JPH0120838B2 - - Google Patents

Description

[Detailed Description of the Invention] The present invention outputs color-coded point-sequential imaging signals by irradiating imaging light onto an imaging surface formed by an imaging element such as a CCD (Charge Coupled Device) through a color filter. In particular, the present invention relates to a color imaging device that uses an optical filter to suppress aliasing components to chroma signals generated by the pixel structure of an imaging element to improve the image quality of captured images. .

Generally, in a color imaging device using a solid-state imaging device such as a CCD, in order to sample and extract color signals, color signals for each primary color (red, green, and blue) are placed in a stripe or mosaic pattern on the front of the imaging surface. A filter is provided, and by receiving imaging light at an image sensor through the color filter, a color-coded imaging output signal is obtained point-sequentially. For example, as shown in FIG. 1, there is a light receiving section 2 formed of light receiving elements 1 arranged in a mosaic pattern corresponding to each picture element for imaging;
an accumulation section 4 formed of an accumulation element 3 that accumulates each signal charge for one field for all picture elements generated according to the amount of imaging light irradiated onto each light receiving element 1; A one-chip solid-state color imaging device is constructed using a frame transfer type CCD 11 comprising a horizontal transfer register section 5 that sends out each signal accumulated in the section 4 for one horizontal line in one field. In this case, a striped color filter 6 as shown in FIG. 2 is used. That is, the color filter 6 is
A red passing filter 6 _R , a green passing filter 6 _G and a blue passing filter 6 _B are arranged in sequence in the horizontal direction as a set, and each color passing filter 6 _R ,
6 _G and 6 _B correspond to the light receiving elements 1 _a , 1 _b , and 1 _c in the vertical line of the CCD 11, respectively. Each light receiving element 1 _a , 1 b , 1 c is irradiated with imaging light through such a color filter 6 , and a signal charge e _R corresponding to the amount of light of the red component for each picture element of the subject image is applied to each light receiving element 1 a , 1 _b , 1 _{c. A} signal charge e _G corresponding to the amount of light of the green component is generated at the light receiving element 1 _b , and a signal charge e _B corresponding to the blue component is generated at the light receiving element 1 _c . Each signal charge e _R , e _G , e _B obtained in each of the light receiving elements 1 _a , 1 _b , 1 _c is vertically transferred field by field to the storage section 4 and then transferred from the storage section 4 to the horizontal transfer register section 5 . By reading out each horizontal line through the horizontal line, a point-sequential imaging output signal can be obtained. A signal processing circuit 10 having a configuration as shown in FIG. 3 is used to convert the image pickup output signal into a color video signal of a predetermined standard television system (for example, NTSC system).

That is, in the signal processing circuit 10 shown in FIG.
A signal separation circuit 12 consisting of _2R , _12G , and _12B separates the red signal _ER , green signal EG, and blue signal EB into a red signal ER, green signal _EG , and blue signal _EB through sampling processing, and converts them into three-channel simultaneous signals. Red signal E _R and green signal converted into simultaneous signals of the above three channels
E _G and blue signal E _B are supplied to voltage variable gain control type amplifiers 13 _R , 13 _G , 13 respectively so that they have the same signal level when capturing a reference white image irradiated with light of a set color temperature. _B adjusts the level and adjusts the white balance. The red signal E _R , green signal E _G and blue signal E _B whose white balance has been adjusted in this manner are transmitted to each process processing circuit 14 _R , 14 _G , 1 arranged on the subsequent stage side.
_4B performs nonlinear signal processing such as γ correction processing, the signal is converted into a luminance signal and each chromaticity signal by a matrix circuit 15, and then converted into a color image of a desired standard television system by a color encoder 16. converted into a signal.

In a color imaging device equipped with the conventional signal processing circuit 10 as described above, a signal charge corresponding to one of the three primary color signals is generated for each pixel by the color filter 6 disposed in front of the imaging surface of the CCD 11. e _R , e _G , e _B
Since _the colors are _{coded to} obtain The signal level of the luminance signal E _Y formed from the imaging output signal based on E _Y = E _R +
It becomes E _G + E _B.

Therefore, in view of the problems of the prior art as described above, the present applicant forms a color filter that is written on the front side of the imaging surface. By using a complementary color filter as at least one of the red passing filter and the blue passing filter, a color imaging device with a new configuration is provided, which improves the utilization efficiency of imaging light and ensures white balance and brightness level balance. I am proposing it first.

In other words, in the color imaging device previously proposed by the applicant, as shown in the spectral characteristics shown in FIG. 4, there is a solid line in the diagram in which the cutoff characteristics are adjusted to the long wavelength side of the green component G of the three primary color components of the imaging light. A cyan color passing filter 20 with the spectral characteristics shown in _Cy , that is, a complementary color filter for the red component R, or a yellow passing filter 20 with the spectral characteristics shown by the broken line in the figure whose cutoff characteristics are adjusted to the short wavelength side of the green component G. _In other words, a complementary color filter for blue component B is optically color coded and imaged by a color filter 21 having a configuration as shown in FIG. 5, which is used as at least one of the blue and red filters. It is configured as follows. The color filter 21 is
Frame transfer type shown in Figure 1 above
When the CCD 11 is used as an image sensor and applied to a solid-state color image sensor, it is formed into a stripe shape arranged at a repeating pitch T matching the arrangement pitch of the light receiving elements of the CCD 11 in the horizontal line direction.

For example, in the color filter 21 shown in FIG. 5, a set of all-color pass filters, that is, a white pass filter 20 _W , a yellow pass filter 20 _Ye , and a cyan pass filter 20 _Cy , each formed in a stripe shape, is used as a set. They are arranged sequentially in the horizontal direction. The above white passing filter _20W is:
It passes through the white component W in the imaging light, which includes all the color components of the three primary color components, red component R, green component G, and blue component B, expressed as W=R+G+B . . . Further, the yellow passing filter 20 _Ye passes the yellow component Ye in the imaging light expressed by the second equation: Ye=G+R.
Furthermore, the cyan color passing filter 20 _Cy passes the cyan color component Cy in the imaging light expressed by the third equation: Cy=G+B. Therefore, the color filter 21 as described above is used.
When imaging is performed by disposing the CCD 11 in front of the imaging surface, the light-receiving element _1a located at the position corresponding to the white passing filter _20W e _W = e _R + e _G according to the light amount of the white component W of the imaging light. +e _B ...The signal charge e _W expressed by the fourth formula is obtained, and in the light receiving element 1 _b located at the position corresponding to the yellow filter 20 _Cy , e _Cy = e _G + e according to the light intensity of the yellow component Ye of the imaging light. _R ...The signal charge e _Ye expressed by the fifth formula is obtained, and furthermore, at the light receiving element at the position corresponding to the cyan color passing filter 20 _Cy , e _Cy = e _G + e according to the light amount of the cyan color component Cy of the imaging light. _B ...The signal charge e _Cy given by the sixth formula is obtained.

Each signal charge e _W , e _Ye , e _Cy obtained by each light receiving element 1 _a , 1 _b , 1 _c of the CCD 11 provided with the color filter 21 is obtained as a point-sequential imaging output signal, and the next signal is obtained. By performing the processing, a luminance signal _EY and each primary color signal _ER , _EG , _EB can be obtained.

In other words, the luminance signal E _Y has the above-mentioned signal charges e _W ,
Treating the point-sequential imaging output signals from e _Ye and e _Cy as color signals for complementary colors, E _Y = E _W + E _Ye + E _Cy = (E _R + E _G + E _B ) + (E _G + E _R ) + (
E _G + E _B ) = 2・E _R +3・E _G +2・E _B ...By performing signal synthesis according to formula 7, the output is approximately 3 times lower than when color coding is performed using a primary color passing filter.
It is possible to obtain twice the signal level.

In addition, each primary color signal E _R , E _G , E _B is obtained by converting the above point sequential imaging output signal into a simultaneous signal, and then E _W −E _Cy = ( _ER + E _G + E _B ) − (E _G + E _B ) = E _R ...The red signal E _R is obtained by performing signal synthesis according to formula 8, and E _Cy +E _Ye -E _W = (E _G +E _B ) + (E _G +E _R ) - (E _R +E _G +E _B
) = E _G ... Green component E _G is obtained by performing signal synthesis according to formula 9, and furthermore, E _W −E _Cy = ( _ER + E _G + E _B ) − (E _G + E _R ) = E _B ...The blue signal E _B is obtained by performing signal synthesis according to equation 10.

Note that if the luminance signal _EY is formed from the respective primary color signals E _R , _EG , and _EB obtained by the above-described signal processing, the signal level of the luminance signal _EY cannot be increased.

Generally, a green pass filter requires a band pass spectral characteristic that allows the green component G to pass, as shown by hatching in FIG. Although it is extremely difficult to accurately set the cutoff characteristics in manufacturing, the color filter 21 having the configuration shown in FIG. 20 _Cy and yellow passing filter 20 _Ye , each filter 20
Since the blocking characteristics of _Cy , 20 _Ye are matched to the long wavelength side and short wavelength side of the green component G, the spectral characteristics can be set accurately and manufacturing is easier compared to a color filter using a conventional primary color passing filter. .

Next, FIG. 6 is a block diagram showing a specific circuit configuration of a signal processing circuit of a color imaging device constructed using the CCD 11 provided with the above-mentioned color filter 21.

In the specific example shown in FIG.
A color filter 21 for color coding as described above is disposed on the imaging surface of the CCD 11 to which imaging light for forming a subject image is irradiated via an infrared cut filter 32. A filter 100 is provided.
The CCD 11 receives the color-coded imaging light through the color filter 21 and outputs an imaging output signal, and the reference clock signal generator 33
A vertical transfer clock signal φ _V and a horizontal transfer clock signal φ _H are supplied by the first and second drive circuits 34 and 35, which operate according to the reference clock signal from the transfer clock signal φ _V and φ _H , respectively. Then, the signal charges e _W , e _Ye , and e _Cy are transferred, and point-sequential imaging output signals based on the signal charges e _W , e _Ye , and e _Cy are output.

The point-sequential imaging output signal outputted from the CCD 11 is supplied via the preamplifier 36 to a luminance level adjuster 51 provided in the luminance signal line system and a signal separation circuit 40 provided in the chrominance signal line system. . The signal separation circuit 40 is composed of three sampling and holding circuits _40W , _40Ye , and _40Cy , including a first sampling and holding circuit 40.
Sample the white signal E _W by the signal charge e _W of the white component in the point-sequential imaging output signal by _W ;
The second sampling and holding circuit 40 _Ye similarly samples the yellow signal E _Ye , and the third sampling and holding circuit 40 _Cy samples the yellow signal E _Cy , so that each color signal is
Convert E _W , E _Ye , and E _Cy into simultaneous signals. The color signals E _W , E _Ye , and E _Cy converted into simultaneous signals by the signal separation circuit 40 are supplied to first and second matrix circuits 41 and 42 .

In the first matrix circuit 41, each color signal
For E _W , E _Ye , and E _{Cy ,} perform the signal synthesis shown in Equations 8, 9, and 10 above to generate a red signal.
E _R , forming a green signal E _G and a blue signal E _B.

A red signal E _R , a green signal E _G and a blue signal E _B formed by the first matrix circuit 41
are voltage variable gain control amplifiers 43 _R , 43 _G , 43 _B that constitute the white balance adjuster 43
After the signal level is adjusted through the process circuits 44 _R , 44 _G , and 44 _B , nonlinear signal processing such as γ correction is performed on the signals, and the signals are supplied to the third matrix circuit 45 . In the third matrix circuit 45, the respective color signals E _R ′, E _G ′,
Based on E _B ′, the color difference signal E _R ′−E _G ′ and E _B ′−
A color difference signal E _G ' is formed and each color difference signal is supplied to the color encoder 60, and the gain control amplifiers 47 _R and 47 _B are connected via the switch circuit 46.
is supplied to. Here, E _R ′, E _G ′, and E _B ′ indicate each color signal subjected to gamma correction. The color encoder 60 outputs the color difference signal E from the third matrix circuit 45 based on the synchronization signal supplied from the synchronization signal generator 37 which is activated by the reference clock signal from the reference clock signal generator 33. _R ′−E _G ′, E _B ′−E _G ′, and the luminance signal supplied via the luminance signal line system described later.
An NTSC color video signal is synthesized from E _Y ′ and output.

Here, the white balance adjuster 43 is
The gain of the amplifier 43 _G for the green signal E _G is fixed, and the gains of the amplifiers 43 _R and 43 _B for the red signal E _R and blue signal E _B can be set freely, and the white balance adjustment operation is possible. At the same time, an image of a reference white color irradiated with light of a set color temperature is captured and transmitted to each gain control amplifier 47 _R , via a switch circuit 46 .
47 _B third matrix circuit 45
_The above red signal _{E R and blue} signal
The gains of the amplifiers 43 _R and 43 _B for E _B are automatically controlled, and the gains of the amplifiers 43 _R and 43 _B are set so that the signal levels of the color signals E _R , E _G , and E _B are equal.

Further, the brightness signal line system includes a brightness level adjuster 5 connected in cascade to the output side of the preamplifier 36 described above.
1, a process processing circuit 52 and an aperture correction circuit 53. The brightness level regulator 51 is composed of a one-channel voltage variable gain control amplifier 51A whose gain is sequentially controlled by a control signal from a control signal generator 50, and includes a preamplifier from the CCD 11. The signal level of the point sequential imaging output signal supplied via 36 is adjusted in the state of the point sequential signal. The control signal generator 50 uses the second matrix circuit 42 provided in the color signal line system to generate E _W −E _Ye =E _B . . . Equation 11 E _W −E _Cy = E _R . Each color signal E _B obtained by signal synthesis of Equation 12,
A timing signal indicating the timing of E _R is supplied via buffer amplifiers 48a and 48b, and a synchronization signal from the synchronization signal generator 37 is also supplied. A control signal synchronized with the timing of the signals E _R , E _G , and E _B is supplied to the voltage variable gain control amplifier 51A.

In general, in a color imaging device that optically spatially samples a subject image using a color filter and performs color coding to obtain each color signal,
Since the subcarrier is modulated by the frequency component resulting from the spatial sampling and aliasing distortion occurs, it is necessary to adjust the brightness level to eliminate the aliasing distortion in the electrical signal processing system. For example, when imaging is performed using a CCD 11 with an ideal picture element configuration in which each picture element is arranged at equal intervals as shown in FIG. 7A, through a color filter formed by a primary color passing filter, as shown in FIG. As shown in B, each color signal E _R , E _G , E _B has the same phase angle θ ₁ , θ ₂ ,
Since it appears at the sampling frequency 1/T determined by the pitch T of the color filter above with θ ₃ ,
If the signal levels l _R , l _G , l _{B of each color signal E R , E G , E G for an achromatic} image are adjusted so that l _R = l _G = l _B ...... Equation 13 is obtained. , aliasing distortion can be eliminated. Moreover, in this case, the above-mentioned luminance level balance and the above-mentioned white balance are compatible.

However, as in this specific example, when the subject image is spatially sampled and color coded through the color filter 21 formed by a complementary color filter, the white balance adjusts the level of the three primary color signals, Luminance level balance must be performed for complementary color signals, and the above-mentioned white balance and luminance level balance become incompatible. For example, considering the spectral characteristics of the color filter 21 in this specific example, the pass coefficient of the red component R in the cyan pass filter 20 _Cy is r _Cy , the pass coefficient of the green component G is g _Cy , and the pass coefficient of the red component R in the cyan pass filter 20 _W is The passage coefficient of the red component R is r _w , the passage coefficient of the green component G is g _w , and the blue component B is
The passage coefficient of the green component G in the yellow passage filter 20 _Ye is denoted by b _W , and the passage coefficient of the green component G in the yellow passage filter 20 Ye is g _Ye ,
Letting the passage coefficient of blue component B be b _Ye , the amount of light passing through each filter 20 _Cy , 20 _W , 20 _Ye is Y _Cy , Y _W ,
The spectral characteristics of the color filter 21 indicated by Y _Ye are as follows: Therefore, the red component R captured through the color filter 21 is as follows: R=-{Y _Cy・(g _W・b _Ye −b _W・g _Ye )+Y _W・(−b _W・g
_Ye )+Y _Ye・(g _Cy・b _W )}/｜r・g・b|……This becomes the 15th formula. Here, r _Cy = r _W = r, g _Cy = g _W = g _Ye =
If g, b _W = b _Ye = b, the above red component R is _as follows _: You can ask for it. Now, if the above equation 16 is guaranteed when lighting at a certain color temperature, then G _Cy (r + g) = G _W (r + g + b) = G _Ye (g + b) ...... Adjust the level so that it becomes equation 17. By doing this, you can balance the brightness level. In this way, in order to white balance each primary color signal obtained by dividing the luminance level balance, it is necessary to return the signal levels to Y _Cy , Y _W , Y _Ye , which requires complex signal processing. will have to be done.

Additionally, in general, the pixel structure of CCD11 is
As shown in Figure A, a light receiving element 1 _a for each color signal,
1 _b and 1 _c are each separated by an overflow drain OFD, so even if each light receiving element 1 _a , 1 _b , 1 _c is arranged at a predetermined pitch T, the phase angle θ of the obtained color signal is ₁₂ , θ ₂₃ , and θ ₁₃ are not equal to each other. Therefore, in the CCD 11 with such a pixel structure, the luminance level balance when performing spatial sampling of an object using a primary color filter is as follows: cosθ ₁₂ =l ₃ ² −l ₁ ² −l ₂ ² /2l ₁ l ₂ ... Equation 18 cosθ ₂₃ = l ₁ ² + l ₃ ² − _{l 2} ² /2l ₁ l ₃ ... Each signal level l ₁ , within the range where the general equation Equation 19 holds true.
It is necessary to set l ₂ and l ₃ .

However, in this specific example using the signal processing circuit shown in FIG. 6 as described above, white balance and brightness level balance can be performed independently, so that the pixel structure shown in FIG. CCD
Color filter 21 formed by the complementary color filter in step 11
Obtain a color-coded imaging output signal by
White balance and brightness level balance can be adjusted reliably.

In addition to the color filter 21 applied in the above-described specific example, the cyan color passing filter 20 may also be used.
Any color filter formed by using _Cy or yellow passing filter 20 _Ye for red or blue color coding may be used.

For example, as shown in FIG. 9A, a color filter 21A is formed of a cyan passing filter 20 _Cy , a yellow passing filter 20 _Ye , and a complementary color filter for green, that is, a magenta passing filter 20M.
When using E _Y = E _M + E _Ye + E _Cy = (E _R + E _B ) + (E _G + E _R ) + (E _G +
E _B )=2・E _R +2・E _G +2・E _B ...The luminance signal E _Y can be formed by signal synthesis according to equation 20, and each primary color signal E _R , E _G , E _B is It can be formed by the following signal synthesis.

E _Ye +E _M −E _Cy =2・E _R E _Ye +E _Cy −E _M =2・E _G E _Cy +E _M −E _Ye =2・E _B ...Formula 21 Also, as shown in Figure 9B , a color filter 21 formed from a cyan passing filter 20 _Cy , a yellow passing filter 20 _Ye , and a green passing filter 20 _G.
By using B, each signal E _Y , E _R , E _G , _B can be formed by the following signal synthesis.

E _Y =E _Cy +E _Ye +E _G =E _R +3・E _G +E _B
...Formula 22 E _Ye −E _G = E _R E _G = E _G E _Cy −E _G = E _B ...Formula 23 In addition, as shown in Figure 9C, cyan color passing filter 20 _Cy and red light passing Color filter 21C formed by filter _20R and green passing filter _20G
By using , each signal E _Y , E _R , E _G , and E _B can be formed by the following signal processing.

E _Y = E _R + E _G + E _Cy = E _R +2・E _G + E _B ... 24th formula E _R = E _R E _G = E _G E _Cy −E _G = E _B ... 25th equation and 9th As shown in Figure D, if a color filter 21D formed by a yellow passing filter 20 _Ye , a green passing filter 20G, and a blue passing filter _20B is used, each signal E _Y , E _R , E _G , E _B can be changed as follows. It can be formed by signal synthesis.

E _Y =E _G +E _B +E _Ye =E _R +2E _G +E _B ...26th formula E _Cy -E _G =E _R E _G =E _G E _B =E _B ...27th formula Explanation of the above specific example As is clear from the above, at least one color of the point-sequential imaging color signal output obtained by imaging through a color filter for color coding of the imaging light is a color signal of a color component other than the three primary colors. In a color imaging device in which the spectral characteristics of the filter are set, a brightness level adjuster is disposed in a brightness signal line that forms and outputs a brightness signal from the point-sequential imaging color signal output, and the brightness level is adjusted. The feature is that a white balance adjuster is installed in the color signal line that separates the point-sequential imaged color signal on the front stage of the device, converts it into three primary color signals, and outputs it, making it possible to perform color coding using a primary color filter. It is possible to improve the utilization efficiency of imaging light and to perform highly sensitive color imaging compared to the case where the imaging light is used as an image, and moreover, it is possible to reliably adjust the white balance and the brightness level balance.

By the way, as mentioned above, in a color imaging device using a solid-state imaging device such as a CCD, the subcarrier is modulated by the sideband component generated in the electrical signal processing system, that is, the frequency component of spatial sampling due to the pixel structure of the imaging device. The aliasing distortion that occurs in the optical system for forming a subject image on the imaging surface of the image sensor can be eliminated by adjusting the brightness level in the signal processing system. It is necessary to remove it optically.

For example _, in the case of a CCD 11 with an equally spaced pixel structure as shown in FIG. +l _G ′〓δ(t
-n・T−T/3) +l _B ′〓δ(t−n・T−2/3・T)}〓re
ct(t/T/3)... can be approximately expressed by the 28th equation. However, in Equation 28, l _R ′=l _R (t)=f(t),
l _B ′=l _B (t)〓f(t), l _G ′=l _G (t)〓f(t)
, and l(t) represents the intensity distribution of the object projected image,
f(t) indicates the spread function of the LPF effect due to the MTF of the imaging lens and the CCD aperture, and rect( ) is a function indicating the sample-hold effect. Then, when the above formula 28 is Fourier transformed, O(f)=[L _R ′(f)==δ(f-1/T・n)+
L _G ′(f)〓〓〓δ(f-1/T・n)e ^-j2 〓 ^n/3 +L _B ′(f)〓〓〓δ(f-1/Tn)e ^-j2 〓 ^2
n/3 ]×sinc(ππ/3f)...The 29th equation becomes the 29th equation. Therefore, when imaging an achromatic object with a single repetition frequency of P, L _R ′(f)=L _G ′(f)=L _B ′(f)=
δ(f-P)+δ(f+P)/2, so O(f)=sinc(πT/3P)・[=δ(f+P-1
/Tn)+〓δ(f±P-1/Tn)・e ^-j2 〓 ^n/3 +〓δ(f±P-1/Tn)・e ^-j2 〓 ^2n/3 〕...
...An imaging output signal expressed by the 30th equation is obtained. The frequency components of the imaging output signal shown in Equation 30 above are distributed as shown in FIG. 10, and in FIG. There is. In addition,
In the NTSC signal, the modulation axis is set as shown in FIG.

Therefore, when the color filter for color coding is formed by the three primary color filters 6 of the red passing filter 6 _R , the green passing filter 6 _G , and the blue passing filter 6 _B , the aliasing distortion component in the optical system is 120°. Since _{there is} a _phase difference _of
°/360) +0.45・cos(2πf _f t−2πn/3) _B ×cos(2πf _c t−2π103°/360)+0.59・cos(2
πf _f t−4πn/3) _G ×cos(2πf _c t−2π209°/360)
...It can be shown in Equation 31, which is Equation 31. Here, the 31st
In the equation, f _f =n/T-P, f _c is the frequency of the subcarrier of the NTSC signal. The signal component shown in Equation 31 above has both side band components of f _c ±f _f , but it has a lower side band component close to the color carrier band side, that is, f _c −f _f =P−(f _c -n/T) components, they are as shown in FIGS. 12A and 12B. Here, FIG. 12A shows a signal component with a frequency of n=1, that is, 1/T, and FIG. 12B shows a signal component with a frequency of n=2, that is, 2/T. The signal components with a frequency of 1/T shown in FIG. It falls and becomes an interference wave to the captured image. Also, the above 12th
The signal component of the frequency 2/T shown in FIG. B does not become an interference wave because its combined vector is approximately zero. In order to reduce the signal level of the signal component of the frequency 1/T mentioned above, it is necessary to suppress the P component that falls within the color signal band when n = 1 before sampling it with the CCD 11, and to An optical low-pass filter may be used. An optical low-pass filter formed from a single crystal plate is n=2N+
1 (however, N=1, 2...) is known to have a filter characteristic as shown in FIG. 13, in which the trap point is N=1, 2, . That is,
In a birefringent material 90 such as a crystal plate, the main cross section defined by the wavefront normal of light and the optical axis y of the crystal is shown in FIG. 14, and the crystal is divided into normal light components and extraordinary light components. Since the light passes through the crystal, the distance between the paths of the normal light component and the abnormal light component that passed through the crystal is 2d.
Then, it is possible to provide a low-pass filter characteristic H _L (f) as shown in Equation 32, which is equivalent to an electrical transversal filter.

H _L (f)=1/2e ^+j2 〓 ^fd +1/2e ^-j2 〓 ^fd = cos (2πfd) ...Equation 32 As is clear from Equation 32, by setting 2d=T/2, the above A low-pass filter characteristic as shown in FIG. 13 can be provided.
If an optical low-pass filter with filter characteristics as shown in FIG. 13 above is installed in the optical system of a color image pickup device, the signal component with a frequency of 1/T will be suppressed in the optical system, so the signal component of the three primary color filters will be used. This is effective in removing aliasing distortion caused by spatial sampling when two-color coding is applied to a color filter.

In addition, color filters _21C as _shown in FIG _.
When color coding is performed using the color filter 21C formed at 0 _G , an aliasing distortion component as shown in equation 30 above occurs, but each color signal E _R ,
E _Cy and E _G are each primary color signal in the matrix circuit 41.
The blue signal component E B is decomposed into E _R , E _G , and E _B , and the blue signal component E _B is formed by signal synthesis according to equation 25 above. In this case, if the signal levels of each color signal E _R , E _Cy , E _G are l _R , l _Cy , l _G , then l _B = l _Cy − l _G , and assuming the state that l _R = l _G = l _Cy . , l _B =l _G /2, so it is necessary to amplify approximately twice (0.59/0.45)×2. The aliasing distortion component in the signal after the above matrix circuit is also amplified approximately twice, so C _f = 0.63・cos (2πf _f t) cos (2πf _f t + 2π13.5°/360) + 0.45 × 0.5゜/0.45×2・cos {2πf _f t+(−2πn/3
−4πn/3+n)/2} cos(2πf _c t−2π103°/360)+0.59・cos(2
πf _f t−4πn/3) cos(2πf _c t−2π209°/360)…
...Equation 33 The aliasing distortion component can be expressed by Equation 33.
The lower sideband component of the signal shown in equation 33 above is:
They have a phase relationship as shown in FIGS. 15A and 15B. FIG. 15A shows a signal component with a frequency of n=1, that is, 1/T, and FIG. 15B shows a signal component with a frequency of n=2, that is, 2/T. As is clear from FIGS. 15A and 15B above, in this case, the aliasing distortion cannot be removed unless both the high-order components of n=1 and n=2 are suppressed in the optical system. .

Furthermore, as the color filter 21, the fifth
When the white passing filter 21 _w , the yellow passing filter 21 _Ye , and the cyan passing filter 20 _Cy are formed as shown in the figure, the synthesis process shown in equations 8 to 10 is performed as described above. Each color signal E _R , E _G , and E _B will be formed. Assuming that the signal level l _W of the white component is l _W =3, and the signal level of each color component is l _Cy =l _Ye =2, An aliasing distortion component as shown in Equation 34 occurs. The lower sideband component of the aliasing distortion component has a phase relationship as shown in FIGS. 16A and 16B, and optically suppresses signal components at frequencies n=1 and n=2. There is a need.

Therefore, in the present invention, an optical low-pass filter 100 having filter characteristics as shown in FIG. 17 and having trap points for the frequency components n=1 and n=2 is arranged in the imaging optical system. By removing aliasing distortion caused by spatial sampling in the optical system, images are taken without interference waves. That is, in the color imaging device shown in FIG. 6 described above, by disposing the optical low-pass filter 100 in front of the imaging surface of the CCD 11, 1/1/2 of the subject image formed on the imaging surface of the CCD 11 is reduced. By optically suppressing each frequency component of T and 2/T, aliasing distortion components in the optical system are removed.

As the optical low-pass filter 100,
For example, as shown in FIG. 18, a first crystal plate 100A having a separation distance of _2d1 and each optical main cross section in a direction of 45° with respect to the optical main cross section A of the first crystal plate 100A. B, C, and each optical main cross section B, C makes an angle of 90° to each other.
A structure formed by laminating crystal plates 100B and 100C can be used. Note that the separation distances 2d ₂ and 2d 3 of the second and third crystal plates 100B and 100C are equal to each other, and d ₂ = _{d 3 =} 2.
√2d is set to ₁ . In the optical low-pass filter 100 configured as described above, light passes through a beam path as schematically shown in FIG. 19. That is, the light that is perpendicularly incident on the point (P ₀ ) on the first crystal plate 100A is composed of a normal light component that vibrates in a direction perpendicular to the optical main section A of the crystal plate 100A and the optical component. As shown in FIG. 20A, the normal light component that passes through the crystal plate 100A without being refracted passes through the point (P ₀₁ ) and is separated into the extraordinary light component that vibrates in a direction parallel to the main cross section A. Further, the extraordinary light component is refracted and incident on the second crystal plate 100B through a point (P ₀₁ '). Said second crystal plate 100
The light incident on each point (P ₀₁ , P ₀₂ ) on B consists of a normal light component vibrating in a direction perpendicular to the optical main section B of the crystal plate 100B and a normal light component parallel to the optical main section B. 20th to the extraordinary light component that vibrates in the direction of
separated as shown in Figure B. The light incident on the point (P ₀₁ ) on the second crystal plate 100B is
The normal light component passes through the point (P ₀₂ ) to the third crystal plate 1
00C, and the extraordinary light component is incident on the third crystal plate 100C through a point (P ₀₂ '). In addition, the normal light component of the light incident on the point (P ₀₁ ') on the second crystal plate 100B is incident on the third crystal plate 100C through the point (P ₁₂ ), and the abnormal light component is incident on the third crystal plate 100C through the point (P 12 ). The light component is incident on the third crystal plate 100C through a point (P ₁₂ '). here,
The second and third crystal plates 100B and 100C
has optical main cross sections B and C that form an angle of 90 degrees to each other as described above, so that the second crystal plate 10
The normal light component of 0B becomes the extraordinary light component of the third crystal plate 100C, and the second crystal plate 10
The abnormal light component of 0B becomes the normal light component of the third crystal plate 100C. Therefore, the third crystal plate 1
The normal light components of the second crystal plate 100B that are incident on each point (P ₀₂ , P ₁₂ ) on 00C are all reflected as abnormal light components of the third crystal plate 100C at each point (P ₀₃ ′, P ₁₃ '), and the third
The extraordinary light components of the second crystal plate 100B that are incident on each point (P ₀₂ ′, P ₁₂ ′) on the crystal plate 100C are all converted into normal light components of the third crystal plate 100C at each point (P 02 ′, P 12 ′). P ₂₃ , P ₃₃ ). That is, in the optical low-pass filter 100, the light incident on the point (P ₀ ) on the first crystal plate 100A is arranged in a straight line from the third crystal plate 100C as shown in FIG. 20C. The light is then separated into light components that pass through four points (P ₀₃ ′, P ₁₃ ′, P ₂₃ , P ₃₃ ) and is emitted. Note that each arrow in FIG. 20 indicates the vibration direction of each light component. Therefore, when the light incident on the point P ₀ is received by the CCD 11, O _L (f) = 1/4 (e ^+j2 〓 ^f1d1 +e ^-j2 〓 ^f1d1 ) + 1/ on the horizontal line.
4 (e ^+j2 〓 ^f23d2 +e ^-j2 〓 ^{f2 3d1} ) = 1/2cos (2πf ₁ d ₁ ) + 1/2cos (2πf ₂ 3d
₁ ) ... Imaging is performed in a state where the optical low-pass filter characteristic O _L (f) as shown in Equation 35 is given. Filter characteristics shown in equation 35 above
If O _L (f) is illustrated, it will be as shown in the above-mentioned FIG. 17.

As is clear from the above description, according to the present invention, the color filters are repeatedly arranged at a pitch of T in the horizontal direction, and at least one of the plurality of unit filters in one repetition period passes a complementary color component. In a color imaging device configured to take an image by color-coding the imaging light through a color filter and outputting a point-sequential imaging signal, at least 1/T and 2/T of frequency components are used. By disposing an optical filter with a trap point in the imaging optical system, the aliasing distortion component to the chroma signal generated due to the pixel structure of the imaging device can be optically removed. Since the interference is suppressed in the system, it is possible to capture images with good image quality without including interference waves, and the intended purpose can be fully achieved.

Note that the above-mentioned optical low-pass filter can be formed using a retardation filter or the like in addition to a crystal plate.

[Brief explanation of drawings]

FIG. 1 is a plan view schematically showing the structure of a frame transfer type CCD. Figure 2 shows the above
1 is a schematic plan view showing the configuration of a primary color filter conventionally used for color coding when a one-chip solid-state color imaging device is configured using a CCD. FIG. 3 is a block diagram showing the configuration of a signal processing circuit in a conventional solid-state color imaging device. FIG. 4 is a characteristic diagram showing the spectral characteristics of the color filter applied to the color imaging device according to the present invention. FIG. 5 is a schematic plan view showing a configuration example of a color filter applied to a color imaging device according to the present invention. Figure 6 shows the CCD shown in Figure 1 above.
FIG. 2 is a block diagram showing an example of the configuration of a signal processing circuit applied to an embodiment of the invention when a solid-state color imaging device is constructed using the following. FIGS. 7A and 7B are explanatory diagrams for explaining a brightness level balance adjustment operation when using a CCD having an ideal picture element structure. FIGS. 8A and 8B are explanatory diagrams for explaining the brightness level balance adjustment operation when using a CCD having a concretely realized pixel structure.
FIG. 9A, FIG. 9B, FIG. 9C, and FIG. 9D are schematic plan views showing configuration examples of each color filter applicable to the color imaging device according to the present invention. FIG. 10 is a frequency distribution diagram of signals in an imaging device that performs spatial sampling. FIG. 11 is an explanatory diagram showing the modulation axis of the NTSC signal. 1st
FIG. 2 is an explanatory diagram for explaining how aliasing distortion occurs when color coding is performed using three primary color filters. FIG. 13 is a characteristic diagram showing the filter characteristics of an optical low-pass filter necessary to remove the above-mentioned aliasing distortion. FIG. 14 is a schematic optical main sectional view for explaining the operation of a filter using one crystal plate. FIGS. 15 and 16 are explanatory diagrams for explaining how aliasing distortion occurs when color coding is performed using complementary color filters. FIG. 17 is a characteristic diagram showing the filter characteristics of an optical low-pass filter necessary to eliminate the above-mentioned aliasing distortion. FIG. 18 is a schematic perspective view showing a configuration example of the optical low-pass filter. FIGS. 19 and 20 are schematic diagrams for explaining the state of light passing through the optical filter. 11...CCD, 20 _Ye , 20 _Cy , 20 _W , 20
_M , 20 _G , 20 _R , 20 _B ...each color passing filter,
21, 21A, 21B, 21C, 21D... Color filter for color coding, 40... Signal separation circuit, 40 _W , 40 _Ye , 40 _Cy ... Sampling hold circuit, 41, 42, 45... Matrix circuit, 43...White balance adjuster, 43 _R ,
43 _G , 43 _B , 51A... Voltage controlled variable gain amplifier, 44 _R , 44 _G , 44 _B , 52... Process processing circuit, 46... Switch circuit, 50... Control signal generator, 51... Brightness Level balance adjuster, 60... Color encoder, 100... Optical low pass filter, 100A, 100B, 10
0C...Crystal plate.

Claims (1)

[Claims] 1 Color filters are repeatedly arranged at a pitch of T in the horizontal direction, and at least one of the plurality of unit filters in one repetition period passes a complementary color component. In a color imaging device that performs color coding to capture images and outputs point-sequential imaging signals, an optical filter that has trap points for each frequency component of at least 1/T and 2/T is used for imaging. A color imaging device characterized in that it is arranged in an optical system of.