JP3079839B2 - Solid-state imaging device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a solid-state imaging device using a solid-state imaging device, and more particularly to a solid-state imaging device in which a color separation filter is provided in an all-pixel readout imaging device to obtain a color video signal.

[0002]

2. Description of the Related Art A color camera using a solid-state image pickup device in an image pickup section includes a multi-chip system using a plurality of image pickup devices and a single-chip system using a single image pickup device. In the single-panel system, color separation filters having different spectral characteristics are provided for each pixel, and a color signal and a luminance signal are obtained by calculating pixel signals corresponding to these color separation filters. In such a single-plate system, the spatial sampling frequency of each signal is inevitably lowered, and a false signal, that is, a moire is generated in the luminance signal and the color signal due to the folding of the high-frequency component of the image.

[0003] By the way, in the conventional image pickup device, the vertical direction 2
A PD (photodiode) mix system is used in which pixels are added and read out. As an example of an imaging method using the imaging element of this method, there is a color difference sequential method. FIG. 9 shows an example of a color separation filter array used in the color difference sequential system. FIG.
0 is a sampling carrier diagram in the color separation filter array of FIG. 9, and Px and Py indicate horizontal and vertical pixel intervals, respectively. In the color separation filter array shown in FIG. 9, since color carriers exist in the horizontal direction of the spatial frequency, luminance moire occurs in the horizontal direction of the spatial frequency, which causes a reduction in resolution in the horizontal direction.

A luminance signal is obtained by adding color signals corresponding to several different color separation filters, and the addition ratio when adding the respective color signals to obtain a luminance signal is set to the reciprocal of the sensitivity ratio of each color signal. It is known that luminance moire can be minimized by doing so. However, since the sensitivity of each color signal depends on the spectral characteristics of the input light, moiré cannot be simultaneously minimized for any incident light.

[0005] A general image has a property that color correlation is strong in a local area. Utilizing this property, a method for suppressing moiré against any incident light is disclosed in Japanese Patent Application Laid-Open No. 2-166988,
Or "The Journal of the Institute of Television Engineers of Japan" (vol. 46, No. 9, pp. 115
3 to 1160 (1992)).

Hereinafter, this method will be briefly described. For example, in the color difference sequential method using the PD mix image sensor and the color filter array of FIG. 9, Mg + Ye and G + Cy or Mg + Cy and G + Ye are output in dot sequence. When there is no color change, the ratio of Mg + Ye: G + Cy: Mg + Cy: G + Ye is considered to be constant. Therefore, for example, Mg +
In the row where Ye and G + Cy are output, Mg + Ye
For the pixel position where is output, G + Cy is interpolated by the operation of (G + Cy) '= (G + Cy) L / (Mg + Ye) L * (Mg + Ye), and for the pixel position where G + Cy is output, (Mg + Ye) The calculation of '= (Mg + Ye) L / (G + Cy) L * (G + Cy) interpolates Mg + Ye. (Mg + Y
e) L and (G + Cy) L are Mg + Ye and G + C, respectively.
This is the low-frequency component (L) of y, and is obtained by subjecting the Mg + Ye and G + Cy signals to low-pass filtering. That is, the luminance moiré is suppressed by performing the above-described interpolation calculation in accordance with the ratio of the color signal in which the color moiré is suppressed by the low-pass filter.

When a color separation filter array in which color carriers are present in the horizontal direction is used as in the color separation filter array shown in FIG. 9, a low-pass filter for suppressing color moire is 1 / 4Px in the horizontal direction. , But with such a filter, the horizontal
In order to suppress color moire around Px, it is necessary to use a low-pass filter having a relatively narrow band.

[0008]

However, when the above-described interpolation calculation is performed, if there is a sharp color change point in the subject, a false signal due to an error between the band-limited color signal and the actual color. And the error at the color change point becomes larger as the band of the low-pass filter becomes narrower, which causes a problem that a false signal is generated in a wide range, and it is difficult to suppress the moiré of the luminance near 1 / 2Px in the horizontal direction. Met.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a solid-state imaging device capable of solving the above-mentioned conventional problems, and in particular, providing an output in which luminance moire in horizontal and vertical spatial frequency directions and a false signal generated by an interpolation operation are suppressed. It is in.

[0010]

In order to achieve the above-mentioned object, the present invention uses an all-pixel readout image pickup device provided with a color separation filter having a predetermined repetition arrangement of vertical M pixels × horizontal N pixels. By performing a two-dimensional operation on signals from the K pixel × horizontal L pixel area, color signals are separated,
The configuration is such that the interpolation operation of the luminance signal is performed according to the ratio of the separated color signals.

[0011]

According to the above arrangement, pixel interpolation is performed on color signals from surrounding pixel signals, and in particular, color signals with little color moiré are obtained in the horizontal and vertical directions, and interpolation calculation is performed using color signals with little color moiré. Accordingly, it is possible to obtain a high-resolution luminance signal with less luminance moire especially in the horizontal and vertical directions.

[0012]

Embodiments of the present invention will be described below with reference to the drawings.

(Embodiment 1) FIG. 1 is a block diagram showing an embodiment of a solid-state imaging device according to the present invention. In FIG. 1, reference numeral 11 denotes a color separation filter, 12 denotes an all-pixel readout imaging device, and 13 denotes a color signal. Generating means, 14 is a pixel interpolation calculating means, 15 is a luminance signal generating means, and 16 is a high-pass filter. FIG. 4 shows an example in which M = 4 and N = 2 of the color separation filter array used in the all-pixel reading solid-state imaging device. FIG. 5 is a sampling carrier diagram when the color separation filter arrangement shown in FIG. 4 is used. Where Px,
Py is a horizontal and vertical pixel interval, respectively.

First, a color signal generation method in the color signal generation means will be described. In order to perform color separation in the color separation filter array of FIG. 4, for example, the color separation filter W,
W signal, G signal, Ye signal corresponding to G, Ye, Cy,
The red signal r can be color-separated by the signal processing means for adding the Cy signals with weights of 1, -1, 1, and -1 respectively. In this case, the horizontal spatial sampling period of the red signal r is the pixel interval. Since it is doubled, color moiré occurs around a horizontal spatial frequency of 1 / 2Px and a vertical spatial frequency of 0.

In the color separation filter arrangement shown in FIG. 4, the sampling phases of the red signal r and the blue signal b are inverted every two horizontal lines, and the lines whose sampling phases are inverted are interpolated with each other. Moire in the horizontal direction can be suppressed. In the interpolation, for example, in the color separation filter arrangement shown in FIG. 4, the W, G, Ye, and Cy signals obtained from the upper half 2 × 2 pixel regions are respectively 1, and −.
The same weighting is applied to the red signal r separated by adding weights of 1, 1, and −1 and the W, G, Ye, and Cy signals obtained from the lower half of the vertical 2 pixels × horizontal 2 pixel area. This is done by taking the average of the red signals r that have been added and separated.

That is, in the color separation filter array of FIG. 4, W, G, Y obtained from a vertical 4 pixel × horizontal 2 pixel area.
The red signal r is separated by adding weights of 1, -1, 1, and -1 to the e and Cy signals, respectively. The blue signal b can also be separated by adding the signals corresponding to the color separation filters W, G, Ye, and Cy in the 4 × 2 pixel area with weights of 1, −1, −1, and 1, respectively. , Red signal r.

FIGS. 6A and 6B show W, G, and W in the above-described calculation for performing color separation of the color signals r and b, respectively.
The weights for the Ye and Cy signals are arranged in correspondence with the positions of W, G, Ye and Cy of the color separation filter in FIG. That is, the color signals r and b are weighted as shown in FIGS. 6A and 6B and added to each pixel value of 4 × 2 regions corresponding to the color separation filter array shown in FIG. Separated by the calculating means. The green signal g is obtained by adding all the pixel values of the vertical 4 × horizontal 2 area of the color separation filter array of FIG. 4 to be 4r + 8g + 4b. Can be separated by subtracting 4r and 4b from.

FIG. 6C shows the weighting for the W, G, Ye, and Cy regions in the operation for separating the g signal.
Are arranged in correspondence with the positions of W, G, Ye, and Cy of the color separation filters. That is, the g signal can be separated by weighting the W, G, Ye, and Cy signals with −1, 3, 1, and 1, respectively, and adding them. By subjecting the color signals r, b, and g obtained by the above operations to low-pass filtering in the horizontal and vertical directions, a color signal in which color moire of a high spatial frequency component is suppressed is obtained. Can be.

When the above-described color signal generation processing is performed on the color separation filter array of FIG.
Color moire occurs around the vertical spatial frequency 1 / 2Py. This color moiré has a vertical spatial frequency of 1 / 2P of incident light.
This is due to the r (red) and b (blue) components near y.
Therefore, it can be suppressed by an optical low-pass filter such as a crystal filter that suppresses the vicinity of the vertical spatial frequency 1 / 2Py. However, since the quartz low-pass filter suppresses not only the r and b components but also the g (green) component, the resolution of the luminance in the vertical direction is deteriorated. Therefore, it is desirable to selectively suppress the spatial high frequency components of r and b while leaving g.

To realize this, Japanese Patent Laid-Open Publication No.
No. 915 discloses a color-selective optical low-pass filter. A color-selective optical low-pass filter is formed by combining glass materials having different refractive indexes for each wavelength of light, that is, each color.
Different low-pass filter characteristics can be obtained for each color. This color-selective optical low-pass filter selectively suppresses the vicinity of the vertical spatial frequency 付 近 Py of r and b while leaving the vertical spatial frequency component of g, thereby suppressing the degradation of the resolution of the luminance in the vertical direction. , Can suppress color moiré.

According to the above-described method, the color signals RL (red), GL (green), and BL having little color moire in the horizontal and vertical directions are provided.
(Blue) is obtained. Using this color signal, G 'at the position of the W pixel, W' at the position of the G pixel, Cy 'at the position of the Ye pixel, and Ye' at the position of the Cy pixel are obtained by the following equation.

W '= WL / GL * G G' = GL / WL * W Ye '= YeL / CyL * Cy Cy' = CyL / YeL * Ye where WL = RL + GL + BL YeL = RL + GL CyL = GL + BL In the case of the separation filter array, the above-described interpolation calculation is performed for each pixel position of the all-pixel reading photoelectric conversion unit.
And W and G ', or W' and G, or Ye and Cy ',
Alternatively, Ye ′ and Cy are aligned. Where
It is considered that the spectral characteristics of W + G and Ye + Cy are almost equal.
Therefore, at each pixel position, W ′ + G or W +
By using G ′, Ye + Cy ′, or Ye ′ + Cy as a luminance signal, a luminance signal with less moire in the horizontal and vertical directions can be obtained. Since the low-pass filter for obtaining WL, GL, YeL, and CyL in the above-described interpolation calculation has a zero point at the horizontal 1 / 2Px, the low-pass filter has a higher horizontal position than the case where the color separation filter array of FIG. 9 is used. Moiré near horizontal 1 / 2Px is suppressed by a low-pass filter having a wide band in the direction. for that reason,
As compared with the case of using the color separation filter of FIG. 9, the range of generation of the false signal by the interpolation calculation is narrow, and the moire near the horizontal １ ／ Px can be suppressed.

From the W, G, Ye, Cy signals subjected to the interpolation operation in the interpolation operation means 14, the luminance signal generation means 1
5, a luminance signal is generated, and the high-pass filter 16
RL, GL, B
The signal equally added to L is output from the output terminal 18.

(Embodiment 2) FIG. 2 shows an embodiment in which a correction for suppressing a false signal at a color change point is performed. Hereinafter, the embodiment of FIG. 2 will be described.

In FIG. 2, reference numeral 19 denotes a means for detecting a color change point. Reference numeral 20 denotes a transversal filter whose tap coefficient is determined by the output of the color change point detecting means 19. FIG. 7 is a block diagram showing an example of the configuration of the color change point detecting means 19 and the transversal filter.
The false signal at the color change point can be suppressed by detecting the color change point and subjecting the W, G, Ye, and Cy signals at the color change point to low-pass filtering. FIG. 7 shows a false signal at a vertical (horizontal line) color change point.
W: An example of a circuit that suppresses a false signal generated due to a change in G.

The color change point detecting means 19 detects the W: G change point by monitoring the ratio of WL / GL for two adjacent upper and lower lines. Color detecting means 19
The operation in is represented by the following equation. Where W
L1 and GL1 are WL and GL corresponding to the same pixel position, and WL2 and GL2 are W corresponding to the position of a pixel one pixel above or one pixel below the pixel corresponding to WL1 and GL1.
L and GL are shown.

Α = log {(WL1 / GL1) / (WL2 / GL2)} a = F (α) Here, F (α) continuously changes from 0 to 、, and α
Is a function that becomes larger as the absolute value of becomes larger. FIG. 8 shows an example of the function F (α).

At the horizontal line color change point, a false signal is generated near the horizontal spatial frequency 1 / 2Px. Therefore, a false signal can be suppressed by suppressing the vicinity of the horizontal spatial frequency 1 / 2Px. The transversal filter 20 in the example of FIG. 7 is a filter having the characteristic of H (z) = a / 2 + (1-a) z- ¹ + (a / 2) z- ² . The transversal filter may be constituted by a digital FIR filter.

FIG. 7 shows an example of a circuit for suppressing a false signal generated at a change point of W: G. A false signal at a change point of Ye: Cy can be suppressed by the same means.

(Embodiment 3) A false signal generated at a color change point becomes larger as the ratio (ie, color) of the color signal fluctuates more. By adding a certain value to all of the output signals corresponding to the color separation filters W, G, Ye, and Cy in FIG. 4, the variation in the ratio of the color signals can be suppressed, and the false signal can be suppressed.

(Embodiment 4) Among false signals generated at a color change point, a false signal generated due to a change in the r (red) and b (blue) components of the incident light is a color-selective optical low-pass filter. 3, can be suppressed by inserting between the CCD and the lens (not shown) as shown in FIG. 3 to selectively restrict the spatial frequency high frequency components of r and b. FIG. 3 is a block diagram showing the embodiment.

Although the embodiment described above is an example in the case of M = 4, N = 2, other combinations of M and N may be adopted. In the above description, the color separation filter shown in FIG. 4 is used. However, if the color separation filter has no color sampling carrier in the horizontal and vertical directions of the spatial frequency, another color separation filter array is used. No problem.

[0033]

As described above, according to the solid-state imaging device of the present invention, a predetermined color separation filter having a repetitive array of M pixels × N pixels, and an optical / electrical All-pixel read-out photoelectric conversion means for performing conversion, and color signal generation means for generating a color-separated color signal by performing two-dimensional operation on a signal from the K × L pixel area subjected to light-to-electric conversion A pixel interpolation operation by multiplying an output pixel signal of the all-pixel reading photoelectric conversion unit by a ratio of a color signal generated by the color signal generation unit or a ratio of a signal obtained by addition and subtraction of the color signal. And an output of the pixel interpolation calculating means.
Force signal and the output signal of the all-pixel readout photoelectric conversion unit.
To generate a luminance signal by adding
And a two-dimensional operation on an output signal of a predetermined color separation filter array to obtain a color signal in which color moiré is suppressed, and a color signal with less color moiré is provided. By performing interpolation calculation using
It is possible to obtain a high-resolution luminance signal with little luminance moire in the horizontal and vertical directions.

In particular, in the case of a color separation filter in which color carriers do not exist in the horizontal and vertical directions of an image, a signal free of color moire in the horizontal and vertical directions can be obtained by combining with a color selective optical low-pass filter. By performing a predetermined pixel interpolation operation in accordance with the ratio of the color signals with less color moiré, a luminance signal with suppressed luminance moiré can be obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a first embodiment of a solid-state imaging device according to the present invention;

FIG. 2 is a block diagram showing a second embodiment of the solid-state imaging device according to the present invention;

FIG. 3 is a block diagram showing a third embodiment of the solid-state imaging device according to the present invention;

FIG. 4 is a diagram showing an example of a color separation filter array when M = 4 and N = 2;

FIG. 5 is a sampling carrier diagram when the color separation filter array shown in FIG. 4 is used.

FIG. 6 is a diagram illustrating weighting for W, G, Ye, and Cy signals;

FIG. 7 is a circuit diagram showing an example of a circuit for suppressing a false signal generated at a change point of W: G.

FIG. 8 is a diagram showing an example of a function F (α).

FIG. 9 is a diagram showing an example of a color separation filter array used in a color difference sequential method.

10 is a sampling carrier diagram in the color separation filter array of FIG.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 Color separation filter 12 All pixel readout imaging device 13 Color signal generation means 14 Pixel interpolation calculation means 15 Luminance signal generation means 16 High pass filter 18 Output terminal 19 Color change point detection means 20 Transversal filter 22 Color selective optical low pass filter

Continuation of front page (72) Inventor Shogo Sasaki 1006 Kadoma, Kazuma, Osaka Prefecture Matsushita Electric Industrial Co., Ltd. (56) References JP-A-4-304089 (JP, A) JP-A-5-145934 (JP) , A) (58) Field surveyed (Int. Cl. ⁷ , DB name) H04N ^9/ 04-9/11

Claims (4)

(57) [Claims] 1. A predetermined color separation filter having a repeating arrangement of M pixels × N pixels, all-pixel read photoelectric conversion means for performing light-to-electric conversion on incident light passing through the color separation filter, and light-to-electric conversion A color signal generating means for generating a color-separated color signal by performing a two-dimensional operation on the signal from the K × L pixel area subjected to Te, a pixel interpolation operation means for performing pixel interpolation operation by multiplying the ratio or ratios of the signal obtained by subtraction of the chrominance signal generated color signal in the color signal generating means, the pixel interpolation calculation means
Output signal and output signal of the all-pixel readout photoelectric conversion means
And generate a luminance signal by adding
A solid-state imaging device comprising: a luminance signal generation unit . 2. A signal processing means A for monitoring a horizontal or vertical color change point by performing a predetermined operation on an output of the color signal generating means, and a frequency characteristic processing for an output of the luminance signal generating means. , And the signal processing means B is constituted by a transversal filter whose coefficient is variable, and the tap coefficient of the transversal filter is determined by the output of the signal processing means A. The solid-state imaging device according to claim 1, wherein: 3. The method according to claim 1 , wherein said pixel interpolation calculating means reads said all pixels.
3. The solid-state imaging device according to claim 1, wherein a fixed value is added to an output pixel signal of the output photoelectric conversion unit . 4. The solid-state imaging device according to claim 1, wherein light passing through the color-selective optical low-pass filter is used as incident light.