JPS61227492A - Solid-state image pickup device - Google Patents

Abstract

PURPOSE:To decrease remarkably color moires of a single board color camera by designing the titled device that a signal interpolation hardly causing a false signal is applied even to an object having a pattern changed in a width several times of the picture element interval of a solid-state image pickup element. CONSTITUTION:An output signal (a) when an image of a dark/light object is formed on a solid-state image pickup element 1 is fed sequentially to delay circuits 2a-2f connected in series. The delay quantity of each delay circuit is set equal to the sampling period of a picture element signal. B, G, R picture element signals obtained from output lines 3a-3g are processed by a subtraction circuit 4, an amplifier circuit 5 having amplification factor of 1/3 and 2/3, an adder circuit 6, a division circuit 7, an adder circuit 8, a multiplication circuit 9 and an amplifier circuit 10 having an amplification factor of 1/2, and the output signal is fed to a gate circuit 11 together with a signal from an output line 3d. A signal from an oscillation circuit 12 synchronously with the drive pulse of the solid-state image pickup element 1 controls the gate circuit 11 to obtain the R, G, B signals with arranged phase where the sampling frequency shown in signals e-g is equal to the sampling frequency of the picture element and generation of an extreme false signals is suppressed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of Invention Q] The present invention relates to a color imaging device, and particularly to a method and device for reducing color moire suitable for a single-chip color camera using a solid-state imaging device.

[Background of the invention]

Currently, single-chip color cameras that obtain color video signals from a single solid-state image sensor are in practical use. In such a camera, a plurality of color signals are obtained by periodically associating several types of color filters with different transmitted light to each pixel of a solid-state image sensor.

Therefore, the spatial sampling frequency of each color signal is lowered to a fraction of the pixel sampling frequency, and color moire is likely to occur.

A method for reducing color moire in a single-panel color camera is disclosed in, for example, Japanese Patent Laid-Open No. 131819/1983.

This method performs the following operations. For example, if the color filters shown in Figure 1 are combined so that each filter piece corresponds to a pixel of a solid-state image sensor on a one-to-one basis, Figure 2 (a)
The signal shown is obtained. Figure 2 obtained by separating this 6)
For the R, G, and B signals shown in ~(d), for example, at time t1
An R signal is obtained at t, but G and B signals are not obtained, and at time t
! The G signal is obtained, but the R1B signal is not obtained. For example, at time t1, the magnitude of R1 is compared to see whether it is closer to Re or US, and if it is closer to Ro, it is Go or B.

Then, if it is close to R or neither, G and B signals are interpolated using Gl and Bt. With these actions, R, G,
When the B signal is interpolated, three signals with a high sampling frequency and the same phase are obtained as shown in FIG. 2(e) to @. As a result, at the boundaries of objects where the interval of change is sufficiently large compared to the repeating period of pixels, the signal that should originally be obtained is interpolated almost correctly, reducing color moire that occurs at these boundaries. .

However, if the subject has a pattern that changes at an interval several times the pixel repetition period, interpolation cannot be performed correctly, and there is a problem in that color moire increases.

For example, when a bright and dark subject is imaged on the color filter shown in FIG. 1 in the relative relationship shown in FIG. 3, a signal shown in FIG. 4(a) is obtained. When this signal is separated into t-R, G, and B and subjected to the conventional processing described above, the signals shown in FIGS. 4(b) to 4(d) are obtained. Comparing these signals with the subject of FIG. 3, the G signal at t3 is clearly different from the signal that should originally be obtained. This result t
In No. 3, the signal appears as if a green object was imaged, and false colors appear on the reproduced image.

[Purpose of the invention]

The purpose of the present invention is to reduce color moire in a single-chip color camera by performing signal interpolation that prevents the generation of false colors even for objects with patterns that change at intervals several times the pixel thickness period of a solid-state image sensor. The purpose is to provide a method.

[Summary of the invention]

In order to achieve the above object, the solid-state color imaging device of the present invention samples and separates a plurality of color signals sequentially obtained from a solid-state imaging device at each relevant phase, and a signal at a specific time of an arbitrary color signal and a signal before that. Find the amount of change between the same color signal obtained and the same color signal obtained later, and calculate the magnitude of the color signal obtained before and after the other color signal at a specific time and the time at which it is obtained. The feature is that interpolation is performed using a signal obtained from the difference and the amount of change described above.

[Embodiments of the invention]

Hereinafter, the present invention will be explained in detail using an embodiment having the configuration shown in FIG.

In FIG. 5, it is assumed that the solid-state image sensor 1 is combined with the color filter shown in FIG. 1, for example. At this time, when a bright and dark subject is imaged in the relative relationship shown in FIG. 3, a signal shown in FIG. 6(a) is obtained from the output. The delay circuit 2a is connected in series.

Add to 2b, . . . , 2f sequentially. Here, the delay amount of each delay circuit is set equal to the sampling period of the pixel signal. As a result, for example, at time t6, the output lines 3a, 3
b, ..., 3g to time t6゜tS, respectively
+ ”””e”B obtained at 0, ,G,, approximately 2゜B
t e Gl * Rt e B(l pixel signals are obtained simultaneously.

Note that the R, , G, and B signals obtained by separating the signal in FIG. 6(a) by sampling are shown in FIGS. 6(b) to 6(d), respectively. Here, taking the signal obtained at time t3 as an example, B1 is obtained for the B signal, but there are no R and G signals. First, if we approximate that the change in the subject between Bo and 81 is linear, then time tl and 1. is B, -B
1 and 1, respectively, so the magnitudes Ba1 and Ba2 of the B signal at these times are as follows.

Similarly, if it is approximated that the change in the subject between B+ and B2 is linear, time t4. Magnitude of B signal of ts B a 4
s B a @ are as follows.

Bo %B1 question and Bt x13. Considering that the changes in the B and G signals during that period are also linear like the B signal, for example, the relationship between the R signal (Rs) obtained at time tl and the signal (R,,3) obtained at tl is as follows. The following relationship can be used.

From equations (1) and (5), equation 3 can be obtained using the following relationship.

The relationship between the R signal (R-'a) obtained at time t3 after rejuvenation and the R signal (Rz) obtained at time t4, and time 1. G signal (G-s, G-'s) at time t! and 1. The relationship with the G signal (G1 and G snow) obtained in is as follows.

Here, if the R signal (Rss) at time t3 is determined as the average value signal of R a 3 approximated from R1 obtained at time 1K and R, '3 approximated from R snow obtained at time t4, the following is obtained. It is.

. . . αO Similarly, the G signal (Gas) at time t3 is as follows.

(11) The above explanation was made using the signal to be obtained at time t3 as an example, but signals to be obtained at other times can also be expressed as R,, G,
By replacing the relationship of B according to the order in which each color signal is obtained, the G0,00 formula can be applied as is.

In the embodiment shown in FIG. 5, which is an example of realizing the above signal processing, Be obtained from the output line 3g, for example at time t6, is added to the positive polarity input of the subtraction circuit 4a, and B obtained from the output line 3d is input to the negative polarity. Add to input and add to 5b. Similarly, Bt obtained from the output line 3a is added to the positive polarity input of the subtraction circuit 4b, and B1 is added to the negative polarity input. The resulting output signals of the subtraction circuit 4b are applied to amplifier circuits 5C and 5d having an amplification factor of 1 and 1, respectively. Furthermore, the output signals of the amplifier circuits 5a, 5b, 5c, and 5d are added to the adder circuits 6a and 6b together with B1 obtained from the output line 3d, respectively.
, 6c, 6d.

Here, the output signal of the adder circuit 6b is added to the division signal input of the divider circuit 7a, while R1 obtained from the output line 3f is added to the divider signal input. The output signal of the adder circuit 6C is added to the division signal input of the divider circuit 7b, and the signal obtained from the output line 3C is added to the divider signal input. The output signals of the division circuits 7a and 7b obtained in this way are added together by the addition circuit 8a, and together with B1 obtained from the output line 3d, the multiplication circuit 9a
In addition to this, a signal corresponding to R13 expressed by equation 00 is obtained from the output of the amplifier circuit 10a with an amplification factor of Y.

Similarly, the output signal of the adder circuit 6a is added to the division signal input of the divider circuit 7C, G1 obtained from the output line 3e is added to the divider signal input, and the output signal of the adder circuit 6d is added to the divider signal input of the divider circuit 7C.
In addition to the divided signal input, G: obtained from the output line 3b is added to the divided signal input. The output signals of the divider circuits 7C and 7d are added together in the adder circuit 8b, and are added to the multiplier circuit 9b together with BK obtained from the output line 3d. After passing through the amplifier circuit 10b with the amplification factor i, the output is Gt expressed by the formula 00.
A signal corresponding to 3 is obtained.

Note that the output signals of the amplifier circuits 10a and 10b are sent to the gate circuit 11a together with the signal obtained from the output line 3d.

11b, llc. At this time, the gate circuits 11a, 11b, and 11C are controlled by a signal from the oscillation circuit 12 synchronized with the driving pulse of the solid-state image sensor, so that the B signal, G signal, and R signal can be obtained separately from their respective outputs. .

As a result of the above, from the signal in Figure 6(a), (e) to (g)
In this case, the sampling frequency is equal to the pixel sampling frequency, and it is possible to obtain R, B, and G signals having the same phase. Although these signals are slightly different from the signals that should originally be obtained, such as the G signal corresponding to times t6 and t7,
Overall, generation of extremely false signals is suppressed compared to the conventional example shown in FIG. 4(C). As a result, false colors appearing on the reproduced image are reduced.

The embodiment shown in FIG. 5 is one method for realizing the αO9αυ formula, for example, as shown in FIG. 7, subtraction circuits 4a, 4b and addition circuits 6a, 6b are used.

Addition/subtraction circuit 13a has the functions of 5c and 6d.

13b, 13C, 13d, etc., G0゜00
Any configuration that satisfies the formula or a modified version of the following formula is possible.

・・・・・・・・・QO' ・・・・・・・・・αυ′ In the example shown in Fig. 5, the signal at a certain time is calculated as the average of the values predicted from the signals of the same color obtained before and after that time. However, it is also effective to add the values at a ratio corresponding to the difference between the time and the time when the previous and subsequent signals are obtained.

For example, when calculating the R signal corresponding to time t3, since the difference from time t3 to time t1 when R1 is obtained is twice the difference from time t4 when R snow is obtained, equation (6) predicted from R1 is used. P calculated by adding 1 times Ra 3 shown and 1 times R1'3 shown in equation (7) predicted from R2
L*'3 is used as an interpolation signal. Similarly, to obtain the G signal corresponding to time t3, Gt is measured using the G signal shown in equation (9).
, '3 is added and GA's is used as an interpolation signal. As a result, Rt's.

GA's are each expressed by the following formula.

・・・・・・・・・α2 ・・・・・・・・・(I3 An example for realizing the α2.13 formula is shown in FIG.

In the embodiment shown in FIG. 8, the signals obtained from the output lines 3f and 3c are transmitted to an amplifier circuit 13a with an amplification factor of 1 and an amplifier circuit 13 with an amplification factor of 2 times. b to the green signal inputs of the divider circuits 7a and 7b, and at the same time, by changing the amplification factor of the amplifier circuit 10a to 1, the Rt expressed by the α2 formula is
's get. Similarly, the signals obtained from the output lines 3e and 3b are passed through an amplifier circuit 13C with a double amplification factor and an amplifier circuit 13d with a single amplification factor, respectively, to divider circuits 7c and 7.
In addition to inputting the green signal d, the amplification factor of the amplifier circuit 10b is changed to 0 to obtain Gt'sf shown by the α2 formula.

The R, B, and G signals obtained in this way are shown in Fig. 6 (h) to (j
) as shown. These signals, like the signals (e) to (g) obtained in the embodiment shown in FIG. 5, are close to the signals that should originally be obtained. Similarly to the embodiment shown in FIG. 5, configurations other than those shown in FIG. 8 or configurations corresponding to modified expressions are possible as long as the relationships of equations (1B and 03 are satisfied).

Another method for synthesizing interpolated signals is the difference between the value of a signal at a certain time predicted from signals of the same color obtained before and after that time, the magnitude of the signal obtained at that time, and the magnitude of the same color signals obtained before and after that time. It is effective to find it from the coefficient corresponding to .

For example, when calculating the R signal corresponding to time t3, the difference between B1 obtained at time t3 and Bo obtained at time 1o (IBo Bs1) and the difference between Bl and B2 obtained at time t6 (IB-B11) are calculated. and express how close B1 is to Be or Bt. This fi' Rs3 shown in equation (6) predicted from Rs and R1'3 shown in equation (7) predicted from Rs
When used as the coefficient of , the interpolated signal R1 "3 at time t3 becomes the following equation.

......Similarly to αXun, the G signal (Gt"3) corresponding to time t3 is GGl
The above coefficients may be used for Ga3 shown in equation (8) predicted from 02 and G,'3 shown in equation (9) predicted from 02,
It becomes as follows.

・・・・・・・・・(is α4. An embodiment for realizing α9 is shown in FIG.

In FIG. 9, for example, the signal obtained from the outputs of the subtraction circuits 4a and 4b at time t6 is Be-B5.
and Bs Bt, respectively, to the absolute value circuit 14.
a, 14b and converted into an absolute value signal. Furthermore, the outputs of the absolute value circuits 14a and 14b are added to the adder circuit 15, and at the same time they are added to the divided signal inputs of the divider circuits 16a and 16b, respectively, and the output signal of the adder circuit 15 is added to the divider signal input. A corresponding signal is obtained. Furthermore, the division circuit 16
The output of a is applied to the multiplication circuit 17b together with the output of the division circuit 7b, and the obtained output signal is applied to the addition circuit 8a. At the same time, the output of the division circuit 16b is applied to the multiplication circuit 17a together with the output of the division circuit 7a, and this output is applied to the addition circuit 8a. Similarly, the output of the division circuit 16a is added to the multiplication circuit 17d together with the output of the division circuit 7d, and the obtained output and the output of the division circuit 16b are combined with the output of the division circuit 7C to the multiplication circuit 1.
7C and the obtained output are added to an adder circuit 8bK.

If the outputs of adder circuits 8a and 8b are added to multiplier circuits 9a and 9b together with the output obtained from output line 3d, α4 is obtained from the output.
), (R1"3°Gi"3 shown in equation 15 is obtained. At this time, the amplification factors of the amplifier circuits IQa and 10b may be set to 1 or may be omitted.

Note that the detector 18 detects the case where both the outputs of the absolute value circuits 14a and 14b are KO, and in this case, the output is 1.

Control is performed so that a signal corresponding to 100 is obtained from the outputs of the division circuits 16a and 16b.

The R, , B, and G signals obtained by the above operation are shown in Figure 6.
As shown in ) to ), the signal is very close to the signal that should originally be obtained, so it is possible to significantly reduce the occurrence of false colors on the reproduced image. Also, as long as the relationship (14) and (151) are satisfied, similar results can be obtained with another configuration or another configuration in which the formula is modified, as in the embodiment shown in FIG.

Furthermore, in the embodiments of FIGS. 5 and 7 to 9, the delay circuit 2a
Although the delay circuits 193 to 2f are connected in series, the delay circuits 193 to 19ft whose delay time is 1 to 6 times the pixel sampling period may be connected in parallel as shown in FIG.

Note that although it is circuit-wise difficult to implement the multiplication circuit, the division circuit, etc. with an analog circuit, arithmetic processing using the digital signal is possible by converting the output of the solid-state image sensor 1 into a digital signal.

Further, although the present invention has been described using the color filter shown in FIG. 1 as an example, it is clear that the present invention can be applied to time-series signals in which several kinds of signals repeat condylarly.

Therefore, those using complementary color filters such as (a) and Φ) in Fig. 11, those with *b return other than 3 pixels such as (C) and (d), and those that do not have the same color in the vertical direction such as (e) ,
It is applicable to a single-chip color camera combined with any color filter, such as one corresponding to a solid-state imaging element in which pixels are horizontally shifted every other line, such as (f).

〔Effect of the invention〕

As described above, according to the present invention, it is possible to perform signal interpolation in which K signal is less likely to occur even for subjects with patterns that change at several times the pixel spacing of the solid-state pixel sensor, and a single-chip color camera Color moiré can be significantly reduced.

[Brief explanation of the drawing]

Figure 1 is a diagram showing an example of a single-plate color filter, Figure 2 is a schematic diagram of signals to explain the operation of the conventional example, and Figure 3 is a diagram of a subject to point out problems with the conventional example. FIG. 4 is a diagram showing a schematic diagram of a signal obtained from the subject shown in FIG. 3; FIG. 11 is a diagram showing a schematic diagram of a signal obtained in an embodiment of the present invention, and FIG. 11 is a diagram showing an example of a color filter of a single-chip color camera to which the present invention is applicable. 4... Subtraction circuit, 6... Addition circuit, 7... Division circuit, 8... Addition circuit, 9... Multiplication circuit, 11...
Gate circuit, 13... Amplification circuit, 14... Absolute value circuit, 15... Addition circuit, 16... Division circuit, 17.
...Multiplication circuit, 18... Steam 1 Zutomi Z Eyelin 0. ：・ 、・′、；°゛、・ ・ ． Tomi 3 Figure'fJd Figure (d,) (, -7,,,,, -1:・ :100 5 Scale q6 Group 6 Figure (ML)B ・effect ・Mimi・
2 m″if group 7 prisoner g4 ρ summer g figure 14i υ ￥J lo figure 11

Claims (1)

[Scope of Claims] 1. An image sensor in which pixels are arranged at least in the horizontal direction, and a filter in which a plurality of types of micro optical color filters whose transmitted light differs from each other are repeatedly arranged in correspondence with each pixel of the image sensor. In the imaging device using the above, the pixel signals sequentially obtained from the image sensor are separated into a plurality of series of pixel signal groups corresponding to the plurality of types of micro optical color filters, and any one of the plurality of series of pixel signal groups is separated. a first pixel signal obtained at a specific time belonging to the first pixel signal group selected in , a second pixel signal obtained immediately before the first pixel signal belonging to the first pixel signal group, and the above. a third pixel signal obtained immediately after the first pixel signal belonging to the first pixel signal group; and a second pixel signal different from the first pixel signal group among the plurality of pixel signal groups the fourth signal obtained immediately before the specific time belonging to the group and the second signal belonging to the group;
A fifth signal obtained immediately after the specific time belonging to the pixel signal group of is obtained, and corresponds to a second time at which the fourth signal obtained from the first signal and the second signal is obtained. a first approximate signal obtained from the ratio of the sixth signal to the first signal and the fourth signal, the first signal and the third signal;
A second approximate signal obtained from the ratio of the first signal and the seventh signal corresponding to the third time at which the fifth signal obtained from the fifth signal is obtained. Obtained, the above 1st
An interpolation signal corresponding to the specific time at which the first signal of the second pixel signal group is obtained from the approximate signal and the second approximate signal is determined. 2. The first signal is obtained by linearly approximating the signal change between the fourth time when the second signal is obtained and the specific time.
Determine the sixth signal from the signal and the second signal,
The seventh signal is obtained from the first signal and the third signal by linearly approximating the signal change between the fifth time when the third signal is obtained and the specific time. A solid-state imaging device according to claim 1, characterized in that: 3. The solid-state imaging device according to claim 1, wherein the interpolation signal is a signal corresponding to an average value of the first approximation signal and the second approximation signal. 4. A signal obtained by amplifying the first approximate signal with an amplification factor corresponding to the difference between the third time at which the fifth signal is obtained and the specific time, and the second approximate signal are amplified by the fourth approximation signal. Claim 1, characterized in that the interpolation signal is obtained by adding the second time at which the signal is obtained and a signal amplified with an amplification factor corresponding to the difference between the specific time and the specific time. solid-state imaging device. 5. Obtain a first difference between the second signal and the first signal and a second difference between the third signal and the first signal, and convert the first approximate signal to the second signal. A signal amplified with an amplification factor corresponding to the magnitude obtained by dividing the difference by the sum of the first difference and the second difference, and the second approximate signal.
and a signal amplified by an amplification factor corresponding to the magnitude divided by the sum of the second difference and the interpolated signal is obtained by adding the signal amplified by an amplification factor corresponding to the magnitude divided by the sum of the difference and the second difference. Imaging device. 6. When the first difference and the second difference are both 0, a signal corresponding to 1/2 of the sum signal of the first approximation signal and the second approximation signal is used as the interpolation signal. A solid-state imaging device according to claim 5, characterized in that: