JPS60264170A - Solid-state image pickup device - Google Patents

Abstract

PURPOSE:To obtain a television signal with high resolution with an inexpensive solid-state image pickup element employed by shifting by a specific interval in horizontal and vertical directions respectively two solid-state image pickup elements arranged by a half of horizontal picture element interval in the horizontal direction two object images at each field of 4n+2, 4n+3, 4n+4 to 4n+1 field. CONSTITUTION:An object image of light obtained from an object is resolved into two by a resolving optical system 22 and the object image is formed respectively to a CCD I 23 and a CCDII24. The relation of position of the CCD I 23 and the CCDII24 is shifted by 2d in the horizontal direction. Further, vibrators 25, 26 move the position of the photodetection part of the two CCDs by horizontal picture element interval 4d/4 in the horizontal direction (i.e., d) and vertical picture element interval 2h/2 in vertical direction (i.e., h) in the (4n+2) field to the (4n+1) field, by 2d in the horizontal direction in the (4n+3) field and by 3d in the horizontal direction and (h) in the vertical direction in the (4n+4) field to provide picture element being 4 times in horizontal direction and times in vertical direction to the picture element number of one CCD.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention relates to a solid-state imaging device used in a color television camera in a high-definition television system.

Conventional configuration and its problems In recent years, the standard television system (NT8C) has been introduced.

For example, as a high-resolution, high-definition television system that replaces PAL, SECAM, etc.),
A high-definition television system is proposed for OL7NllL44P37-42 that compresses the band four times and transmits it.

The camera section of a conventional high-definition television system will be described below.

Figure 1 is a configuration diagram of the camera section of a conventional high-definition television system, where 1, 2, and 3 are image pickup tubes for red (R), green (G), and blue, respectively, and 4 is R. , G, and H video signals to obtain a luminance signal (■), a color difference signal (CW), and ([N), and 6 is a matrix circuit that converts the luminance signal to a frequency of t, /4 using a clock of 5 Hz. D converter, 6 is field offset sampling that samples each field so that it has an offset relationship in the horizontal direction, 7 is a still image filter that passes the band required for still images, and 8 is a video filter that passes the band required for moving images. Filter, 9 is frame memo! 171.10 is motion vector detection that detects vectors for camera pan and tilt movements;
11 is a frame memory whose readout time is controlled by a motion vector detection signal; 12 is a video detection device that detects moving parts between frames; 13 is a switch that switches between still image mode and video mode using a video detection signal; 14 is an in-field memory This is sub-sampling in which sampling is performed in an offset relationship for each horizontal scan.

The operation of the camera section of the high-definition television system configured as described above will be described below.

First, the signals obtained from the R, G, 'B image pickup tubes are synthesized in matrix 4, and a luminance signal (1) and a color difference signal (CW
) and (ON) are obtained. Next, the luminance signal (7) is subjected to AD conversion at a clock cycle t1/4 in an AD converter. The timing of this person transducer coincides with the spatial sampling points a, b, c, d, and e shown in FIG. In Figure 2,
◯ indicates the sampling point of the 4n+1st field, ◯ indicates the 4n+2nd field, . indicates the sampling point of the 4n1003rd field, and I indicates the sampling point of the 4n1004th field. That is, the spatial interval d corresponds to t1/4, and the sampling points are arranged horizontally and vertically in a square. Next, the field offset sampling 6 is used to sample spatial sampling points a, b, c, and d shown in FIG. That is, sampling is performed at a period t1/2 corresponding to a spatial interval of 2d, and the phase relationship between the sampling points of the odd and even horizontal scans is set to 180 degrees, so that the relationship between the sampling points of the so-called odd field and even field is an offset relationship. Make it.

Next, the video signal output from the field offset sampling 6 is band-limited by the still image filter 7 and the moving image filter 8, which are two-dimensional low-pass filters that pass only the shaded areas shown in FIGS. 3 and 4, respectively. On the other hand, the motion vector detector 10 compares the output of the filled offset sampling 6 with the current signal, which is obtained by delaying the output of the filled offset sampling 6 by one frame period in the frame memory 9, and the output of the filled offset sampling 6. Calculate vectors for pan and tilt movements.

Next, the output of the filled offset sampling 6 is stored in the frame memory 111, and when read out one frame later, it is moved by the vector amount calculated in the motion vector detection 1o.

That is, the readout time of the frame memory I111 is controlled so that the output of the frame memory i11 and the output of the filled offset sampling 6 are obtained at exactly the same timing with respect to panning and tilting movements of the camera section. The video detection unit 12 compares the output obtained with the field offset output and detects movement other than panning and tilting of the camera unit, such as zooming, or a partially moved portion within the screen. Next, depending on the signal detected by the moving image detection 12, the output of the still image filter and the output of the moving image filter 8 are switched by the switch 13, and input to the sub-sampling 14, 4n+
In the 1st field, sampling point a is sampled, in the 4n+2nd field, sampling point I is sampled, in the 4n+3rd field, sampling point Ci is sampled, and in the 4n+4th field, sampling point d is sampled. The period of this sampling is t, and the frequency is S W / 4 H
It becomes z.

Therefore, one screen is composed of four fields, and for the still image part (including panning and tilting of the camera), the four fields are two-dimensionally synthesized using frame memory on the receiving side, and sample points a, b.

By interpolating the signals at c and d to obtain the signal at sample point e, a video signal having the spatial frequency region indicated by diagonal lines in FIG. 3 can be reproduced on the monitor.

As there is no correlation between fields in the video part,
For example, in the case of the 4n+1th field, by interpolating the signal at sampling points to and e using sampling point a in that field, a video signal having the spatial frequency region shown in FIG. 4 can be reproduced on the monitor.

As a result of the above, even though it is transmitted to the receiving side at the sub-sampling frequency sW/4H2, in still images, the video signal in the horizontal frequency band SW/2H2 can be transmitted because the sampling positions between fields are offset. This is because there is a correlation between the legs and the fields, so aliasing interference can be removed.

Also, regarding videos, there is no correlation between fields, so
When transmitting a video signal in the horizontal frequency band SW/2H2, aliasing interference occurs, so the video signal band is SW/2H2.
It is necessary to keep it to 4H2.

Therefore, by sampling the video signal of the moving image filter 8 for a moving part in the screen and the still image filter 7 for a stationary part using the subsampling 14, an image with less aliasing can be obtained.

In a high-quality system, the sampling frequency SW in the mu D converter is approximately e4MHz, and approximately 1500 sampling points are required horizontally and approximately 1000 vertically within the effective screen. , If a solid-state image sensor for R, G, and B is used in place of the B image sensor tube, a solid-state image sensor with a pixel count of approximately 160 pixels horizontally and approximately 1000 pixels vertically is required, and the processing after the matrix is the same as that for the image sensor tube. The same video signal can be obtained with, for example, 2
/3 inch size solid-state image sensor can be used as a prize with about 400 horizontal pixels and 500 vertical pixels, and 1 horizontal pixel.
500. With the current technology, it is impossible to satisfy various performances with a vertical 1000 solid-state image sensor. Moreover, if it is manufactured in a size larger than 1 inch, it will be very expensive. Therefore, if we were to use a solid-state image sensor to configure the camera section of a high-definition television system, it would be impossible to use a 2/3-inch solid-state image sensor, and we would have to use a very expensive 1-inch or larger solid-state image sensor. The problem was that it had to be used.

Purpose of the Invention The present invention solves the above-mentioned problems of the conventional example, and aims to provide a solid-state imaging device for a high-definition television system using a solid-state imaging device that is very inexpensive and can be manufactured as a prize. .

Structure of the Invention The present invention provides a first image forming apparatus which is arranged on the imaging plane of two identical subject images.
a second solid-state image sensor; The relative positional relationship of the second solid-state image sensor is shifted between the first and second solid-state image sensors by a length corresponding to one half of the horizontal pixel interval in the horizontal direction of the solid-state image sensor; The relative positional relationship between the subject image and the first and second solid-state image sensing devices is determined as follows: for a 4n+1 field (n is an integer), for a 4n+2 field, one quarter of the horizontal pixel interval in the horizontal direction, and vertical in the vertical direction. 1/2 of the pixel interval
For n+3 fields, 1/2 of the horizontal pixel spacing in the horizontal direction.4 For n+1 fields, 3/4 of the horizontal pixel spacing in the horizontal direction, 3/4 of the vertical pixel spacing in the vertical direction, and 2/2 of the vertical pixel spacing in the vertical direction The first solid-state image sensor is constructed with a shift of one-fold. By performing pixel synthesis in the horizontal direction using the second solid-state image sensor and selecting four pixel positions of the solid-state image sensor relative to the subject image for each field, the number of pixels is four times as many in the horizontal direction and twice as many in the vertical direction. A solid-state imaging device having a resolution equivalent to that of a solid-state imaging device can be obtained.

DESCRIPTION OF EMBODIMENTS FIG. 6 is a block diagram of a solid-state imaging device according to a first embodiment of the present invention, and FIG. 6 shows a horizontal pixel interval of 4 used in the first embodiment.
d. First and second solid-state image sensors in which light-receiving parts are squarely arranged with a vertical pixel interval of 2h (here, a COD image sensor is used as the solid-state image sensor; hereinafter abbreviated as "can").

), Fig. 7 is the configuration diagram of Fig. 1. Figure 8 is a positional relationship diagram for the subject image between the second COD and the 10th COD for the subject image.
It is a positional relationship diagram of CCD. In Figure 7, Ank is n
The light receiving part of the first COD in the vertical row in the horizontal row, Bnk is n
The horizontal row shows the light receiving part of the second COD in the vertical row, and the eighth
In the figure, mnka is n in the 4n+1 field.
The light receiving part of the first COD in the vertical row in the horizontal row, mnkb is n in the 4n+2 field, the light receiving part of the first COD in the vertical row in the horizontal row, mnkc is the 4n+3th field p
λnkd indicates the light receiving portion of the first COD in the vertical row in the n horizontal row in the 4n+4th field.

In FIG. 6, 6 is a D converter, 7 is a still image filler, 8 is a moving image filler, 9 is a frame memory 1, 10 is a motion vector detector, 11 is a frame memory II, 12 is a moving image detector, and 13 is a moving image filter. Switch 14 is subsampling,
The above has the same configuration as that shown in Fig. 1, with 21 being a lens, 22 being a resolving optical system that creates two subject images, and 2
3.24 is the can shown in FIG. 6 which is arranged on the imaging plane of the two subject images, and the relative positional relationship between the subject image and the two CODs corresponding thereto is such that the horizontal pixel interval is the same in the horizontal direction. CGD I and acDm, 25 and 28, which are arranged 1/2 of 4d, that is, 2d apart, are CC1DI2
3 and CGDl[24 to 4n+174-old, 4n
+2 field, 1/4 of the horizontal pixel interval 4d, or d1, 1/2 of the vertical pixel interval 2h, 4n
27 is a transducer for shifting the output of CCDI23 by 2d horizontally in the +3 field, 3d1 horizontally in the 4n+4 field, and h in the vertical direction in the 4n+4 field.
Time t1/! corresponding to the deviation amount 2d of DI23! The delay line 1128 that delays CCDI23 and CGDl[2
29, 30.

31 is 4n+2.4n+3.4n+4 for 4n+1
A horizontal shift amount d in the relative positional relationship between the subject image and the CCDI 23 and C0D 1124 caused by the transducers 25 and 26 in the field.

2d, t1/4 corresponding to 3d. t1/2, 3t, /
Delay line 11 for delaying the time of 4. I, IV, and 32 are field selectors that select and output one of the four manual inputs for each field.

The operation of the solid-state imaging device of this embodiment configured as described above will be explained.

First, light obtained from the object is passed through the lens 21, and the same object image is separated into two by the separation optical system 22, and these object images are focused on the CCDI 23 and the 0CDn 24, respectively. The positional relationship between CCDI23 and C0D1 [24 is 2d apart in the horizontal direction as shown in FIG.
The outputs of and CG D If 24 are shown in Figure 9 a and b, respectively.
The signals shown in FIG. 9 and i are the signals with the timing shown in FIG. Mu12
CC shown in FIG. 7 corresponds to j in FIG. 9, respectively.
Light receiving section B11 of DI[24. Corresponds to B12.

Next, by delay line I27, C1C is
The output of DI24 is delayed to obtain the signal shown in Figure 9C, and then C
The outputs of the CDI 23 and the delay line 27 are combined to obtain the signal shown in FIG. 9d. In addition, as shown in FIG. 8 using the oscillators 25 and 26, the positions of the light receiving parts of the CCDI 23 and the CCDIr 24 are set to 4th n+1 + th n+2. 4thn+3. For the 4n11th field to move by the 4n+4 (n: n is an integer) field, the 4n+2 field is d in the horizontal direction, h1 in the vertical direction, the 4n+3 field is 2d in the horizontal direction, and the 4n+4 field is μ in the horizontal direction. In case of 3D, it is shifted by 3d in the horizontal direction and by h in the vertical direction.

Therefore, for the 4n+1 field, the 4n+2 field
In the field, the delay line lI29 causes the time t1/4 corresponding to the horizontal deviation amount d, and in the 4n13th field, the delay line 130 causes the time t1/2 corresponding to the horizontal deviation amount 2d, and in the 4n+4th field, the delay line IV
31, the time 3t corresponding to the horizontal deviation amount 3d
The amount of positional deviation in each field is corrected by delaying by 1/4, and the field selector 32 outputs the synthesis circuit 28 output in the 4n+ field, the delay line 1129 output in the 4n+2 field, and the delay line 130 in the 4n+3 field. The output, the 4n+4th field, selects the delay line IV31 output.

If the sampling points of the signal obtained by this field selector 32 are converted into two dimensions, they are two fields, that is, one frame, and are the sampling point a shown in FIG.

This corresponds to b, c, and d. If only one CCDI23 is used, as shown in Fig.
b l AnnCr Arranged at the positions of And, corresponding to Figure 2 a, c, b, d, and C
When using two cards, GDI23 and C0DII24, C
GDI23 light receiving section Ann & 1Annb r Ann
nnc tAnd for c tAnd, respectively.
Since the light-receiving part of OODM2a exists at the position of tAnnd IAnnb, the 4n+1 and 4n+3 fields receive light from Munna and Annc in Figure 8, and the 4n+2 and 4n+474- fields receive light from Munnb and Munnd in Figure 8. The outputs of the parts are continuously obtained by the feel route selector 32, and the mnna tom nnc in FIG. 8 corresponds to the sampling points a and b in FIG.
This is because they correspond to sampling points c and d in FIG.

Therefore, the output obtained by converting the output of the field selector 32 by the D converter 6 is the same in sampling timing as the signal obtained by the field offset sampling 6 in the conventional example shown in FIG. 6
The processing from output to subsampling 14 is the same as the processing from field offset sampling 6 to subsampling 14 in FIG. 1, and the same video signal is obtained.

As described above, according to this embodiment, two cans are arranged with respect to two subject images shifted by 1/2 of the horizontal pixel interval 4d of the COD, that is, 2d, and a 4n+2 field is arranged for a 4n+1 field. Then, in the horizontal direction, 1/4 of the horizontal pixel interval 4d, or d1, in the vertical direction, 1/2 of the vertical pixel interval 2h, or 4n+3 feed μ, 2 in the horizontal direction
In the d14n+4 field, the relative positions of the light receiving parts of the two CODs with respect to the subject image are moved by 3d in the horizontal direction and h in the vertical direction, and the number of pixels in one COD is 4 in the horizontal direction. ccn, which has twice as many pixels in the vertical direction, has fewer pixels, and is therefore cheaper
A high-resolution video signal can be obtained using this method, and it can be fully used as a brightness signal for a high-definition television signal. It is also clear that a color high-definition television signal can be obtained by providing a set of two CODs for each of R, G, and B as shown in this embodiment.

In the embodiment, the COD was moved using a vibrator, but the cap-4 may be fixed and the subject image may be moved.

Further, in the embodiment, a COD is used as the solid-state image sensor, but any other solid-state image sensor may be used as long as it can read data from all the light receiving sections to the transfer section within the vertical blanking period.

Further, in the embodiment, delay line I, 11 . I, IV, 27, 29, 30.31 are used, but the phase of the readout node of the solid-state image sensor is corrected by the time corresponding to the positional deviation of the field. Alternatively, delay lines I, n, II, IV and field selector 32 can be omitted. Furthermore, by inserting a spatial low-pass filter with an f-band, the spatial frequency of the subject image indicated by the diagonal lines in Figure 3, between the lens and the COD, the still image filter can be omitted and similar results can be obtained. Needless to say, it can be used.

Effects of the Invention The present invention uses two solid-state image sensors arranged horizontally shifted by half the horizontal pixel interval from each other for two subject images, for a 4n+1 field, and for a 4n+2 field. 1/4 of the interval, i.e. d, 1/2 of the vertical pixel interval in the vertical direction, 2d in the horizontal direction for 4n+3 fields, 3d in the horizontal direction and h in the vertical direction for 4n+4 fields μ. This makes it possible to obtain high-resolution television signals using a very inexpensive solid-state image sensor that can be easily manufactured. It is possible to realize an excellent solid-state imaging device that does not cause frame afterimages (results caused by reading from half of the light-receiving sections in odd and even fields) that occurs in interline type COD for reading.

[Brief explanation of the drawing]

Figure 1 is a configuration diagram of the camera section of a conventional high-definition television system, Figure 2 is a sampling arrangement diagram showing the sampling position of each part in Figure 1, and Figures 3 and 4 are one-dimensional spatial low-pass Characteristic diagram of the filter, Fig. 5 is a block diagram of a solid-state imaging device in an embodiment of the present invention, Fig. 6 is a block diagram of a COD image sensor used in the embodiment, and Fig. 7 is a diagram between two COD light receiving sections. FIG. 8 is a layout diagram of one COD light receiving section in 4n+1.4n+2.4n+3.4n+4 fields, and FIGS. 9a to 9d are waveform diagrams at various parts of FIG. 5 of the embodiment. . 22...Resolving optical system, 23.24...
C.O.D. 25.66... vibrator, 27, 29, 3Q,...
31...Delay line, 28...Synthesizing circuit,
32...Filled selector. Name of agent: Patent attorney Toshio Nakao and 1 other person 2nd
Figure 3 Figure 4 Horizontal Air Viewing Seimoku Light 0 Loto c> c3 2 - - 8 Satoshi Ishi

Claims (1)

[Claims] The first and second solid-state imaging devices are arranged on the imaging planes of two identical subject images, and the first and second solid-state imaging devices are arranged on the imaging planes of two identical subject images. The relative positional relationship of the light-receiving parts of the second solid-state image sensor is determined between the light-receiving parts of the first and second solid-state image sensors, which corresponds to one half of the horizontal pixel interval of the solid-state image sensor in the horizontal direction. By shifting the subject image by the length, the relative positional relationship between the subject image and the first and second solid-state image sensors is set to 1/4 of the horizontal pixel interval in the horizontal direction in the 4n+2 field compared to the 4n+1 field (n is an integer). 1/2 of the vertical pixel spacing in the vertical direction
, 1/2 of the horizontal pixel spacing in the 4n-4-3 field, 3/4 of the horizontal pixel spacing in the horizontal direction, and 1/2 of the vertical pixel spacing in the vertical direction for 4n fields.
A solid-state imaging device characterized by being shifted.