since it is sufficient correspondence to slow changes in brightness location due to frequency drift, for example, once every 10 times, the vertical read reference pulse V _p
What is necessary is just to take in the sensor output for comparison which shifted the position.

[Fourth Embodiment] In the above embodiment, the correction process was performed on the assumption that the fluorescent lamp blinking period _Tf is known. However, actually, the region where the power supply frequency is 50 Hz and the region where the power supply frequency is 60 Hz are used. There is. Therefore, in the present embodiment, an imaging apparatus including a unit that automatically detects the fluorescent lamp blink period _Tf will be described.

First, the principle of automatically detecting the flicker cycle _Tf of the fluorescent lamp will be described. The normal read output S (m, n) read at the time t _n is given by:

[0090]

(Equation 5)

Is obtained. Also, the normal read start time t _n
S ”at a time (t _n −t ₀ ) shifted by t ₀ time from
(m, n) is

[0092]

(Equation 6)

Is obtained. Then, S ″ (m, n) and S ′ (m, n)
The difference ΔS (m, n) from

[0094]

(Equation 7)

Is obtained. This ΔS (m, n) is t ₀ = (2N
+1) In the case of T _f / 2, the maximum value is obtained. Therefore,

[0096]

(Equation 8)

[0097] The time t _{0 (50 Hz)} = 10 ms (for 50 _Hz) and t _{0 (60 Hz)} = 8.3 ms (for 50 Hz)
Is applied and calculated. Note that N _f is the number of horizontal lines read during the blink cycle T _f . Then, determining [Delta] S _Ave. and _(50Hz) ΔS _{Av e.} Comparing the magnitudes of the _{(60 Hz),} for example, if _{ΔS Ave. (50Hz)> ΔS Ave.} (60Hz) power supply frequency is 50Hz Can be.

Next, an apparatus provided with a means for detecting the frequency of the blinking cycle will be described. FIG. 12 is a block diagram illustrating a schematic configuration of a solid-state imaging device according to the fourth embodiment of the present invention. In FIG. 12, the same parts as those in FIG. 1 are denoted by the same reference numerals, and detailed description thereof will be omitted.

First, the TG / SG 17 calculates the time difference between the comparison pulse V _{p ′} and the normal read reference pulse V _p by t
Each signal is sent to the timing generator 16 corresponding to _{0 (50 Hz)} . Then, comparative sensor output signal S "(m, n) and the sensor output signal S (m, n) is, .Derutaesu _Ave. calculation unit 111 to be transferred to the [Delta] S _Ave. calculation unit 111 compares sensor output signal S ”(m, n) and sensor output signal S
ΔS _{Ave. (50 Hz)} is obtained from (m, n) based on the equation (13), and the result is transferred to the T _f determining unit 112. The time difference between the comparison pulse V _{p ′} and the normal read reference pulse V _p is also calculated as ΔS by the time difference corresponding to t _{0 (60 Hz).
Ave.} seeking _{(60 Hz),} T _f determining unit 112 for transferring the result to the T _f determining unit 112, [Delta] S _Ave. and time difference t ₀ transferred
Determined from, [Delta] S _Ave. and _{(50H z)} and compares the [Delta] S _Ave. and _{(60 Hz),} for example, if _{ΔS Ave. (50Hz)> ΔS Ave.} (60Hz) power supply frequency is 50Hz .

As described above, according to the present apparatus, the blink period can be automatically detected, and there is no need to input the power supply frequency.

The present invention is not limited to the above embodiment, but can be implemented in various modifications without departing from the scope of the invention.

[0102]

As described above, according to the present invention, the first electric signal (comparative sensor output signal) and the first electric signal are read from the photoelectric conversion unit at a predetermined time from the read time. A gain correction coefficient for a change in luminance of the amount of light incident on the photoelectric conversion unit provided in the row direction is calculated from the second electric signal (sensor output signal) and a second correction signal based on the calculated gain correction coefficient. By correcting the electric signal of No. 2 with a correction unit such as a multiplier, a change in the luminance of the incident light amount can be suppressed.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of a solid-state imaging device according to a first embodiment.

FIG. 2 is a diagram illustrating a circuit configuration of a pixel cell of the solid-state imaging device in FIG. 1;

FIG. 3 is a diagram showing a timing chart for explaining the principle of the present invention.

FIG. 4 is a diagram showing a timing chart for explaining the operation of the solid-state imaging device in FIG. 1;

FIG. 5 is a timing chart for explaining the operation of the solid-state imaging device in FIG. 1;

FIG. 6 is a timing chart illustrating the operation of the solid-state imaging device in FIG. 1;

FIG. 7 is a block diagram illustrating a schematic configuration of a solid-state imaging device according to a second embodiment.

8 is a diagram showing a timing chart for explaining the operation of the solid-state imaging device in FIG. 7;

FIG. 9 is a block diagram illustrating a schematic configuration of a solid-state imaging device according to a third embodiment.

FIG. 10 is a diagram showing a timing chart for explaining the operation of the solid-state imaging device in FIG. 9;

FIG. 11 is a diagram showing a timing chart for explaining the operation of the solid-state imaging device in FIG. 9;

FIG. 12 is a block diagram illustrating a schematic configuration of a solid-state imaging device according to a fourth embodiment.

FIG. 13 is a diagram showing a flicker cycle of a fluorescent lamp.

FIG. 14 is a diagram showing a state of an image photographed under a fluorescent lamp using a conventional MOS image sensor.

FIG. 15 is a diagram illustrating a cause of flicker generation.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 10 ... Image sensor part 11 ... Pixel cell 12 ... Vertical register 13 ... Scanning part 14 ... Horizontal register 15 ... Amplifier 16 ... Timing generator 17 ... TG / SG 18 ... A / D converter 19 ... Selector A 20 ... Line memory A 21 ... Line memory B 22 ... Selector B 23 ... Frame memory 24 ... Gain correction coefficient calculator 25 ... Coefficient memory 26 ... Multiplier 27 ... Selector C 28 ... Camera signal processing unit 61 ... Selector D 62 ... Frame memory A 63 ... Frame memory B 81: Selector E 82: Coefficient register A 83: Coefficient register B 84: Selector F 111: ΔS _Ave. Calculation unit 112: T _f determination unit

Claims (4)

[Claims] A plurality of photoelectric conversion units arranged two-dimensionally in a row direction and a column direction to generate an electric signal according to an amount of incident light;
A vertical signal line from which an electric signal from the photoelectric conversion unit is sequentially read out for each column, a vertical control unit that controls reading of an electric signal generated by the photoelectric conversion unit to the vertical signal line, and the vertical control unit A signal processing method for a solid-state imaging device, comprising: a horizontal control unit configured to control a transfer of an electric signal read to the vertical signal line in a horizontal direction. The first electric signal read from the photoelectric conversion unit And a second electric signal read from the photoelectric conversion unit at a time Δt shifted from the read time of the first electric signal, based on a change in the amount of light incident on the photoelectric conversion unit provided in the row direction. A signal processing method for a solid-state imaging device, comprising: calculating a gain correction coefficient; and correcting a second electric signal based on the gain correction coefficient. 2. The time Δt is defined by a first signal S read from the photoelectric conversion unit and a first signal S.
The time difference ΔS from the second signal S ″ read from the photoelectric conversion unit from which the first signal S is read 11 is shifted by the time Δt ′ from the read time of the signal S, and the time Δt ′ is changed. 2. The method according to claim 1, further comprising: determining a plurality of times; and determining a period of a luminance change by comparing a plurality of ΔS obtained by changing the time Δt ′. 3. A signal processing method for a solid-state imaging device. 3. A plurality of photoelectric conversion units arranged two-dimensionally in a row direction and a column direction and generating an electric signal according to an amount of incident light;
A vertical signal line from which an electric signal from the photoelectric conversion unit is sequentially read out for each column, a vertical control unit that controls reading of an electric signal generated by the photoelectric conversion unit to the vertical signal line, and a vertical control unit A solid-state imaging device comprising: a horizontal control unit that controls a horizontal transfer of an electric signal read to the vertical signal line; a first signal read from the photoelectric conversion unit; From the second signal read from the photoelectric conversion unit from which the first signal is read by shifting the time Δt from the signal reading time to the luminance change of the amount of light incident on the photoelectric conversion unit provided in the row direction Gain correction coefficient calculating means for calculating a gain correction coefficient for, based on the calculated gain correction coefficient,
A solid-state imaging device comprising: a correction unit that corrects a second signal. 4. A plurality of photoelectric conversion units arranged two-dimensionally in a row direction and a column direction and generating an electric signal according to an amount of incident light;
A vertical signal line from which an electric signal from the photoelectric conversion unit is sequentially read out for each column, a vertical control unit that controls reading of an electric signal generated by the photoelectric conversion unit to the vertical signal line, and a vertical control unit A solid-state imaging device comprising: a horizontal control unit that controls a horizontal transfer of an electric signal read to the vertical signal line; a first signal S read from the photoelectric conversion unit;
The second signal read from the photoelectric conversion unit from which the first signal S has been read is shifted from the read time of the signal S by the time Δt ′.
Means for calculating a plurality of differences ΔS from the signal S ″ by changing the time Δt ′, and comparing a plurality of ΔS obtained by changing the time Δt ′ to determine a cycle of luminance change. A solid-state imaging device comprising: