JPS6149564A - Solid-state image pickup device - Google Patents

Abstract

PURPOSE:To obtain color video signals of high sensitivity and smooth picture quality by mixing and reading out every two picture element signals different in position components from each other in the horizontal scanning direction. CONSTITUTION:The signal electric charge which is stored in each photo diode in accordance with an incident light is read out to a corresponding vertical transfer V-CCD2 through a MOS switch 6 in a field A when a vertical transfer pulse phiA goes to the high level in a vertical blanking period. For example, signals loads of photosensitive elements P11 and P22 are mixed and stored in a gate electrode A1. Signal loads are stored in electrodes A1, A2...A'1, A'2... similarly, and these signal loads are transferred to a horizontal shift register H- CCD3 in parallel in the cycle of one horizontal period and are taken out as the video signal from an output amplifier. In a field B, the similar operation is performed. By this mixing and reading out signal loads, video signals of high sensitivity and smooth picture quality are outputted.

Description

[Detailed Description of the Invention] Industrial Application Field The present invention constitutes a color camera by combining color filters arranged in a mosaic pattern containing various color components such as blue, green, and red, and one solid-state image sensor. The present invention relates to a solid-state imaging device for a so-called single-chip color camera. - Conventional configurations and their problems Solid-state image sensors have been actively developed because they have many advantages over image pickup tubes. Among them, interline transfer method CAN (hereinafter abbreviated as IT-ccn)
is considered particularly promising because it has low noise in principle.

A description will be given below using FIG. 1 showing the overall configuration of a conventional IT-ccn.

In this configuration, the signal charges accumulated in each of the photoelectric conversion elements 1 (hereinafter abbreviated as pixels) are transferred to the vertical transfer CCD2 (hereinafter abbreviated as v-ccD) as shown by the arrow, and then further transferred to the horizontal transfer CCD3. (hereinafter abbreviated as H-ccD), and sequentially transferred to the charge detection section 4 for detection and output. This is the basic operation of T-COD. In the IT-COD in which pixels are arranged in a square matrix in this manner, the horizontal resolution of the solid-state image sensor is determined by the number of pixels in the horizontal direction. For example, in a currently commonly used device with a horizontal pixel count of about 380, the horizontal resolution is about 28 OTV lines, and with the current integrated circuit technology, increasing the horizontal pixel count unnecessarily is disadvantageous in terms of yield.

On the other hand, due to the configuration of IT-COD, it is easy to arrange pixels closely in the vertical direction, but in the horizontal direction, it is necessary to insert a V-GCD with a large occupying area between each pixel, and the incident light Since light becomes a false signal (hereinafter abbreviated as smear) for the image sensor, it is necessary to completely block the light, and the aperture 6 of each pixel for the incident light is substantially small as shown by the broken line. However, photosensitive elements in the horizontal direction can only be made intermittently. For this reason, when capturing images using conventional IT-CC;D, problems such as discontinuity in image quality (appearing as a jagged outline) and moiré phenomena for images containing oblique components have become a major problem. There is.

Next, focusing on the scanning method, we will first explain the detection of black and white image signals.

As is well known, the color television image transmission and reception system in Japan and the United States uses the NTSG system9, and the interlace scanning system is the standard method for measuring and depicting image information. In order to reproduce images on television receivers in ordinary homes, even solid-state imaging devices must comply with these methods.

FIG. 2 is a diagram for explaining interlace scanning,
In the case where, for example, characters are imaged on the imaging surface of a solid-state image sensor and an image signal is extracted, the photosensitive pixel row P1+Pz+P
M... indicates a state in which photoelectric conversion is performed according to each scan from top to bottom and from left to right, and a video signal is detected corresponding to the image.

In the same figure, when normal interlaced scanning is performed, initially + Pl + P5 + p5 + ... PM
After the A field is interlaced scanned in this order, Pz and P4 are scanned to fill in the middle.
Scanning of the B field is completed in the order of °Pa, . . . PM. That is, a complete image A, that is, the entire screen (one frame) is constructed by scanning the A field and the B field. Thus, in the NTSC system, interlaced scanning is used to create a television image, which consists of j0 frames per second, or the sum of so[il] A-field scans and 30 B-field scans per second. In this method, a signal is detected from each light-receiving element once every π seconds, and such signal readout is called frame readout.

On the other hand, a scanning method called pseudo-interlace scanning is used in the field of imaging for black-and-white images of solid-state imaging devices. In this case, as shown in Figure 2, initially (P1 + Pz
) − (Ps, Pa) iPs + Pa), −・
- Simultaneous scanning is performed by adjusting the various workmanship, and the signals obtained through each signal line are not mixed in an external circuit, and after A field scanning is performed, then (Pz
+P3), (Pa, Ps), (Pa, Pz), (Pa, Pz), etc. By performing simultaneous reading with vertically shifted one line combinations, B field scanning is performed and one frame scanning is completed. It is a method. In this case, signals are detected for all light-receiving elements once every field, that is, once every π seconds, and such signal reading is called field reading.

In particular, in solid-state image sensors, the actual light-receiving element is often a photodiode integrated on a semiconductor Si substrate, and based on the following principles (1) to (3), field reading is not possible even if the vertical resolution is sacrificed to some extent. Interlaced scanning is desired.

(1) In the current semiconductor integrated circuit technology, it is difficult to make completely equivalent photodiode groups that correspond to frame readout A feed/read and B field readout. On the other hand, in field readout, all the photodiodes are read out field by field, so the flicker development described above can be improved.

(2) In field readout, each light receiving element is 1/6
Since it is read out every 0 seconds, the apparent afterimage is preferably shorter than frame readout.

(3) The maximum signal amount that can be handled in field readout for each scan within one frame period is two times each light receiving element, which is twice the dynamic range compared to one frame readout. can get.

Next, FIG. 3 shows an example of a conventional pixel arrangement for performing the above-mentioned pseudo-interlaced scanning, in which photosensitive pixels (represented by O in the figure, such as photodiodes, phototransistors, etc.) are arranged in a square. is P + + P 2'+ in Figure 2
(11) corresponding to each line of P3 +...
,12,13,...),(21,22゜23,
...) (31, 32, 33, ...) at regular intervals, and the horizontal intervals are Ll, L2, . . . corresponding to each pixel column in FIG. It is lined up with L5...

When field reading is performed using the method described above, for example, in field A, pixel (11, 21).

Each set of signals (12, 22), (13, 23), etc. is read out at the same time, Ll, L2 . L3...
... is guided to the v-ccD corresponding to each of the following. For the image sensor, this is the center of gravity A1 of each combination of pixels.
+ A2 + A3 . . . operate as if there were valid pixels, and in the vertical direction, averaging between pixels is performed, resulting in smooth image quality.

In the end, in the A field, each line of A, A', A", etc. is horizontally scanned sequentially, and in the B field, B, B', B
The field readout features (1), (2),
(3) is maintained.

However, the resolution in this case is heavy+[r'A+, A2
, 15..., horizontally L 1 + L2
+L5...at intervals, and vertically A,
The intervals are determined by B, A', B', A''...

Generally, in solid-state image sensors made of Si integrated circuits, there is a limit to the number of pixels that can be integrated into a unit area, and it is necessary to achieve the highest image quality with a limited number of pixels. In the above method, the characteristics of field readout can be obtained without deteriorating the resolution in both the horizontal and vertical directions, but the deterioration in image quality due to pixel discontinuity in the horizontal direction cannot be improved.

Purpose of the Invention In view of the above problems, the purpose of the present invention is to
In the T-CCD image sensor, we created an element configuration that can simultaneously average pixel signals not only vertically but also horizontally, and achieved smooth image quality with less moiré while maintaining conventional resolution. Furthermore, while maintaining the characteristics of field readout, high-quality color images can be obtained, and I
The aim is to create a color camera that takes advantage of the high sensitivity characteristic of T-COD.

The combination of two pixels in the diagonal direction makes it possible to average pixel signals in the horizontal and vertical directions.

Therefore, if this configuration can be implemented in IT-CG'D, the spacing between pixel centroids will remain the same as in the conventional case, while maintaining the conventional resolution and taking advantage of the characteristics of field readout. Pixel signals having adjacent components in the direction are mixed and averaged by reading out, and the discontinuous characteristic of image quality that tends to occur in a square arrangement of pixels can be alleviated, and the image quality can be further improved. Even if the aperture ratio of the photosensitive section is reduced to reduce blur, the smoothness of the image quality in the horizontal direction can be maintained sufficiently.

DESCRIPTION OF THE EMBODIMENT FIG. 4 is an example of pixel arrangement for pseudo-interlaced scanning in an embodiment of the present invention, and the A field reading is (A, A
', A" ...-...), but in the B field (B
, B', B"...) are horizontally scanned in this order.

More specifically, for example, horizontal scanning A consists of pixels (11,
22), (12, 23), (13, 24)...
The signals of each set are mixed and read out, and horizontal scanning A' is for pixels (31, 42) and (32, 43).

(33, 44)..., but on the other hand B, the pixel (21
゜32), (22,33), (23,34)...
・But in B', pixels (41, 52), (42, 53)
.

(43, 54)... are similarly configured to be read out in a mixed manner.

Next, FIG. 5 shows a part of the solid-state imaging device of the present invention.
This is an example of a T-I CD embodying the configuration shown in FIG. 4. Prior to colorization, the basic electrical operation and characteristics of this solid-state image sensor will first be explained in detail.

In the same figure, the imaging section is a photosensitive element 1 (P++) consisting of a photodiode as shown in the example of P41.

B12. B13...) and 4 prisons rI/(Rusφ
Column 2 of v-ccD driven by person, φB, φC9φD, φm.

φBK Fig. 6 MOS switch 6 (GAl, GA2...
,GBl,'GB2. ...), and further H-CO
D..., 3, a charge detection section 4,
Gate electrode A 1+ A forming V −CCiD2
1'...and B1. B1'...is φ person,
It has the ability to store signal charges under each gate electrode by applying φB, and the charges under each gate electrode can be transferred toward the H-CCD 3 by driving the sequential application of four-phase pulses. Here, explanation will be given using four-phase pulses in order to increase the maximum charge transfer amount of the v-CCD 2, but of course a two-phase type that is common in CCDs may also be used.

The clock pulses φ and φ8 shown in FIG. 6 are used to drive the MOS switch 6 and V-I CD 2.
It has three levels of vH, and φG and φD have two levels of Vt, , and vM that drive the V-CCD.

Regarding the basic operation, the signal charges accumulated in each photodiode according to the incident light are transferred to the vertical transfer pulse φm during the vertical planking period in the A field, and when the (B field TECH) becomes high level vH, the MOS switch 6
The data is read out to the corresponding V-CCD 2 through.

That is, in exact correspondence with FIG. 4, for example, the gate electrode A
1, the signal charges of Plj and B22 are mixed and accumulated.

Other A,, A2...A1', A2'...
. . . is exactly the same, and these signal charges are transferred to the horizontal shift registers H-G-D in parallel with the cycle of one horizontal period.
3 and taken out as a video signal from the output amplifier. This kind of signal flow is normal IT-(Ic
: Same as D sensor. This is exactly the same in the B field, and even if the basic configuration shown in FIG. 4 is configured with IT-ccD, the characteristics can be improved as described in the previous section.

The above is the basic operation of the device according to the present invention, and the video signal obtained here can only reproduce black and white images. A major feature of the present invention is that when using an image sensor having a structure capable of mixed readout as in this embodiment, by associating a mosaic color filter with each of the light receiving elements, it maintains the advantages obtained with field readout. It is possible to obtain a smooth color video signal with high sensitivity and image quality relatively easily.

Fig. 7 shows an embodiment of a mosaic color filter consisting of various color filters attached with similar numbers to correspond to each photo diode number in Fig. 4 to obtain a color signal. Filter each part with (B)
, green (G), and red (R).Ye in the figure means a filter that transmits yellow, that is, the G-1-H component, and ay means cyan and G4-B.
The component consists of four color elements: Mg is magenta, transmits B and R components, and G (transmits green light). Therefore, if this filter is attached to the element shown in Fig. 5, A,
B In each field, (Mg+07) and (
G-1-Ye) in the horizontal direction, and the n-1th
In the -1 line, (G + Cy) and (Mg + Ye) are repeated 9 times, and at this time, the luminance signal output of each line is
n, Yn++, Yn-(Mg+C:Y)+(G×Ye)=2R+3G+
2BYn++ = (G + Gy) + (Mg + Ye) = 2R
+sG+2B Still, each color difference component is On, O
If n++, On = (Mg+cy)-(G+Y
e )=2B −GOn++ = (Mg−1−Ye
) −(G+Cy) = 2R−G but B, G, R
are the respective color component outputs. Once the luminance signal and two types of color difference signals are obtained in this way, signal processing is performed in exactly the same way as in the case of single-tube color image pickup tubes, which are currently available in many commercially available tubes combined with stripe filters. In particular, a shicolor video signal can be obtained.

The experimental results show that the horizontal level is 400. Using the image sensor shown in Fig. 5 with 500 vertical light-receiving elements and the filter shown in Fig. 7, it is extremely clear with a TV resolution of 240 $ horizontally and 350 vertical lines, with no jagged edges and almost no moiré. I got no color image. At the same time, the advantages associated with the above-mentioned pseudo-interlaced scanning can be fully utilized.

Effects of the Invention As described above, by using the IT-I CD according to the present invention, each pixel can be scanned while averaging each pixel in the horizontal and vertical directions with a relatively simple structure, and the smear phenomenon can be suppressed to produce smooth heights with less jaggedness. It is possible to obtain high-quality color images, and also enjoy the field readout feature that is advantageous for solid-state imaging. Furthermore, when considering future solid-state cameras, since the present invention can average pixels within the element, there is no need to go to the trouble of making corrections in an external circuit, and the circuit system as a whole can be made smaller. are doing.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram of an IT-I CD as a conventional example.
Fig. 2 is a diagram for explaining interlaced scanning and ω pseudo-interlaced scanning; Fig. 3. Fig. 4 is a diagram of a conventional example and an improved example to explain the relationship between pixel arrangement and resolution for mixed readout, and Fig. 5 is a first implementation of the present invention that embodies Fig. 4 with IT-COD. FIG. 6 is an example of the main pulses necessary for driving the solid-state imaging device in this example, and FIG. 7 is a configuration diagram of a mosaic color filter. 1...Photoelectric conversion element , 2. River... Vertical transfer CO
D, 3...Horizontal transfer COD, 4...Charge detection section, 5...Light frontage section, 6...M
OS switch, PI j + PI 2, etc.
・Photodiode, φM, φB, φG, φD...
・・Main driving pulse, U + U' * BlHB+'
etc. Gate electrode, GAl, GA2. G.B.
l, GB26, etc...Mixed read MOS switch. Name of agent: Patent attorney Toshio Nakao (1st person)
Figure 2 p. Figure 3 Figure 4 LI L? L3L4Ls C♂ ta α

Claims (4)

[Claims] (1) In a charge-coupled (CCD) type solid-state image sensor, during horizontal scanning, each image sensor has two different positional components along the scanning direction.
A solid-state imaging device characterized by a structure that reads out individual pixel signals while mixing them. (2) The solid-state imaging device according to claim 1, wherein pixel signals of two pixels adjacent to each other with a vertical charge transfer means in between are mixed. (3) The solid-state imaging device according to claim 1, wherein pixel signals are read out by different combinations of pixels for mixed readout in the A field and the B field. (4) The solid-state imaging device according to claim 1, characterized in that it is formed by a combination of mosaic color filters made up of four types of color filter elements.