JPS6081991A - Color solid state image pickup device - Google Patents

Abstract

PURPOSE:To make highly efficient A/D conversion and to enlarge a time margin by changing the number of digital output bits of each A/D converters and performing digital signal processing according to the kind of analog signal fed from a solid state image pickup element. CONSTITUTION:A row of filters of photodetector is arranged in repetition of R, G, B in the direction of scanning and three A/D converters 602, 603, 604 are provided in a signal transferring section 601. Signals taken out directly from the solid state image pickup element are R, G, B signals and portioned to converters 602-604 that process exclusive signals of R, G, B through the transferring section 601. They are A/D converted separately, and separated to Y signal, R-Y signal and B-Y signal by a digital signal processing section 605 and outputted as signals of NTSC standard system through a D/A converter 607. At this time, R, G, B signals are converted by the number of bits at least 1 bit smaller than the number of necessary bits when A/D conversion is applied to the Y signal. Thus, highly efficient A/D conversion is performed.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a color solid-state imaging device that can be used for digital signal processing in color video cameras and the like.

Conventional Structure and its Problems An example of the structure of a color mosaic filter that can generally be used in a color solid-state imaging device is shown in FIG. That is,
For the light receiving pixels 101 arranged two-dimensionally, a magenta filter M, 10 is placed on the left half of the first pixel 102 of the nH line.
3, a cyan filter CA 104 is provided in the right half of the second pixel 105, a green filter G106 is provided in the left half of the second pixel 105, and a yellow filter Y8107 is provided in the right half.

Further, a green filter G109 is provided in the left half of the first pixel 108 of the (n+1)H line, a cyan filter C110 is provided in the right half, a magenta filter Mg112 is provided in the left half of the second pixel 111, and a magenta filter Mg112 is provided in the right half. Yellow filter Y. 113 will be installed. nH line, (n+1)H
If the lines are a first filter row 114 and a second filter row 116, each of which is constructed by repeating the above-mentioned filter arrangement, then the mosaic color filter is created by vertically connecting these first filter rows and second filter rows. They are arranged sequentially in the direction.

With the filter having such a configuration, each light receiving element 1o1 can obtain a photoelectric conversion signal corresponding to the respective filter. FIG. 2 is a schematic diagram showing the element output of each line. rr 'J r b are red, green, and
Indicates the yellow signal level.

That is, the same element outputs from the nH line are sequentially taken out. Similarly, for the (n+1)H line, the luminance signal components of each horizontal line 5nH2S(n+1)H
becomes . However, K1. 2 is a constant +f, and is the frequency of the sampling frequency f80% of the solid-state image sensor.

The low frequency component Y of these signals is Y = K for each line.
1 (2r + 3 g + 2 b), which matches,
A luminance signal Y without line shading can be extracted by simply passing it through the low-pass 3 filter. In addition, the modulated color difference component appears as (r--") and (b once) are modulated with a frequency f of 2 2 waves every 1 ray/ray. Therefore, the color signal H,
It can be extracted using a bandpass filter.

FIG. 3 shows a specific example of a block diagram of an analog signal processing circuit that performs such signal processing, as proposed by the present applicant in Japanese Patent Application No. 57-22325. In the same figure, 30
Reference numeral 1 denotes a solid-state image sensor in which the color mosaic filter shown in FIG. 1 is glued onto a light-receiving element, and its output signal has a polar frequency f. passes through a low-pass filter 302, and only the luminance signal component Y remains.

On the other hand, the center frequency f. The bandpass filter 303 extracts the modulated color difference signal component spatially modulated by the color mosaic filter, detects it in the detection circuit 304, and r
9 and b-1 are sequentially taken out, and a 1H period delay line 305 and a switching circuit 306 that switches every 1H period produce simultaneous signals of the first color difference signal 4 and the second color difference signal 1. is generated and converted into an NTSC standard signal by an encoder 307 along with a luminance signal Y, and output from an output terminal 30B.

FIG. 3 is a block diagram showing typical conventional analog signal processing, but if this is simply replaced with digital signal processing, a block diagram as shown in FIG. 4 is obtained. That is, one analog signal read from the solid-state image sensor 401 is converted into a digital signal through one A/D converter 402. Then, the luminance signal and the color signal are separated into digital signal processing 403 and 404, and the color signal passes through the synchronization circuit 405, is mixed with the luminance signal in the encoder 406, and finally is processed by the D/A converter 40. ,
It is output as a signal according to the NTSC standard.

Reading from the solid-state image sensor is shown in FIG. When scanning each line, for example, the nH line 501, all pixel signals of the nH line are transferred to the signal transfer unit 501 at once.
02 and read out as one analog signal.

By performing digital signal processing as shown in Figure 4,
The solid-state image sensor can be controlled pixel by pixel. Furthermore, once the sampling frequency is set correctly, level and timing adjustments due to variations in signal circuits, as seen in analog signal processing, do not occur.

However, in the simply replaced digital signal processing as shown in FIG. The information is a mixed signal containing both luminance and chrominance signals. Originally, chrominance signals can be A/D converted using a smaller number of bits than luminance signals, but cannot be handled as mixed signals.

Therefore, in order to obtain sufficient information as a video, A
It is necessary to increase the number of digital output bits of the A/D converter 402 than when handling the nine color codes of the luminance signal separately, but this would reduce the time margin of the A/D converter, making it extremely difficult to implement. The problem is that it is difficult to

Furthermore, as the number of conversion bits increases, the power consumption of the A/D converter also increases, and the efficiency deteriorates significantly.

OBJECTS OF THE INVENTION The present invention obviates these conventional drawbacks,
The number of digital output bits is small, the time margin is large,
The present invention provides a color solid-state imaging device that can utilize an A/D converter with low power consumption.

Structure of the Invention The color solid-state imaging device of the present invention is provided with A
The A/D converter performs digital signal processing by changing the number of digital output bits, and is not simply replacing the conventional analog circuit with a digital circuit.
This is a converter with a reduced number of digital output bits.

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

As shown in FIG. 6, the filter rows of light receiving elements are arranged in a repeating pattern of R (red), G (green), and B (blue) in the scanning direction, and three A/D filters are arranged for the signal transfer unit 601. Converters 602, 603, 604 are provided. The signals directly extracted from the solid-state image sensor are R, G, and B signals, and these are passed through the signal transfer unit 601 to the A/D: +y Bartaro 02.603. 604
be distributed to. Then, these signals are separately A/D
It is converted and separated into Y signal, RY signal, and BY signal by digital signal processing unit 605, and encoder 6o6°D/A:
The signal is output as an NTSC standard signal via the ry bartaro o7.

The relationship between the read timing from the signal transfer unit 801 and the data sampling timing of the A/D converters 602, 603, and 604 is as shown in FIG. The sampling clock used in the A/D converter has different phases, and data is sequentially taken in at the rising edge of the clock.

Once the R, G, and B color signals are determined, the luminance signal Y is expressed as Y=0.3OR+0.59G+0.11B, so from the above formula, the signal must be changed to R to maintain the bit accuracy of the Y signal.
, G, and B, the Y signal itself can be converted into A/D
It can be seen that it is acceptable to perform conversion using at least one bit less than the required number of bits for conversion.

FIG. 8 schematically shows the configuration of the digital signal processing section 605, and shows that R, G, and B signals are converted into Y, RY, and BY signals.

Sult said that n bits are necessary to maintain the accuracy of the Y signal.
A/D converter 801.80 for R, G, B signals
The conversion bit numbers of 2.803 are n-2 and n-1, respectively.
, n-2 bits. If these R, G, and B signals are added as they are, a Y signal with n-bit precision can be obtained.

That is, this Y signal is expressed by the following equation.

Y=0.2SR+0.50G+0.25B For example, Y
Assuming that 8 bits are required to A/D convert a signal, the number of conversion bits for the R, G, and B signals is sufficient to be 6 bits, γ pit, and 6 bits, respectively. Although the accuracy is somewhat lower from a mathematical perspective, there is no problem in practical use. - In addition, the R-Y and B-Y signals are obtained by shifting the R and B signals digitally converted by the n-2 focus to the upper part by 2 bits.
By taking the difference from the Y signal, it can be easily obtained as an n-bit signal.

If the number of conversion bits for each of the R, G, and B signals is set to 8 bits, the bit precision will be improved more than necessary, and it is meaningless.

On the contrary, compared to the above case, the time margin of the A/D converter becomes smaller and the power consumption also becomes larger.

By reducing the number of conversion bits of the A/D converter,
This increases the time margin and significantly eases the accuracy requirements.

FIG. 9 shows R,
This is an example in which G and B signals are extracted.

The signals read directly from the image sensor 901 are R2G, B
Cy, rather than the signal, as shown in FIG.

Repeated signal e of W, Y, G, CW, Y, G,...
yl @ 902 and the signal 903 shifted by 2 pixels in the same repetition,
They are configured to be taken out at the same time. Therefore, the analog signal processing unit 904 performs calculations on the read analog signals 902 and 903 to obtain R and G signals.

This creates a B analog signal. That is, R,
The B signal is obtained as follows.

R=-((W-Cy)+(Y,-G) 1B=-(
(W-Ye) + (Cy-G) 1 Once R, G, B signals are obtained, A/D conversion is performed in the same way as in Fig. 8, and Y, R-Y, B- Y signal can be taken out.

In this way, the signal read directly from the solid-state image sensor is
Even if it is not an R, G, B signal, R, G can be calculated by analog calculation.
, B signals can be obtained, so efficient A/B signals can be obtained for them.
D conversion can be realized.

In FIG. 11, the signal from the solid-state image sensor is subjected to analog processing in front of the A/D converter, and Y, RY.

This is an example in which a BY signal is obtained. A/D converters 1102 and 1103 are for luminance signals, respectively.

This is for color signals. The solid-state image sensor 1101 has a filter having the same configuration as the conventional example shown in FIG.

The signal directly read out from the solid-state image sensor 11o1 is a mixed signal, but it can be considered to be the Y signal itself. This is because the proportion of color signals included in the mixed signal is at most about 25 pixels, which poses no practical problem. The color signal is extracted from the mixed signal by an analog color separation circuit 11.
R-Y by o4.

It is taken out as a B-Y signal. However, strictly speaking r 9
．． The b9 signal is treated as the R-Y and B-Y22 signals, respectively.

For these two luminance signals, the number of digital output bits is m and m-1 pinto (m is an integer of 2 or more)) (DA/D near 7 converter 1102, 1).
103, those digital signals can be obtained efficiently. For example, if 7 pins are required to A/D convert the Y signal, 6 bits is sufficient as the number of conversion pins for the RY and BY signals.

If color separation processing is performed after A/D conversion as in the conventional example shown in FIG. 4, the bit accuracy of the color signal will drop significantly. As mentioned above, there are at most two color signals included in the composite signal.
Since the number of pixels is about 6, if the color signal is extracted after A/D conversion of the mixed signal with m-bit precision, it can only be obtained with m-2 bit precision, making it impossible to reproduce faithful colors.

Therefore, by implementing the method shown in FIG. 11, the number of output bits can be reduced and the bit accuracy can be maintained.

Note that since the chrominance signal does not require as wide a frequency band as the luminance signal, it is also possible to lower the sampling clock frequency of the A/D converter for the chrominance signal.

Effects of the Invention As described above, the present invention performs digital signal processing by changing the number of digital output bits of the A/D converter provided for each type of analog signal from the solid-state image sensor. It is possible to use an A/D converter that can efficiently A/D convert each signal, has a large time margin, and has low power consumption. In addition, the number of digital output bits of the A/D converter can be slightly reduced, and the requirements for conversion speed and accuracy are eased, so the circuit system can be simplified and the number of parts reduced.

This is extremely useful in terms of cost reduction, facilitates the configuration when integrated into an IC, and is extremely advantageous in practice.

[Brief explanation of drawings]

Fig. 1 is a plan view showing the configuration of a color mosaic filter in a conventional example, Fig. 2 a and b are schematic diagrams of output signals from a solid-state imaging device in a conventional example, and Fig. 3 is an analog signal of a conventional solid-state imaging device. Processing block diagram; FIG. 4 is a block diagram when the same device is replaced with digital signal processing; FIG. 5 is a schematic diagram showing signal readout from a conventional solid-state image sensor;
FIG. 6 is a schematic diagram showing an embodiment of the present invention, FIG. 7 is a timing chart when FIG. 6 is operated, FIG. 8 is a schematic diagram of a digital signal processing section in an embodiment of the present invention, and FIG. FIG. 9 is a block diagram showing an embodiment of the present invention using a solid-state image sensor with simultaneous two-line readout, FIG. FIG. 1 is a block diagram showing an embodiment of the present invention using an image sensor. 601 - Signal transfer unit, 602, 603, 604...
...A/D converter, 6o5...digital signal processing unit, 6060-encoder, 607...D/
Ako7verta, 801, 8'02, 803...4
/D converter, 901 ... solid-state image sensor, 902
゜903...Analog signal read out from the solid-state image sensor, 904...-Analog signal processing unit, 1101...
・Solid-state image sensor, 1102, 1103...A/D converter, 11o4...Analog color separation circuit, 11o
5.--Digital synchronization circuit.

Claims (5)

[Claims] (1) Digital signal processing is performed by changing the number of digital output bits of the A/D converters provided for each type of analog signal, depending on the type of analog signal from the solid-state image sensor. A color solid-state imaging device. (2) Analog signals from the solid-state image sensor are R and G. The patent claim is characterized in that the number of digital output bits of the A/D converter for the R and B analog signals is smaller than the number of digital output bits of the A/D converter for the G analog signal. Range 1
The color solid-state imaging device described in . (3) The number of digital output bits of the A/D converter for B, G, and B analog signals is n-2 and n-2, respectively.
1, n-2 bits (n is an integer of 3 or more), add them to create a Y signal, shift the digital output of the R9B signal upwards by 2 bits each, and take the difference from the Y signal. 3. The color solid-state imaging device according to claim 2, wherein the color solid-state imaging device generates R-Y and B-Y signals. (4) Y by the signal read out from the solid-state image sensor. Create analog signals of R-Y and B-Y, and R-Y. 2. The color solid-state imaging device according to claim 1, wherein the number of digital output bits of the A/D converter for the BY signal is smaller than the number of digital output bits of the A/D converter for the Y signal. (5) A/D for Y signal, RY signal, B-Y signal
The number of digital output bits of the converter is m, respectively.
5. The color solid-state imaging device according to claim 4, wherein the number of bits is m-1, m-1 bits (m is an integer of 2 or more).