JPS6081993A - Solid-state color image pickup device - Google Patents

Abstract

PURPOSE:To enable to form correct luminance signals and chrominance signals at high resolution when driving a solid-state color image pickup element in field accumulation mode by using color filters arranged in prescribed manner. CONSTITUTION:W picture elements are arranged in every other element in each vertical row material eight picture elements, two picture elements in horizontal direction and four picture elements in vertical direction, a unit of repetition, and one element each of Ye picture element and Cy picture element exists in a repetition unit and separated by one horizontal row between. G picture elements are allotted at remaining positions. A solid-state image pickup element provided with color filters of such arrangement is driven in field accumulation mode, and forms a row of signal packets to which vertical two picture elements are added. The signal from the element 1 is branched, and one is led directly to sample and hold circuit 3-7 and another is led to the circuits through a delay circuit 2. Luminance signals Y of wide band are formed through a BPF8 and an LPF9. Further, R signals and B signals from which false signals are removed are obtained through a switching circuit 14.

Description

[Detailed Description of the Invention] Technical Field The present invention relates to a solid-state color imaging device with a color separation function, and more particularly to a solid-state color imaging device driven in field accumulation mode. To obtain a color video signal using a single image sensor, a method is usually adopted in which a color filter is placed on the light-receiving surface of the image sensor and the subject image is spatially sampled for each color component. The arrangement depends on the color image quality obtained, but on the other hand, the arrangement is also limited by the structure and driving method of the solid-state image sensor. Therefore, various color filter arrays have been proposed that are tailored to the characteristics of each solid-state image sensor.

□Among the driving methods for solid-state image sensors, field accumulation mode driving methods have recently attracted particular attention.

(1983 Television Society National Conference 3-16.3-
17.4-9.4-10, etc.) In other words, in the field accumulation mode driving method, by adding and reading the signals of two vertically adjacent pixels, it is possible to accumulate and read out the video signal in the field period. Compared to the frame period accumulation mode, there is a slight afterimage and less moiré in the vertical direction.There are two main types of methods that have been proposed as color filter arrays that can be applied to the field accumulation drive method: being classified. One of them is, for example, Fig. 1(a) K shown in Japanese Unexamined Patent Application Publication No. 1989-57.
As in the arrangement described in Publication No. 3968/1, the two color signals themselves are obtained as modulation components alternately every horizontal scanning period (IH), and the other one is, for example, the first color signal.
As shown in the arrangement described in Japanese Patent Laid-open No. 51-aa9Qo shown in FIG. 5B, a sum signal and a difference signal of two color signals are obtained alternately as modulation components for each IH. here,
R: red, G: green, B: blue, W = R + G
10B represents a transparent filter, Ye=R1G represents a yellow filter, and Cy=G+B represents a negative color filter.

That is, in FIG. 1(a), (W0Ye)-(Cy1G)=2 in both the odd field and the even field on the odd line.
R is obtained as a modulation component, and on even lines t/+C
y) −(Ye+G)=2B is obtained as a modulation component. Therefore, if you synchronize using an IH delay line, R and B
I get a signal. On the other hand, in FIG. 1(b), both the odd field and the (i-number field) are n(Cy+G
)−(Ye×G)=BR is obtained as a modulation component, and for even lines, (Cy×Ye)−(GyG)=B×R is obtained as a modulation component. Therefore, a 52B signal is obtained by taking the sum between the real time signal and the IH delay signal, and a 2R signal is obtained by taking the difference.

Note that the luminance signal is obtained using the same method for both types. In other words, since all pixels contain the G component that makes up the majority of the luminance signal, the entire signal packet sequence added by two vertical pixels is passed through a filter and only the tone component is removed, resulting in a broadband signal. can get.

Comparing the two types of arrays shown in Figure 1 (b)
), one color signal must be obtained by calculations between four signal packets spanning both horizontal and vertical directions, whereas in Fig. 1(a), it is obtained by calculations between two horizontally adjacent signal packets. Therefore, it is considered that the method shown in FIG. 1(a) is superior in that it generates fewer false signals.

The array VC&'1 in FIG. 1(a), however, has the following drawbacks. First, when examining the frequency response of each of the G, R, and B components in the vertical two-pixel addition signal string, the result is as shown in FIG. Therefore, the low frequency component of the luminance signal is R+2G10B, which is the ratio of the brightness signal component in the standard television system, which is 0.3OR100.59G10, considering that the response of the image sensor is generally low in the B region. , ] I
The value is close to B, so there is no problem. On the other hand, the high frequency component 1r of the luminance signal, the state of 1:2G-t-H and the state of 20+R are IH@[alternating. Therefore, not only the components are incomplete, but also an unbalanced signal that differs from one to the other becomes an unbalanced signal.

Furthermore, the following problems exist in color signals as well.

In FIG. 2, the color signal is extracted as a modulation component at the Nyquist limit frequency (fN)K, but when broken down, it consists of the following units.

That is, the R signal is the average of w-cy and Ye-G, and the B signal U
This is the additive average of W-Ye and Cy-G. On the other hand, the spectral transmission characteristics of a color filter that can be realized are generally as shown in FIG. Here, the G filter is obtained by overlapping a Ye filter and a Cy filter. When the above calculation is applied to this spectral characteristic, almost pure R and B signals are obtained with Ye Gy C3'-G, but in addition to the R and B signals that should be obtained with W-Cy and W-Ye. Components of other colored lights are also mixed in. In other words, the modulation component should appear only when the R signal exists in (W 0 Ye) - (Cy 0 G) and the B signal exists in (w 0 cy) - (Ye 1 G). However, this means that modulation components will appear even when other colored lights are present, resulting in an error in color reproduction.

<Purpose of the Invention> The present invention was devised in view of the above-mentioned problems.The present invention maintains high resolution for both the luminance signal and the color signal, and the luminance signal always has constant correctness in both the low and high frequencies. The object of the present invention is to provide a solid-state color imaging device having a color filter array that can obtain a component ratio and a correct signal without false responses in terms of color reproducibility. ) shows an example of a color filter arrangement in the present invention. These are all horizontal 2
Pixels, 8 pixels (4 pixels in the vertical direction) are set as the 11-position of repetition, and within this tit position, W pixels are arranged every summer element in each vertical column.Ye pixels and cy pixels are the repetition unit. There is one element each in the array, occupying positions separated by an interval of VC + horizontal columns. Six pixels are assigned to the remaining positions. Adding these arrays between two vertically adjacent pixels When driven in field accumulation mode, the signal processing method and obtained signals are equivalent and exhibit the excellent characteristics shown below.

The frequency response of the vertical two-pixel addition signal in each array of FIG. 4 (aL(b)) is as shown in FIG. 5.

From this, the R signal and the B signal are obtained alternately for each IH as a modulation component at the frequency fNVC, and the luminance signal is obtained as a wideband signal from which only the color modulation component is removed. That is, a video signal is basically obtained in the same format as the arrangement shown in FIG. 1(a), and the pixels involved in color signal formation are close to each other, which is a feature of the arrangement shown in FIG. 1(a). In addition, the arrays shown in FIG. 4 completely eliminate the drawbacks of the array shown in FIG. 4(a).

First, consider the luminance signal. In this arrangement, W0G=R+
Since a 2G1B signal is always obtained every other signal packet, a signal obtained by sampling and holding this signal is used as a luminance signal low-frequency component. The band of this signal exists up to fN/2. As a result, the luminance signal low frequency 1l-tR9G and B signal component ratios are not only appropriate but also always constant. Further, as the luminance signal high-frequency component, a signal obtained by removing the low-frequency component and the color modulation component near fN from the two-pixel addition signal is used.

As is clear from Fig. 5, this signal is always constant R+
It is a 200B signal, which matches well with the low frequency signal and is an almost ideal brightness high frequency signal.

Next, let's consider color signals. The color signal is VCf as mentioned above.
It is obtained as a modulation component near N, but if you break it down and consider it, the R signal is from Ye-GVc, and the B signal is from Cy-G.
It can be seen that the result is obtained by By the way, since the spectral characteristics of the color filter are as shown in FIG. 3, the R signal and B signal in this case exhibit almost ideal spectral characteristics, and other color light components are hardly mixed. As shown above, each of the arrays shown in FIG. 4 is found to be an extremely excellent array that inherits the advantages of the 1n(a) array without any drawbacks.

FIG. 6 shows an example of a circuit block diagram for actually carrying out signal processing in the color filter array of the present invention shown above. A solid-state image sensing device equipped with color filters arranged as shown in FIGS. 4(a) and 4(b) is driven in a single field accumulation mode to form a signal packet string in which two vertical pixels are added. The signal from Motoko is branched, one is directly,
The other one is guided through the IH delay circuit 2 to the sample and hold circuit 3.4.5°6.7. In circuit 3, all signal packets are sampled and held and guided to bandpass filter 8. Only the uW+G signal packet is sample-held in the circuit 4 and guided to the low-pass filter 9. The filter 8 removes the base frequency component and frequency components near the elm to form a high-frequency luminance signal component. A filter 9 passes only low frequency components of 1 dfN/2 or less to form a low frequency luminance signal component. The signals from filters 8 and 9 are summed to form a broadband luminance signal Y.

On the other hand, since the HW+Ye to v+cy signals are sampled and held in the circuit 5, the output of the circuit 4 is subtracted from the output of the circuit 5, and R or B signals IO which alternate for each IH are obtained and guided to the switch circuit 14. Similarly, circuit 6
Then, the w+cy or W+Ye signal is sampled and held, and the circuit 7TfdW+G signal is sampled and held, so the output of circuit 7 is subtracted from the output of circuit 6, and I1
B to R signals II alternating every 1 are obtained.

Here, when the color signals that alternate every IH are synchronized with the H delay signal, the following false color signal is generated. That is, for an optical image with a large change in luminance in the vertical direction, a phenomenon occurs in which the color signal falsely responds near the point of change. This problem can be solved, for example, by the method described in Japanese Patent Application Laid-Open No. 58-90884. That is, by adding the luminance vertical contour signal to the chrominance signal delayed by ]H, the false response component of the chrominance signal is removed. This technique can be used here as follows. By subtracting the output of the circuit 7 from the output of the square circuit 4, a luminance vertical contour signal 12 is obtained. Therefore signal 1
2'fr: By adding it to the signal 11 after adjusting it to an appropriate magnitude, the IH delay signal 13 in which the false color signal impulse T has been canceled is obtained, and is guided to the switch circuit 14. The switch circuit 14 performs a switching operation for each IH, and outputs R and B signals that are synchronized and have false signals removed.

<Effects> As shown above, according to the present invention, it is difficult to drive a solid-state color image sensor in field accumulation mode.
It becomes possible to form correct luminance signals and color signals with high resolution without false responses.

In addition, the color filter used is a 0V) filter in which 1/2 of all pixels are transparent, which also has the advantage of high light utilization efficiency and high sensitivity.

[Brief explanation of drawings]

Figures 1(a) and (b) are diagrams showing examples of color filter arrays conventionally used in field accumulation mode, and Figure 2 is the frequency of each component in the output signal in the case of Figure 1(a). Figure 3 is a diagram showing the response, Figure 3 is a diagram showing the spectral transmission characteristics of commonly used color filters, Figure 4 (a) and (b)
is a diagram showing the arrangement of color filters according to the present invention, Figure m5 is a diagram showing the frequency response of each component in the output signal in the case of Figures 4 (a) and (b), and Figure 6 is a diagram showing the arrangement of color filters according to the present invention. This is a circuit block diagram for 0 1: solid-state image sensor, 2: IH delay circuit, 3, 4° 5, 6
．． 7: Sample and hold circuit, 8: Band pass filter, 9: Low pass filter, 14: Switch circuit. Agent Patent attorney Aihiko Fuku (2 others) Figure 1 Figure 2 Figure 314 (1) (1)) Figure 4 0 1 fs Figure 5 Figure 6

Claims (1)

[Scope of Claims] (1) In an imaging device consisting of a group of light receiving elements arranged in horizontal and vertical directions, each light receiving element is a first light receiving element sensitive to a first spectral band, a first light receiving element sensitive to a first spectral band, a second spectral band, and a second spectral band. a second light receiving element sensitive to a spectral band, a third light receiving element sensitive to a third spectral band, and a fourth light receiving element sensitive to all of the second and third spectral bands. The fourth light-receiving element occupies all the horizontal elements FJ every other horizontal row, and the first light-receiving element and the second light-receiving element are arranged between the remaining horizontal rows. obtained from the light receiving element group, wherein horizontal rows in which the first light receiving element and the third light receiving element alternate in each element and horizontal rows in which the first light receiving element and the third light receiving element alternate in each element are repeatedly arranged. A method for forming color signals based on signals obtained by
J-=To input hζ force Tan, remainder ~ 4b yen great dispute 7-(2. In the solid-state color imaging device according to claim 1, the signals of two pixels adjacent in the vertical direction are added and read out. (3) In the solid-state color imaging device according to claim 2, a signal obtained by adding the signal of the first light receiving element and the signal of the fourth light receiving element is sampled. A solid-state color imaging device characterized in that a low-frequency component of a luminance signal is formed from a signal obtained by applying a luminance signal.