JPS63182986A - Color separator for image pickup device - Google Patents

Abstract

PURPOSE:To prevent a pseudo color signal from occurring in a part where there is no vertical correlation by detecting the difference between the brightness signal low band component of neighboring three lines, comparing the result with a threshold, sorting the types of vertical correlations, selecting the most suitable arithmetic system depending on the result of the sorting, and thus processing the color sorting. CONSTITUTION:The titled separator is constituted of the following: a LPF 10 to take out an illuminance signal from a picked-up image signal coming through a color filter 9 in the arrangement shown in the figure, a memory to store the information of consecutive three lines from the output of the LPF, difference detection circuits 12, 13, 30 that receive the content of the memory 11 and detect the difference between two lines, comparators 14, 15, 31 to compare these detection results with thresholds supplied to respective terminals 16, 17, 32, a demodulation circuit 34 to take out color component from a picked-up image signal, a memory 18 to stored information of consecutive five lines, arithmetic circuits 20-23, 33 that sort input images in accordance with the comparison results of the comparators 14, 15, 31 depending on the types of the vertical correlations and execute color separation processings by operating depending on the results of the sortings, selection circuits 19, 24 that select the inputs or the outputs of the arithmetic circuits depending of the results of said comparisons, and an output terminal 7 for a blue component or a red component. As stated in the above, the color separation processing is executed separating in a case the vertical correlation exists in an input image and four cases there is no vertical correlation.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a color separation device in an imaging device.

[Conventional technology]

Figure 6 shows, for example, the Journal of the Television Society Vol. 1.37°N.
This is an example of the configuration of a conventional color separation device shown in pages 834-835 of J. O., 10 (1983).
is an image sensor, 2 is an amplifier that amplifies the output of the image sensor 1,
3 is a bandpass filter for extracting color signals, 4 is an IH delay, 5 is an adder, 6 is a subtracter, 7 is a blue component signal output terminal, 8 is a red component signal output terminal, 9 is a color filter,
10 is a low-pass filter for extracting the luminance signal.

Next, a conventional color separation method will be explained. In the image sensor 1, the image passing through the color filter 9 is converted into a bridge image. This color filter array is as shown in FIG. 2, and by passing through this color filter 9, the 0th row and the
j is an integer), the luminance signal and the first color component C are obtained by multiplexing, and the r1+1th row and the n+1+2j
In the row, the luminance signal and the second color component C2 are multiplexed and obtained. For example, taking the nth row as an example, since two lines are added to form one scanning line signal, the output signal Sn of the nth row image sensor 1 is the horizontal repetition frequency ω of the color filter array, / It is spatially modulated by 2π and becomes as follows.

Sn = JWn + Wn + Ye, + Cyn)
ko+ , (J +W, -Ye++ -Cyn) sin
ω, t...(30) (ω = 2π/d, d: repetition period of color filter)
However, as the modulation component, only the fundamental wave component of C3 is shown. Further, the signal amount generated in the nth row by passing through each color filter is expressed by symbols Wyl and Yen l CYn. Next, the output signal of the n+1st row is SR+1 in the same way.
1 = 2 (Wn+++Cyn-++Yen+++ W
n++)+ (W11*++Cy+s++-Yen+
+-Wn++)sir+ω,t・”(31) Here, W=G+R+B...(32)
Ye = G + R... (33) Cy = G + B...
(34), so the equations (30) and (31) become as shown below.

S+t =-(4Gll+ 3R,l+ 3B,%)+
F (Bn +R, 1) sin ω, t... (35) Sn + I'' (4Gre◆t+3Rn, ++3B river) +4
(Bn-+-Rn++) sin t・”(36)
The DC component of these output signals is filtered by low-pass filter 1.
0, and becomes the luminance signal Y.

Y・-(4G +3R+3B)...
(37) Also, if only the modulation components with center frequencies of ω and /2π are extracted by the bandpass filter 3, (35)
The first color component C corresponds to Equation (36) and Equation (36), respectively.
I.

As the second color component Ct, 5lIC=Y(Bll +R11)SinC3t
...(38)S(n+l1c=-i(Bn++-
Rh++ )S1nQ)1 t''・(39) is obtained. Therefore, if there is a correlation between the n-th row and the n+1-th row,
Since it can be regarded as B,,・B,;',,, R,・RI'l+1, the modulation color component snc of the nth row is set to IH delay 4.
and the adder 5 adds the modulated color component S of the n+1st row.
stomach. When added to Inc, a modulated signal B consisting of only blue components is obtained.
When c is subtracted by the subtracter 6, modulation signals Rc consisting only of red components are obtained from terminals 7 and 8, which means that the signals are color-separated. That is, 9.

Bc =Sne+S<r++++c =TBstnωs
t...(40) given ・Rc = Snc-8(11+-+) c = Bst
nω, t...(41) [Problem to be solved by the invention] The conventional color separation method is performed as described above, but n
If there is no correlation between the row and the n+1 row, B, ≠
B I'1011 Rn ≠R79, so R.

When calculating Rc = 4((Bn-Bn++) +(R++ +R+
+, +)) sin (d, t...(42) Be = (((Bn +B1101) +(Rn -R
am)) stnω3t (43) The blue component is included in the red component signal, and the red component is included in the blue component signal, resulting in a false color signal. As described above, the conventional device has problems such as false color signals occurring in areas where there is no vertical correlation.

The present invention has been made to solve the above-mentioned problems, and it is an object of the present invention to provide a color separation device for an imaging device that can prevent the generation of false color signals in areas where there is no vertical correlation.

[Means for solving problems]

A color separation device for an imaging device according to the present invention includes: a difference detection means for detecting a mutual difference between low-frequency components of luminance signals in three adjacent rows; and a comparison means for comparing the detection result with a predetermined threshold value. , a calculation means for classifying the input image according to the type of vertical correlation based on the comparison result and performing different calculations depending on the classification result to perform color separation processing.

[Effect]

In this invention, the difference between the low frequency components of the luminance signals of three adjacent rows is detected, this is compared with a predetermined threshold value, the type of vertical correlation is classified, and the optimal calculation method is selected based on the classification result. and performs color classification processing.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings. 1st
In the figure, ■ is an image sensor, 9 is a color filter, 2 is an amplifier that amplifies the output of the image sensor 1, 10 is a low-pass filter that extracts a luminance signal, 11 is a memory that stores three consecutive lines of information, 12.13 .30 is a difference detection circuit that detects the difference between two lines, 14, 15.31 is a comparator, 16
, R° 32 is a terminal that provides a threshold value to the comparators 14, 15.31, 34 is a demodulation circuit that extracts color components, 18 is a memory that stores information on five consecutive lines, 20 to 23.3
3 is an arithmetic circuit, 1°9, 24 is a comparator 14.15.31
7 is a signal output terminal for the blue component, and 8 is a signal output terminal for the red component.

Next, the operation will be explained. The outputs of the image sensor 1 when capturing an image using the conventional color filter shown in FIG.
When the output is passed through the low-pass filter 10, the luminance signal shown in equation (37) is obtained, and this luminance signal is input to the memory 11. On the other hand, when the output of the image sensor 1 is passed through the demodulation circuit 34, signals obtained by demodulating the color signals shown in equations (38) and (39) (corresponding to the first color component CIl and the second color signal component C2) are obtained. and input into the memory 18. Memory 11. The memory 18 has data in FIGS. 3(a) and (b).
) are stored in each row. However, Y(
t, j).

C3 or R(i, D (first color component C+) + Cl-
Tl (IIJ) (second color component C,) is Y(i,j) =■(4G(i,j) + 3R(i+
j) + 38(i, j) )...(50) CB+lI (fj) = child(B(i, j) +R(i
, j) )...(51) CIl-R(i+j)=child(B(i,j)-R(i,j))...(52) is shown. Hereinafter, to simplify the explanation, the m column will be explained. The memory 11 outputs luminance signals for a total of three lines: the output line and one line above and below it.

Now, if we consider the case where n rows are output as an example, signals of three rows, n-1 row, n row, and n+1 row, will be output, and therefore the output from memory 11 will be Y(m, n- 1),
Y(m,n), Y(m,'n+1).

Then, the difference detection circuit 12 calculates l Y (m, n-1) - Y (m, n) l =
DI = (53) IY (m, n) -Y (m, n+1) l= Dz by the difference detection circuit 13
...(54) The difference detection circuit 30 calculates l Y (n+, n-1) -Y(n+, n+1)
l=DI.=(55) is determined and input to comparators 14, 15, and 31, respectively, which are provided with appropriate threshold values. Terminal 16, R. 32 respectively a, d2 . d, if there is vertical correlation in the input image, D, = 0.01
=0.

0, #O, the outputs of comparators 14, 15, and 31 are low (hereinafter referred to as rLJ) and L, respectively.

It becomes L.

On the other hand, when there is no vertical correlation, it can be roughly divided into four patterns shown in FIG. That is, (11 has a boundary between rows n-1 and n), (b) has a boundary between rows n, and (C) has a boundary between rows n and n+1. In the case, (dl is the case where there is a linear pattern in the nth row. In the case of (a) to (d), the comparator 14.15.3
When examining the output of 1, in case (a), it is the output of the difference detection circuit 12.

is DI>0. The output D2 of the difference detection circuit 13 is Dt#0, the output of the difference detection circuit 30 becomes OS>>O, and the comparator 1
4.15.31 output is high (hereinafter referred to as rHJ), L
, H. Similarly, the outputs of comparators 14, 15.31 are:
When ('b) is H, H, Hl (C) is H, H,
(At the time of (1), it becomes H, H, L.

The memory 18 outputs color signals for a total of five lines, including the line to be output and two lines above and below it. Now, since we are outputting n lines, cl+l (m) is output from the memory 18.

n-2) + cl-1 (lltn-1) + Cl41
(m+n)+Cm-*(m,n+1).

Cm** (m, n+2) is output. Then, based on the output of the comparator 14°15.31, the arithmetic circuits 20 to 23.3
3 is selected by the selection circuit 19, and its output is input to the arithmetic circuit.

First, the outputs of the comparators 14, 15, and 31 are L.

In the case of L, that is, when there is vertical correlation, the arithmetic circuit 20
connected to. The calculation circuit 20 performs the following calculations.

Cal (m+n)=-CIOR(m,n)+Sat Ci-
* (m+n-1)+-! -C+-R(m, n+1)
Noon Noon... (56) C4 (m, n) = -C1, R (m, n) Ni-CI-N (m+n-IL-cs-II (m, n
+1) Da giving...(57) (51), (52) gives (56), (57
) rewriting the equation, Cm (m+n)□UpperX”(B(
m, n) + R (m, n)) 2 previous + work x-L (B (m, nn-1) - R (, n-1) 4,
t +B(m,n+1)-R(m,n+1))C4(m+n
) = Kux-'(B(m,n)+R(m,n))2 Da-kuxA(B(m,n-1)-R(n++n-1)4
and +B(m, nn-1)-R(, n+1)) Furthermore, n-
Since there is a correlation between row 1, row n, and row n+1, B(m, n) = B(m, n-1) #
B(m, n+1) R(m, n) # B(m, n-1
) ＃B(m,n+1), and Sc3(m,n) =-3i-B(m,n)
-... (5s) YuCa (m, n) ” TR (m+n)
”・(59) is obtained, and blue component and red component signals are output from terminals 7 and 8, respectively.

When the outputs of the comparators 14 and 15 are H and L, that is, when there is no vertical correlation as shown in FIG. 4(a), they are connected to the arithmetic circuit 21. The calculation circuit 21 performs the following calculations.

C1 axis, n) = Cl4R(m, n)+-Cm-* (
m, n+1) - (60)
Yu C4 (m, n) = Sat Cm, * (m, n)
-LCm-m (m, n+1) -(61)λ
2 (51), (52), (60), (
61) Rewriting the equation, C3(1,n)-TXT(B(n+,n)+R(m,n
)) + knee x-L(B(ai, n+1)-R(m, n+
1)) t C・(...)′dx * (B (″・n)+R(・
・・)) - 工
n+1)) U t Furthermore, as can be seen from Figure 4 (al), there is a correlation between row n and row n+1, so B (m+ n) ” B (m, n+1) R (m, n
) s R(m, n+1) and Cs (m, n) = “B(m, n)
−(62)C4 (m, n) =TR(m, n) −
(63) is obtained, and blue component and red component signals are output from terminals 7 and 8, respectively.

When the outputs of the comparators 14 and 15 are H, that is, when there is no vertical correlation as shown in FIG. 4(C), they are connected to the arithmetic circuit 22. The calculation circuit 22 performs the following calculations.

C3(m+n)=LCs,+t(m,n)+1C1
-R(m,n-1) ・=(64)λyuCs (n++n)='Cm+* (+w,n)-Ec
1-R(m, n-1) -(65)(51) , (
Rewriting equations (64) and (65) using equation 52), we get C3(m+n)□' xyu(B(m,n)+R(
m, n) ) t + k x-1 (B(m, nn-1) - R(, n-1) )
2, t C4<ylIIn”)=÷X p, (B (m, n
) +R(m+n) )−-yxy(B(m,n
n-1)-R(,n-1)) Furthermore, as shown in Figure 4(C), there is a correlation between row n-1 and row n, so B(lll, n-1) 'q B(m ,n)R(m,n
-1) '; R(m, n) becomes Ca(m, n) = -EB(m, n)
・”(66)Ca (m, n) −4R(m+ n
) ・” (67) is obtained, and terminal 7
．． 8 outputs the blue component and red component, respectively.

When the outputs of the comparators 14 and 15 are H, H, and when the output of the comparator 31 is H, that is, there is no vertical correlation in FIG.
), it is connected to the arithmetic circuit 23. Arithmetic circuit 2
In step 3, the following calculations are performed.

C3(m, n) = a-person(C1, R(m, n-2)
+C1,R(m,n+2))French+TX(Cl1-R(m,n-1)+CII-II(
m, n+1) )...(68) Ct(m+n) = J CB+, I(m+n-2)
+CBAI (m, n+2) ) Harvest in the afternoon Cl1-R (+n, n-1) + CI-R (m,
n+1) 1...(69) (51) , (52) gives (68) , (69
) Rewriting the equation, C5(nl, n) =, X, (
B axis, n-2) + R (m, n-2) + B (m, n+
2) +R(m+n+2) )+-! -X'(B(m
, nn-1) - R (, n-1) tF likelihood + B (m, n + 1) + R (m, n + 1) ) Ca (
m + n ) = 4χ2 (B (m, n-2) + R (m
, n-2)+B(m, n+2)+R(m, n+2))-
Sat x-2 (B (m, nn-1) - R (, n-1) 4 + B (m, n + 1) + R (m, n + 1))"Furthermore, the fourth
As can be seen from Figure (b), there is a correlation between rows n-1 and rows n-1, and rows n+1 and n+2 also have a correlation, so B(m, n-1) #8(m, n-2 ), R(m
, n-1) #R(m, nn-2) B(, n+1)
”+B axis, n+2), R(m, n+1) = R(m
, n+2) becomes Ci(m, n)・,2χy(B(nl, n-1) +B
(m,n+1) ) ”(70)Ca(m,n)−
yXy(R(m+n-1) +R(m,n+1))
.=(71) is obtained, and the blue component and red component are output from the terminal 7.8.

When the outputs of the comparators 14 and 15 are H, H and the output of the comparator 31 is L, that is, in the case fdl in FIG. 4, the arithmetic circuit 33 is selected. The calculation circuit 33 performs the following calculations. Y when imaging a uniform achromatic color with no pattern
, Ct, C2 ratio as Y:Ct:Cz = Kl:
Kt: Kl.

C3(m, n) = 工(-'Y(m, n)+-'Y(
m, n) ) ” (72) 2 to r
Kl C4, (m+n)-earth ("Y(m+n)-"Y(m+n
) ) ・”(73): 1 for LK+ cs(m+n) + Ca(m
, n) correspond to an achromatic color, but even if a thin pattern as shown in FIG. 4(d) becomes an achromatic color, it is better than a false color.

As described above, according to this embodiment, since the input image is divided into various patterns depending on the presence or absence of vertical correlation and color separation processing is performed, a high-quality color image with few false color signals can be obtained. Note that color separation processing is also performed on the (n+1)th line in the same manner as in the embodiment for the nth line.

In the above embodiment, the memory 11 and the memory 18 are used to simultaneously output adjacent rows, but the IH delay 40 to 46 is used for the broken line portion in FIG.
It may also be configured using

Further, in the above embodiment, the threshold values in each comparator were set to the same d+, dt, and dz under each condition, but
This may be done by changing the respective threshold values for each condition. According to this, even more accurate color separation processing (pattern determination when there is no vertical correlation) can be performed.

Further, in the above embodiment, for example, DI > d, + o!
〉dz, DI >d:+condition (Fig. 4 (bl)
In, Cm+, L(lTl+n)+ CI-R(
We did not use the data of man) at all, that is, we set the contribution ratio of this data to O (see equation (68)), but this is CI*R(llll+n)+C++-R(m,n)
The contribution ratio of the data may be lowered compared to the case of other conditions. This means that other conditions (Fig. 4 (al)
The same applies to the case of , (C1).

By the way, if the arithmetic expression of each arithmetic circuit in the above embodiment is generalized, it can be expressed as follows, and in the above embodiment, CI is B+R, C is B-
Let R, C, be B, C4 be R, and each coefficient be p=
% + q = 'A + r = 'A + s =
% (!: Lttaka, this changes the color filter arrangement and changes its own+Cff1 to that L R+B, R-B
tL, p ='/z + q -% + r-'A
+ Assuming S='A, CI and C4 are B and R, respectively.
You can also use it as Similarly, CI +C1 are each B
, R and pl+! L q-(L r-0+ s-1
, C3+C4 may be set as B and R, respectively.
Also, let C0°C2 be B-Y and R-Y, respectively, and p-1,
As q-0°r-0, s-1, C1 and C4 are B-Y, respectively.

As R-Y, C1 and C are each B-'A.
G, R-! /SG (G: green component of color signal),
As p=L q-0+ rswQ, s=1, C3,
C4 may be B-WG and R-'/iG, respectively.

〔Effect of the invention〕

As described above, according to the present invention, the difference between the luminance signals of three adjacent rows is detected, and this is compared with a predetermined threshold.
Cases with vertical correlation of input images and cases without vertical correlation 4
Since color separation processing is performed on a total of five patterns, a high-quality color image with less false color signals can be obtained.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing a color separation device according to an embodiment of the invention, FIG. 2 is a diagram showing a color filter used in an embodiment of the invention and a conventional example, and FIG. 3 is an embodiment of the invention. In the example, a diagram showing how each row of the color signal and luminance signal is stored in the memory, FIG. 4 is a diagram showing four patterns in the case of no vertical correlation classified according to the present invention, and FIG. FIG. 6, which is a diagram showing another embodiment, is a diagram showing an example of a conventional color separation device. 1...Image sensor, 7, 8...Signal output terminal, 9...
・Color filter, 10...Low pass filter, 11.1
8...Memory, 12,13.30...Difference detection circuit,
14°15.31... Comparator, 16, R. 32...
Threshold input terminal, 19.24...Selection circuit, 20~
23° 33... Arithmetic circuit, 34... Demodulation circuit. Note that the same reference numerals in the figures indicate the same or equivalent parts.

Claims (8)

[Claims] (1) In the n-th row and the n+2j-th row (n, j are integers), the luminance signal and the first color signal component C_1 (C_i also represents the amount of the color signal component. i; 1 to 4) are multiplexed. In the color separation device of the imaging device, in which the luminance signal and the second color signal component C_2 are obtained by multiplexing the luminance signal and the second color signal component C_2 in the n+1th row and the n+1+2jth row, A low-pass filter that extracts the component Y (Y also represents the signal amount of the low-frequency component of the luminance signal), and a low-pass filter that receives the output of the low-pass filter to extract the low-frequency component Y (m, n-1) of the luminance signal of three adjacent rows. , Y(m, n), Y(m
, n+1) (m: integer, Y (m, k) also represents the signal amount of the low-frequency component of the luminance signal at the k-th row and m-th column; k: integer); a demodulation circuit that extracts the first and second color signal components C_1 and C_2 from the output signal of the image sensor; and a demodulation circuit that receives the output of the demodulation circuit and extracts the first and second color signal components of adjacent rows. Receiving the signals from the color signal output means for simultaneous output and the luminance signal output means, the following calculation formula |Y(m, n-1)-Y(m, n)|=D_1...(
1) |Y(m,n)-Y(m,n+1)|=D_2...
・(2) |Y(m,n-1)-Y(m,n+1)|=D
_3...Difference detection means for detecting the mutual differences D_1, D_2, and D_3 between the luminance signal low-frequency components of the three adjacent rows based on (3), and the difference detection result and a predetermined threshold value d_1.
- d_1_0, and receiving a signal from the color signal output means, under the first condition that the comparison results are D_1>d_1, D_2>d_2, D_3>d_3, C_1(m, k) (However, k=
n, C_i (m, k) is the amount of the i-th color signal component in the k-th row and m column) by lowering the contribution ratio of the data compared to the case other than the first condition and calculating the following calculation formula ▲ Formula, There are chemical formulas, tables, etc.▼...(4) However, p, q, r, s; Output color signal components C_3 (m, n), C_4 (m,
Under the second condition of D_1>d_4, D_2>d_3, and D_3<d_6, Y:
As C_1:C_2=K_1:K_2:K_3 ▲There are mathematical formulas, chemical formulas, tables, etc.▼...(5) As a result, the output color signal components C_3(m, n), C_4(m,
n), and under the third condition of D_1>d_7, D_2≦d_3, C_1(m, k), C_2(m, k) (however, k
≦n-1) is compared to cases other than the third condition, and the output color signal components C_3(m, n) and C_4(m, n) are calculated using the above equation (4). Under the fourth condition of D_1≦d_9, D_2>d_1_0, the contribution ratio of the data of C_1(m, k) and C_2(m, k) (however, k≧n+1) is calculated based on the fourth condition. a calculation means for calculating output color signal components C_3(m, n), C_4(m, n) by the above-mentioned calculation formula (4) in comparison with other cases. Separation device. (2) When the above comparison result does not meet any of the conditions 1 to 4 above, the above calculation means calculates the output color signal component C_3( m, n), C_4(m,
n), and when the above first condition is met, ▲There are mathematical formulas, chemical formulas, tables, etc.▼ ...(7) The output color signal components C_3(m, n), C_4(m,
n), and when the second condition above is met, ▲There are mathematical formulas, chemical formulas, tables, etc.▼...(8) The output color signal components C_3(m, n), C_4(m,
n), and when the third condition above is satisfied, ▲There are mathematical formulas, chemical formulas, tables, etc.▼...(9) From this, the output color signal components C_3(m, n), C_4(m,
n), and when the above fourth condition is met, ▲There are mathematical formulas, chemical formulas, tables, etc.▼... (10) The output color signal components C_3(m, n), C_4(m,
2. The color separation device for an imaging device according to claim 1, wherein the color separation device calculates the color separation value n). (3) Set the threshold value of the above comparison means to d_1=d
_4=d_7=d_9...(11) d_2=d_5=
d_8=d_1_0...(12) d_3=d_6...
・(13) Claim 1 or 2 characterized in that
A color separation device for an imaging device according to section 1. (4) The first and second color signal components C_1 and C_2 are B+R (B: blue component of the color signal, R: red component of the color signal) and B-R, respectively, and the output color signal component C_3, C_4 are B and R, respectively, and the above coefficients are p = 1/2, q = 1/2, r = 1, respectively.
3. A color separation device for an imaging device according to any one of claims 1 to 3, characterized in that s=-1/2 and s=-1/2. (5) The first and second color signal components C_1 and C_2 are R+B and R-B, respectively, the output color signal components C_3 and C_4 are B and R, respectively, and each coefficient is p=1. /2, q=-1/2, r=
1/2, and s=1/2, the color separation device for an imaging device according to any one of claims 1 to 3. (6) The first and second color signal components C_1 and C_2 are B and R, respectively, the output signal components C_3 and C_4 are B and R, respectively, and the coefficients are p=1 and q=, respectively. 0, r=0, s=1
A color separation device for an imaging device according to any one of claims 1 to 3, characterized in that: (7) The first and second color signal components C_1 and C_2 are BY and RY, respectively, and the output color signal components C_3 and C_4 are BY and RY, respectively.
RY, and the above coefficients are p=1, q=0, r=0, s=1, respectively.
A color separation device for an imaging device according to any one of claims 1 to 3, characterized in that: (8) The first and second color signal components C_1 and C_2 are B-1/2G and R-1/2G (G: green component of the color signal), respectively, and the output color signal components C_3 and C_4 are B-1/each
2G, R-1/2G, and the above coefficients are p=1, q=0, r=0, s=1, respectively.
A color separation device for an imaging device according to any one of claims 1 to 3, characterized in that: