JP3524112B2 - Solid-state imaging device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a solid-state image pickup device.
The present invention relates to an imaging device. 2. Description of the Related Art Solid-state imaging devices include color filters and individual
A single image sensor that forms an image via a color filter is provided.
I am. [0003] The light receiving element on the solid-state image pickup device
Color filters that transmit colored light are sequentially arranged to fix the subject image.
The image is formed on the body image sensor and spatially modulated by the solid-state image sensor.
Various control circuits are driven to obtain color and luminance signals
This is performed using FIG. 7 shows a color filter arrangement of a single-chip color image sensor.
FIG. [0005] To obtain a color signal, the following
Is performed to output a video signal. [0006] In the even-numbered field, the neighbors are located vertically.
Of the photosensitive elements on the n horizontal lines and the (n + 1) horizontal lines
Load and photosensitive elements of n + 2 horizontal line and n + 3 horizontal line
Charge, odd-numbered fields
Of the photosensitive elements on the n + 1 horizontal line and the n + 2 horizontal line
Load and photosensitive elements of n + 2 horizontal line and n + 4 horizontal line
Are added, transferred, and output. By the scanning,
Outputs a signal composed of the primary color signal component ratios as shown below.
Is done. [0007] Even field n line m row Mg + Cy = R + G + 2B; WB m + 1 row G + Ye = R + 2G; GR even field n + 1 line m row G + Cy = 2G + B; GB m + 1 row Mg + Ye = 2R + G + B; WR odd field C = 2G + B; GB m + 1 row Ye + Mg = 2R + G + B; WR even field n + 1 line m row Cy + Mg = R + G + 2B; WB m + 1 row Ye + G = R + 2G; GR
Signal Y = WB + GR / WR + GB is obtained in a time-division manner for each horizontal line, and the
A luminance signal is obtained by adding a predetermined synchronization signal. above
In processing, the signals that make up the luminance signal obtained for each line are
The level difference between lines in the sum signal.
The color filters have been adjusted to avoid this. The color signal between adjacent pixels in the horizontal direction is
The subtraction signal CB = WB-GR CR = WR-GB is obtained in a time-division manner for each horizontal line, and is stored in a memory such as a line memory.
The three types of signals are always obtained using a stage.
Y, CR, CB are multiplied by three predetermined numbers, and out of the three primary colors
One primary color component is extracted. Therefore red, gre
The en and blue primary color signals are processed by the following arithmetic processing.
Can be obtained. [0010] From the RGB signals, a predetermined matrix processing is performed.
By convention, the television signal standard RY / B-
A Y color difference signal is generated and output together with the luminance signal. Further, the RY / BY signal is converted into a quadrature two-phase signal.
Modulates the carrier color signal, which is a television signal standard,
The luminance signal, or the carrier chrominance signal and the luminance signal
Are added and output as a composite video signal. The four types of signals WB, WR, GR,
One of the three primary colors is obtained by multiplying GB by four predetermined numbers.
There is also a method of extracting a color component. Therefore red, gre
The primary color signals en and blue are obtained by performing the following arithmetic processing.
Can be ## EQU2 ## The four types of signals are obtained by the above-mentioned scanning method.
Therefore, they cannot be obtained at the same time because they are output in a time-division manner. So
Here, one horizontal line, one pixel delay element and selection circuit
To obtain the same kind of signal at all times. The image processing device has a pixel array,
When the primary color components are generated from the four signals output by the
A demodulated RGB signal can be obtained from a 4-pixel video signal in the vertical direction.
And For this reason, in an image with no vertical correlation
Causes interference such as false colors. Reduce the effects of this disturbance
Object with a vertical 4-pixel repetition cycle
Of the signal WB whose phase does not change with the field
g, G of G and the WB of Mg, G of G are 18
0deg different phase of GB CCD of GB and WR of Mg
So that the signal gain including the demodulation system from
Defines the coefficient of the arithmetic processing. Next, a configuration example of the color separation matrix will be described below.
Shown below. An image of a subject whose color separation output (R, G, B) is to be set so that (R, G, B) = (X, 0, 0) (0, Y, 0) (0, 0, Z) , Each subject
The following four types of signal data Sig (WB, W
When R, GR, GB) are obtained, Sigx = (WBx, WRx, GRx, GBx) Sigy = (WBy, WRy, GRy, GBy) Sigz = (WBz, WRz, GRz, GBz) Color separation matrix A = [Βiαij] (i = 0, 1, 2 j = 0, 1,
2,3) is a matrix satisfying the following equation. A * Sigx = (X, 0,0) A * Sigy = (0, Y, 0) A * Sigz = (0,0, Z) In particular, αi0 = (f1-f2αi3) / f0 αi1 = (e1-e2αi3) / e0 αi2 = (d1-d2αi3) / d0αi3 (i = 0,1,2) where d0 = b0a1-b1a0 d1 = b0a3-a0b3 d2 = b0a2-b2a0 e0 = a0 e1 = a3-a1d1 / d0 e2 = a2-a1d2 / d0 f0 = WBx f1 = ｛X-WRx e1 / e0-GRx d1 / d0 (i = 0) -WRx e1 / e0-GRx d1 / d0 (i = 1 , 2) f2 = GBx-WRx e2 / e0-GRx d2 / d0 where a0 = WByWRx-WBxWRy a1 = WByGRx-WBxGRy a2 = WBy Bx-WBxGBy a3 = {WByX (i = 0) -WByY (i = 1) 0 (i = 2) b0 = WBzWRx-WBxWRz b1 = WBzGRx-WBxGRz b2 = WBzGBx-WBxGBzb 0) 0 (i = 1) -WBxZ (i = 2) where βi is a so-called white balance process.
Coefficient for processing, for example, each of RGB of the entire subject screen
Factors that make the integrated values equal,
Coefficients that vary depending on the color temperature of the light source
You. Also, even if αi3 is arbitrary, the color separation characteristics are preserved.
It is. Therefore, the dynamic range of the signal is
Or two-dimensional frequency response determined by color filter arrangement
May be selected as desired. Color by the above processing
The separation matrix is uniquely determined. [0021] The conventional signal processing described above.
The hue adjustment that was performed in the 1. Adjustment of CR and CB signal gains 2. Adjustment of RY and BY gains For a linear matrix, etc., all require additional circuits or adjustment methods
Method and adjustment amount cannot be quantitatively analyzed in advance.
The drawback is that the coordinator has to
Was. Also, in the conventional example, the subject
The subject moving vertically or the imaging device
When the camera is moved vertically or panned vertically,
Invert every field if the imager vibrates vertically
Relatively less visually prominent color carriers suppress
Therefore, there is a disadvantage that interference such as false colors is not sufficiently suppressed.
Was. Further, the processing as in the above conventional example is performed.
Is the object whose color separation output (R, G, B) is desired to be such that (R, G, B) = (X, 0, 0) (0, Y, 0) (0, 0, Z). Below the color temperature of the imaged subject
Outside, the disadvantage is that the color separation characteristics change significantly
was there. The present invention solves the above-mentioned drawbacks, and realizes color separation processing.
Hue adjustment and good color separation characteristics
To provide a solid-state imaging device that can maintain
Aim. [0025] The present invention provides a solid-state imaging device.
Color filters that transmit different color lights on the photosensitive elements inside
Next, the subject image is formed on this solid-state image sensor,
The output signal corresponding to the different color light of the solid-state image sensor is specified.
To obtain the primary color components
In the body imaging device, the operation coefficient can be controlled from outside
And the motion vector of the object to be imaged is detected.
Operation to control the operation coefficient according to the motion vector detection output
It is characterized by having a coefficient control means. The operation of such a color separation method can be defined.
A quantitative hue adjustment is performed. Also, the motion vector of the video signal
Switch the coefficients of the color signal demodulation matrix
This allows vertical false colors even in video signals other than still images.
And other disturbances are suppressed. Further, a color for which good color separation characteristics are desired to be maintained.
In the temperature range T0 to T1, the temperature obtained at T0
Color separation matrix α0ijT obtained from mapping data
1. The color separation mat obtained from the subject data obtained in Step 1.
Rix α1ij is obtained, aij = k α0ij + (1−k) α1ij (k = 0 to 0)
1) Perform color separation using a matrix. EXAMPLE An RGB composition aij of an output signal of a solid-state imaging device is shown.
(I = 0,1,2 j = 0,1,2) is shown by the following equation.
Sea urn [Equation 3] The obtained hue adjustment coefficients are Ye, Cy,
Let hue gain coefficients on the MG axis be K0, K1, K2
The color separation matrix αij is defined as follows:
It is. [Equation 4] Where αi0 = ((r * y2-g * x2) * (y1 * z2-y2 * z1) − (g * z2-b * y2) * (x1 * y2-x2 * y1)) / ( (X0 * y2-x2 * y0) * (y1 * z2-y2 * z1)-(x1 * y2-x2 * y1) * (y0 * z2-y2 * z0)); αi1 = ((r * z2-b) * Y2)-(y0 * z2-y2 * z0) * α [i] [0]) / (y1 * z2-y2 * z1); αi2 = (g-α [i] [0] * y0-α [ i] [1] * y1) / y2; where, x0 = ccd_ [WB] [RD] + ccd_ [GR] [RD] or ccd_ [WR] + ccd_ [GB] [RD] y0 = ccd_ [WB] [GN] + ccd_ [GR] [GN] or ccd_ [ WR] [GN] + ccd_ [GB] [GN] z0 = ccd_ [WB] [BL] + ccd_ [GR] [BL] or ccd_ [WR] [BL] + ccd_ [GB] [BL] x1 = ccd_ [ WR] [RD] -ccd_ [GB] [RD]; y1 = ccd_ [WR] [GN] -ccd_ [GB] [GN]; z1 = ccd_ [WR] [BL] -ccd_ [GB] [BL]; x2 = ccd_ [WB] [RD] -ccd_ [GR] [RD]; y2 = ccd_ [WB] [GN] -ccd_ [GR] [GN]; z2 = ccd_ [W ] [BL] -ccd_ [GR] [BL]; ccd_ [WB] [RD] = (K1 * K2 * (a00 + a01) -K2 * K0 * (a01 + a02) + K0 * K1 * (a02 + a00)) / ( 2 * K0 * K1 * K2); ccd_ [WB] [GN] = (K1 * K2 * (a00 + a01) −K2 * K0 * (a01 + a02) + K0 * K1 * (a02 + a00)) / (2 * K0 *) Ccd_ [WB] [BL] = (− K1 * K2 * (a00 + a01) + K2 * K0 * (a01 + a02) + K0 * K1 * (a02 + a00)) / (2 * K0 * K1 * K2) Ccd_ [WR] [RD] = (K1 * K2 * (a10 + a11) −K2 * K0 * (a11 + a12) + K0 * K1 * (a12 + 10)) / (2 * K0 * K1 * K2); ccd_ [WR] [GN] = (K1 * K2 * (a10 + a11) −K2 * K0 * (a11 + a12) + K0 * K1 * (a12 + a10)) / (2 * K0 * K1 * K2); ccd_ [WR] [BL] = (− K1 * K2 * (a10 + a11) + K2 * K0 * (a11 + a12) + K0 * K1 * (a12 + a10)) / (2 * K0) * K1 * K2); ccd_ [GR] [RD] = (K1 * K2 * (a20 + a21) + K2 * K0 * (a21 + a22) + K0 * K1 * (a22 + a20)) / (2 * K0 * K1 * K2) Ccd_ [GR] [GN] = (K1 * K2 * (a20 + a21) + K2 * K0 * (a21 + a22) + K0 * K1 * (a22 + a2) )) / (2 * K0 * K1 * K2); ccd_ [GR] [BL] = (− K1 * K2 * (a20 + a21) + K2 * K0 * (a21 + a22) + K0 * K1 * (a22 + a20)) / ( 2 * K0 * K1 * K2); ccd_ [GB] [RD] = (K1 * K2 * (a30 + a31) + K2 * K0 * (a31 + a32) + K0 * K1 * (a32 + a30)) / (2 * K0 * K1) * K2); ccd_ [GB] [GN] = (K1 * K2 * (a30 + a31) + K2 * K0 * (a31 + a32) + K0 * K1 * (a32 + a30)) / (2 * K0 * K1 * K2); Ccd_ [GB] [BL] = (− K1 * K2 * (a30 + a31) + K2 * K0 * (a31 + a32) + K0 * K1 * (a32 + a30)) / (2 * K0 * K1 * K2); yields RGB signals with more desired adjustments.
You. Hereinafter, the present invention will be described in detail with reference to examples. FIG. 1 is a block diagram showing a first embodiment of the present invention.
It is a block diagram of a body imaging device. 1 is an imaging optical processing system such as a lens, and 2 is an imaging optical processing system.
Focus control system for controlling the focus of the image lens, 3
An aperture control system for controlling the aperture, an aperture for controlling the amount of incident light, 5
Is a photoelectric converter including a micro color separation filter having the color arrangement of FIG.
A two-dimensional color image sensor, which is a replacement element,
A driving circuit for driving the element, comprising a television signal
Within the image sensor during the period corresponding to the vertical blanking period of
Transfer the charge stored in the photoelectric conversion element to the vertical transfer unit
Signal and TV signal validity period for television signal
Horizontal transfer transferred from the vertical transfer unit during the period corresponding to
And a control signal for transferring the electric charge of the transmitting section. Reference numeral 7 denotes reset of an output signal from the image sensor,
Correlated double sampling circuit to remove clock noise
(CDS) and 8 connect the output of CDS7 to the control voltage input terminal.
Variable gain amplifier (A) that varies the gain according to the control voltage
GC) and 9 fix the black level of the input video signal to a predetermined voltage.
The clamp circuit 10 is controlled by the AGC output signal.
Generate luminance signal and color difference signal in vision signal
Signal processing circuit 11 for
A color difference signal RY.BY is generated from the obtained primary color signal.
Matrix processing section 12 for the signal processing circuit 10
Digital data related to the obtained signal
By the data obtained by processing the digital data,
Signal processing circuit 10, focus control system, incident light amount control system, gain
Micro control unit that outputs control signals to the control system
Knit (MCU). Reference numeral 13 denotes a color difference obtained from the signal processing circuit 10.
Generate a carrier chrominance signal that conforms the signal to a given broadcast standard
Chrominance signal modulation circuit. Numeral 14 designates digital signal data as analog signal.
D / A converter for converting into color signals, 15 is a carrier color signal output
Input terminal, 16 is a luminance signal output terminal, 17 is an adder, 18
Is an oscillator, and 20 is digital data corresponding to an input signal.
A / D converter for output, 100-1 is a timing controller
Output digital data according to the input signal
A / D conversion clock horizontal synchronization signal HD, vertical synchronization signal
VD, NTSC, PAL standard switching signal N / P,
Clock CLOCK of each frequency and phase required by the control unit
S, blanking pulse for television signal formation
BLK, burst flag pulse BF, color subcarrier SC,
Line sequential signal ALT, point sequential signal PALT, composite synchronization signal
Generate CSYNC or the like. FIG. 2 shows the signal processing circuit 10 shown in FIG.
FIG. 3 is a diagram showing one configuration example of a unit, where 101 is a fixed length.
(1 horizontal pixel) digital data delay circuit, 102
Digital data of fixed length (1 horizontal line-2 pixels)
Extension circuit 103 calculates the average of the input digital data.
Circuit, 104 is a switch, 105 is gamma correction
Gamma knee processing for level compression of high-brightness areas
Unit, 106 is a multiplier, 107 is an adder, 108 is smoothing
Circuit, 109 is a subtractor, 122 is an MCU,
The MCU captures the data signals required to control the
Data processing unit that performs pre-processing so that
is there. An object not shown is an imaging lens 1 and an aperture control system.
3, the amount of incident light is adjusted, and
Is accumulated as electric charges. The accumulated charge is a video signal
(Not shown) during the period corresponding to the vertical blanking period
The data is transferred to the vertical transfer unit in the image element. Transfer to vertical transfer unit
The charged electric charge is unillustrated in the image sensor during the horizontal blanking period.
To the indicated horizontal transfer unit. Transferred to horizontal transfer unit
The charge is output from the horizontal transfer section during the effective video period,
Is converted to pressure. Below is the sequence of the signal to be read
And the signal composition. Even field n line m row Mg '+ Cy' = xR + yG + zB; WBm + 1 row G '+ Ye' = xR + yG + zB; GR Even field n + 1 line m row G '+ Cy' = xR + yG + zB; GBm + 1 row Mg '+ Ye + xR; WR Odd field n line m row Cy '+ G' = xR + yG + zB; GB m + 1 row Ye '+ Mg' = xR + yG + zB; WR Odd field n + 1 line m row Cy '+ Mg' = xR + yG + zB; WB m + 1 row Ye '+ G' + G + xR; GR The whole signal processing will be described below. The data read from the image sensor by the above scanning is
The captured video signal is reset by the CDS 7 with the clock component.
After the cut noise is removed, the variable gain amplifier 8
Amplified with a gain according to the gain control signal,
To set the black level to approximately the lower limit of the input range of the A / D converter 20
Is converted to a digital data signal.
You. This digital signal is supplied to a one-pixel delay circuit 1
01 DP-1, DP-2 (horizontal line-2) horizontal pixels
Delay circuit 102 DL-1, DP-3, DP-4, DL
-2, input to delay circuit to which DP-5 and DP-6 are connected
The outputs from the delay elements are input to the averaging circuit 103 in a combination of A / D, DP2, DL2, DP6 DP1, DP5 DL1, and DP4. Average circuit output
And the DP3 output is input to the switching circuit 104.
You. The line sequence signal AL is input to the control signal input of the switching circuit.
T, a point sequence signal PALT is input, and
The same kind of color signal data series is output from the output of the replacement circuit
Is done. The output of the switching data is connected to 109.
The CR and CB signals are output. Also
The output of the delay element is also connected to the smoothing circuit 108.
1/4, 1/2, 1/4 gain between three adjacent pixels
Perform integration processing with a difference. This gives the Y signal
You. The Y, CR and CB signals are multiplied by a multiplier 106
And the other input of the multiplier is MCU
Twelve outputs are connected and a multiplier of the output of the switch
Can be controlled from the MCU. This MCU
Multiplied by the coefficient for the so-called white balance
The matrix coefficient calculated by the above method is output.
The result of multiplication of different signals among the outputs of this multiplier is added to the adder
The addition is performed at 107 to obtain a primary color signal component. The output signal data is subjected to gamma knee processing.
Gamma correction processing by the unit 105, the pressure of high-level data
Compression, and two types of color difference signals
Signal RY / BY signal. Each of the two color difference signal data is a modulator
13, a phase reference conforming to a predetermined broadcasting standard is colorized.
-Added according to the burst flag timing signal BF
That is, data that has the same amplitude as the signal data but has the opposite sign
Generate The above four data series are defined as sub-
A data sequence of opposite sign at a frequency four times the carrier frequency
So that they correspond to the four phases of the subcarrier
Output. This output is analyzed by the D / A converter 14.
Not shown, centered on the subcarrier frequency
Is input to the band pass filter. The luminance signal is a color component extracted image of the image sensor.
Lower the chrominance modulated carrier determined by the pixel array
It is removed by the filter 103-d, and gamma processing and knee processing are performed.
The luminance signal is generated by the luminance signal generator. As described above, in this embodiment, the single-chip camera
The hue adjustment of the signal color signal processing unit can be quantitatively performed.
It works. FIG. 3 is a block diagram showing a second embodiment of the present invention.
It is a block diagram of a body imaging device. The figure
1, the same parts as those shown in FIG.
Description is omitted. Reference numeral 30 denotes a motion vector detector used in this embodiment.
It is something that can be done. FIG. 4 shows one configuration of the signal processing circuit shown in FIG.
FIG. 3 is a diagram showing an example, the configuration of which is the circuit configuration shown in FIG.
Is almost the same as A subject (not shown) includes an imaging lens 1 and an aperture.
The amount of incident light is adjusted by the control system 3, and
The charge is stored in the storage unit. The accumulated charge is
Unexpected during the period corresponding to the vertical blanking period of the image signal
Is transferred to the vertical transfer unit in the image sensor shown in FIG. Vertical transfer unit
Charge transferred to the image sensor during the horizontal blanking period
To the horizontal transfer unit (not shown). Transfer to horizontal transfer unit
Charge within the effective video period
Output in the form and converted to voltage. Even field n line m row Mg + Cy = R + G + 2B; WB m + 1 row G + Ye = R + 2G; GR Even field n + 1 line m row G + Cy = 2G + B; GB m + 1 row Mg + Ye = 2R + G + B; WR odd field; = 2G + B; GB m + 1 row Ye + Mg = 2R + G + B; WR Even field n + 1 line m row Cy + Mg = R + G + 2B; WB m + 1 row Ye + G = R + 2G; GR
The captured video signal is reset by the CDS 7 with the clock component.
After the cut noise is removed, the variable gain amplifier 8
Amplified with a gain according to the gain control signal,
To set the black level to approximately the lower limit of the input range of the A / D converter 20
Is converted to a digital data signal.
You. This digital signal is supplied to a one-pixel delay circuit 1
01 DP-1, DP-2 (1 horizontal line-2) horizontal image
Elementary delay circuit 102 DL-1, DP-3, DP-4, D
L-2, DP-5, and DP-6 enter the delay circuit to which they are connected.
The output from each delay element is input to the averaging circuit 103 in a combination of A / D, DP2, DL2, DP6 DP1, DP5 DL1, and DP4. Average circuit output
And the DP3 output is input to the switching circuit 104.
You. The line sequential signal AL is input to the control signal input of the switching circuit.
T, a point-sequential signal PALT is input, and
The same kind of color signal data series is output from the output of the replacement circuit
Is done. The output of the switching data is sent to the multiplier 106.
MCU1 is input to the other input of the multiplier.
2 is connected, and the multiplier of the output of the switch is
It can be controlled from the MCU. Of this multiplier
Adder 107 adds the result of multiplication of different signals among outputs
To obtain a primary color signal component. A vertical pixel having a four-pixel folded frequency component
3) Color demodulation of 10 and 12 in FIG.
Occurs when performing color demodulation processing with a matrix (MAT)
Carriers to perform are Mg of WB, G of GR, and M of WB.
g, GR of GB 180 degrees different from G of GR
G, WR Mg, so WB Mg, GR G,
180 deg different from Mg of WB and G of GR
Phase GB G, WR Mg from CCD to demodulation system
The arithmetic processing is performed so that the gains of the included signals become equal.
Determine the number. An object having a vertical motion component
When color demodulation processing is performed by the color demodulation matrix of FIG.
The generated carriers are Cy of WB, Ye of GR, and W
180 deg phase difference from Cy of B and Ye of GR
Of the WB, and Ye of the WR,
180d with respect to Ye of the WB and Cy of the WB and GR
eg Cy of GB Cy, WR Ye CCD of different phase
So that the signal gains up to the demodulation system are equal.
A coefficient for the arithmetic processing is determined. The above matrix is used for the motion vector detector 3
The MCU performs vertical blur according to the vertical vector detection signal VV of 0.
During the inking period, it is sent to the color signal generation circuit. The output signal data is subjected to gamma knee processing.
Gamma correction processing by the unit 105, the pressure of high-level data
The primary color signal data is reduced by the color difference signal
Color difference signal data according to a predetermined ratio.
You. Each of the two color difference signal data is a modulator
13, a phase reference conforming to a predetermined broadcasting standard is colorized.
-Added according to the burst flag timing signal BF
That is, data that has the same amplitude as the signal data but has the opposite sign
Generate The above four data series are defined as sub-
A data sequence of opposite sign at a frequency four times the carrier frequency
So that they correspond to the four phases of the subcarrier
Output. This output is analyzed by the D / A converter 14.
Signal centered on the subcarrier frequency
Not input to the bandpass filter. Also the luminance signal
Is determined by the pixel arrangement of the color component extraction pixels of the image sensor.
The chrominance signal modulated carrier is removed by the low-pass filter 103, and the
The luminance signal is generated by the luminance signal generator that performs
Generated. In this embodiment, the vertical motion component detector is
Although the luminance signal is used as the detection means, the vertical
Any difference even when using physical detection means such as acceleration in the direction
I do not support it. As described above, the vertical motion level of the video signal
Switching the color demodulation matrix
Color images that do not interfere with vertical false colors other than still images
It becomes possible to obtain a signal. FIG. 5 is a third embodiment of the present invention, and FIG.
FIG. 3 is a diagram illustrating an example of a configuration of a signal processing circuit 10 shown in FIG.
You. A subject (not shown) includes an imaging lens 1 and an aperture.
The amount of incident light is adjusted by the control system 3, and
The charge is stored in the storage unit. The accumulated charge is
Shown in the period corresponding to the vertical blanking period of the image signal
The image data is transferred to the vertical transfer unit in the image sensor that does not use the image data. Vertical transfer
The charge transferred to the
Is sent to a horizontal transfer unit (not shown). Transfer to horizontal transfer section
The transmitted charges are output from the horizontal transfer unit within the effective video period
And converted to a voltage. In the following,
Shows the sequence and signal composition. Even field n line m row Mg '+ Cy' = xR + yG + zB; WBm + 1 row G '+ Ye' = xR + yG + zB; GR even field n + 1 line m row G '+ Cy' = xR + yG + zB; GBm + 1 row Mg '+ YG + xR; WR Odd field n line m row Cy '+ G' = xR + yG + zB; GB m + 1 row Ye '+ Mg' = xR + yG + zB; WR Odd field n + 1 line m row Cy '+ Mg' = xR + yG + zB; WB m + 1 row Ye '+ G' + G + xR; GR An example of the structure of the color separation matrix is shown below.
You. Color temperature range in which good color separation characteristics are desired to be maintained
In T0 to T1, the subject data obtained in T0
I want the color separation output (R, G, B) obtained from the data to be (R, G, B) = (X0, 0, 0) (0, Y0, 0) (0, 0, Z0) Take a picture of the subject
The following four types of signal data Sig (WB, W
When R, GR, GB) are obtained, Sig0x = (WB0x, WR0x, GR0x, GB0)
x) Sig0y = (WB0y, WR0y, GR0y, GB0
y) Sig0z = (WB0z, WR0z, GR0z, GB0
z) Color separation matrix A0 = α0ij (i = 0,1,2 j = 0,1,2,2)
3) is a matrix that satisfies the following equation. A0 * Sig0x = (X0,0,0) A0 * Sig0y = (0, Y0,0) A0 * Sig0z = (0,0, Z0) Particularly, αi0 = (f1-f2αi3 ) / F0αi1 = (e1-e2αi3) / e0αi2 = (d1-d2αi3) / d0αi3 (i = 0,1,2) where d0 = b0a1-b1a0 d1 = b0a3-a0b3 d2 = b0a2-b2a0 e0 = a0 e1 = a3-a1d1 / d0 e2 = a2-a1d2 / d0 f0 = WBx f1 = ｛X-WRx e1 / e0-GRx d1 / d0 (i = 0) -WRx e1 / e0-GRx d1 / D0 (i = 1, 2) f2 = GBx-WRx e2 / e0-GRx d2 / d0 where a0 = WB0yWR0x-WB0xWR 0y a1 = WB0yGR0x-WB0xGR0y a2 = WB0yGB0x-WB0xGB0y a3 = {WB0y X0 (i = 0) - WB0x Y0 (i = 1) 0 (i = 2) b0 = WB0zWR0x-WB0xWR0z b1 = WB0zGR0x-WB0xGR0z b2 = WB0zGB0x-WB0xGB0z b3 = ｛WB0z X0 (i = 0) 0 (i = 1) −WB0x Z0 (i = 2) Obtained from the subject data obtained at T1.
Of the object whose color separation output (R, G, B) is (R1, G1, B1) = (X1, 0, 0) (0, Y1, 0) (0, 0, Z1) And each subject
The following four types of signal data Sig (WB, W
When R, GR, GB) are obtained, Sig1x = (WB1x, WR1x, GR1x, GB1)
x) Sig1y = (WB1y, WR1y, GR1y, GB1
y) Sig1z = (WB1z, WR1z, GR1z, GB1
z) Color separation matrix A1 = α1ij (i = 0,1,2 j = 0,1,2,2)
3) is a matrix that satisfies the following equation. A1 * Sig1x = (X1,0,0) A1 * Sig1y = (0, Y1,0) A1 * Sig1z = (0,0, Z1) In particular, αi0 = (f1-f2αi3) ) / F0αi1 = (e1-e2αi3) / e0αi2 = (d1-d2αi3) / d0αi3 (i = 0,1,2) where d0 = b0a1-b1a0 d1 = b0a3-a0b3 d2 = b0a2-b2a0 e0 = a0 e1 = a3-a1d1 / d0 e2 = a2-a1d2 / d0 f0 = WBx f1 = ｛X-WRx e1 / e0-GRx d1 / d0 (i = 0) -WRx e1 / e0-GRx d1 / D0 (i = 1, 2) f2 = GBx-WRx e2 / e0-GRx d2 / d0 where a0 = WB1yWR1x-WB1xWR 1y a1 = WB1yGR1x-WB1xGR1y a2 = WB1yGB1x-WB1xGB1y a3 = {WB1y X1 (i = 0) - WB1x Y1 (i = 1) 0 (i = 2) b0 = WB1zWR1x-WB1xWR1z b1 = WB1zGR1x-WB1xGR1z b2 = WB1zGB1x-WB1xGB1z b3 = ｛WB1z X1 (i = 0) 0 (i = 1) −WB1x Z1 (i = 2) Color separation matrix A = [βiαij] (i = 0, 1, 2 j = 0, 1,
2,3) A0 (α0ij), A1 (α1ij) obtained from the above
From the color separation matrix A (αij), αij = βi (kα0ij + (1−k) α1ij) (i
= 0,1,2; j = 0,1,2,3) where βi is a so-called white balance process.
Coefficient for processing, for example, each of RGB of the entire subject screen
Coefficients and objects that illuminate
Coefficients that vary according to the color temperature of the light source
You. Also, even if αi3 is arbitrary, the color separation characteristics are preserved.
It is. Therefore, the dynamic range of the signal and the color
A two-dimensional frequency response determined by the
You just have to choose the number that will make you grow. K is the output of the color temperature of the subject.
Determined by conditions such as current frequency and color temperature that emphasize color separation characteristics
Confuse. Hereinafter, the entire signal processing will be described. The above-mentioned scanning reads out from the image sensor.
The captured video signal is reset by the CDS 7 with the clock component.
After the cut noise is removed, the variable gain amplifier 8
Amplified with a gain according to the gain control signal,
To set the black level to approximately the lower limit of the input range of the A / D converter 20
Is converted to a digital data signal.
You. This digital signal is supplied to a one-pixel delay circuit 1
01 DP-1, DP-2 (1 horizontal line-2) horizontal image
Elementary delay circuit 102 DL-1, DP-3, DP-4, D
L-2, DP-5, and DP-6 enter the delay circuit to which they are connected.
The output from each delay element is input to the averaging circuit 103 in a combination of A / D, DP2, DL2, DP6 DP1, DP5 DL1, and DP4. Average circuit output
And the DP3 output is input to the switching circuit 104.
You. The line sequence signal AL is input to the control signal input of the switching circuit.
T, a point sequence signal PALT is input, and
The same kind of color signal data series is output from the output of the replacement circuit
Is done. The output of the switching data is sent to the multiplier 106.
MCU1 is input to the other input of the multiplier.
2 is connected, and the multiplier of the output of the switch is
It can be controlled from the MCU. This MCU
Have multiplied the coefficient for white balance described above.
The matrix coefficient calculated by the above method is output. this
The multiplication result of the different types of signals among the outputs of the multiplier is added to the adder 10.
7 to obtain a primary color signal component. The output signal data is subjected to gamma knee processing.
Gamma correction processing by the unit 105, the pressure of high-level data
The primary color signal data is reduced by the color difference signal
Color difference signal data according to a predetermined ratio.
You. Each of the two color difference signal data is a modulator
13, a phase reference conforming to a predetermined broadcasting standard is colorized.
-Added according to the burst flag timing signal BF
That is, data that has the same amplitude as the signal data but has the opposite sign
Generate The above four data series are defined as sub-
A data sequence of opposite sign at a frequency four times the carrier frequency
So that they correspond to the four phases of the subcarrier
Output. This output is analyzed by the D / A converter 14.
Not shown, centered on the subcarrier frequency
Is input to the band pass filter. The luminance signal is a color component extracted image of the image sensor.
Lower the chrominance modulated carrier determined by the pixel array
It is removed by the filter 103-d, and gamma processing and knee processing are performed.
The luminance signal is generated by the luminance signal generator. Note that when actually performing the color separation processing,
A color separation matrix to perform
There is no level constraint between RGB.
Therefore, in the above matrix, X, Y, Z are other than 0
May be any real number. In the above embodiment, the color
Data approximating the separation matrix was obtained from an RGB pure-color subject in which (R, G, B) = (X, 0, 0) (0, Y, 0) (0, 0, Z). , G, B) = (X0, X1, 0) (0, Y0, Y1) (Z0, 0, Z1)
The color separation matrix can be configured by the method. As described above, according to the present embodiment, the
Maintains good color separation characteristics even when the color temperature of the object changes
An imaging device can be provided. The present invention has been described in detail based on the embodiments.
As described above, in the present invention, for example, the vertical motion vector
By switching the color demodulation matrix,
There is no interference such as vertical false color other than still image, and hue adjustment is accurate
Color video signal with good color separation characteristics
It is possible to obtain.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing a first embodiment of the present invention. FIG. 2 is a diagram showing an example of the internal configuration of a signal processing circuit 10 shown in FIG. FIG. 3 is a block diagram showing a second embodiment of the present invention. FIG. 4 is a diagram showing an example of the internal configuration of the signal processing circuit 10 shown in FIG. FIG. 5 is a block diagram showing a third embodiment of the present invention. FIG. 6 is a diagram illustrating a configuration example of the inside of the signal processing circuit 10 illustrated in FIG. 5; FIG. 7 is a diagram illustrating a color filter arrangement of a single-chip color image sensor. [Description of Signs] 1 Imaging optical processing system 2 Focus control system 3 Aperture control system 4 Aperture 5 2D color imaging device 6 Drive circuit 7 Correlated double sampling circuit (CDS) 8 Variable gain amplifier (AGC) 9 Clamp circuit 10 Signal Processing circuit 11 Color difference matrix processing unit 12 Micro control unit (MCU) 13 Carrier color signal modulation circuit 14 D / A converter 15 Carrier color signal output terminal 16 Luminance signal output terminal 17 Adder 18 Oscillator 20 A / D converter 30 Motion vector detector 100-1 Timing controller 101 Fixed-length (one horizontal pixel) digital data delay circuit 102 Fixed-length (one horizontal line-two pixels) digital data delay circuit 103 Circuit for calculating the averaging of input digital data 103- d Low-pass filter 104 Switching device 105 Gamma knee processing unit 10 6 Multiplier 107 Adder 108 Smoothing circuit 109 Subtractor 122 Data processing unit

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H04N ^9/ 04-9/11

Claims (1)

(57) [Claim 1] A color filter that transmits different color lights is sequentially arranged on a photosensitive element in a solid-state imaging device, and a subject image is formed on the solid-state imaging device, In a solid-state imaging device that performs an arithmetic process of multiplying output signals corresponding to different color lights of an imaging element by a predetermined arithmetic coefficient to obtain a primary color component, the arithmetic coefficient can be externally controlled, and a motion vector of a subject to be imaged is detected. The solid-state imaging device further includes an operation coefficient control unit that controls an operation coefficient according to the motion vector detection output.