JPH0325992B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an apparatus for automatically performing vertical centering adjustment for registration of a color television camera.

The imaging device for color television uses multiple imaging elements with different spectral sensitivity characteristics.
For example, a system that is equipped with an image pickup tube and synthesizes video signals from multiple image pickup tubes to obtain a color video signal is widely adopted. It is necessary to accurately superimpose images using signals, so-called registration.

This registration includes size adjustment to match the scanning range of the optical image to be captured, correction of geometric distortion caused by variations in the imaging optical system and imaging element system, and centering adjustment to match the relative position of the scanning range. Among these, the centering adjustment method is
Conventionally, a manual method as shown in FIG. 1 has been mainly adopted.

In the figure, 1 is the G (green) channel imaging system, 2 is the R (red) channel imaging system, and 3 is the B channel imaging system.
(blue) channel imaging system, 4 is the viewfinder (or monitor), 5 is the operator's eye, 2v,
3v is a knob for centering adjustment, and the operator operates knobs 2v and 3v as necessary while looking at the image on the viewfinder 4 to adjust G.
Adjustments were made so that the center portions of the R and B images coincided with each other using the image as a reference.

Therefore, this method has the disadvantage that the operation is complicated and requires a considerably long adjustment time.

In particular, this kind of registration tends to shift when the usage conditions of the television camera change, and as a result, readjustment is often required for portable cameras etc., so the manual method described above is not sufficient. It was not possible to make any adjustments, and it was not possible to obtain a video signal of excellent image quality.

In order to solve this problem, a so-called auto-centering device, which automatically adjusts centering when necessary, has been proposed and has become widely used.

An example of such an autocentering device is shown in FIG.

In the figure, 1 to 3 are the same imaging systems for each channel as in Figure 1, 6 and 7 are preprocessing circuits, 8 and 9 are binarization circuits, 10 is an exclusive OR circuit, and 1
1 and 12 are first and second edge signal generation circuits;
3 is an arithmetic circuit, 14 is a control circuit, 15 and 16 are centering correction devices, and 17 and 17' are interlocked changeover switches.

The preprocessing circuits 6 and 7 are circuits that perform signal processing such as nonlinear amplification and signal clamping (to ensure binarization).

Binarization circuits 8 and 9 output signals of "0" and "1" depending on the signal level.

Exclusive OR circuit 10 is preprocessing circuit 6
A binarized signal G obtained by processing the video signal of the G channel from the preprocessing circuit 7 in the binarization circuit 8 and a binarized signal RorB obtained by processing the video signal from the R or B channel from the preprocessing circuit 7 in the binarization circuit 9. Exclusive OR signal E is generated by extracting only the portion where the level difference between

The edge signal generation circuits 11 and 12 are circuits that extract the edge portions of the binary video signals G and RorB (hereinafter RorB will be referred to as It is configured.

The arithmetic circuit 13 receives the exclusive OR signal E and the edge signals Ge, Xe, performs the following arithmetic operations, and generates two types of discrimination signals Y, Z. For example, it is a circuit that uses four AND circuits and one It consists of an inverter and two OR circuits.

Y=E・Ge+・Xe Z=E・Xe+・Ge or Y=E・Ge Z=E・Xe or Y=E・Ge・e+・e・Xe Z=E・e・Xe+・Ge・e In these arithmetic expressions, the signals with a bar above the characters such as eg represent inverted signals, and therefore the above signal represents the signal with the polarity of signal E inverted. .

The control circuit 14 operates a centering correction device 15 or 16 in response to a signal representing a centering deviation.
This circuit operates to send a control signal to the imaging system 2 or 3 to align the center of the scanning range of the imaging system 2 or 3 with the imaging system 1 of the G channel, and may be configured by a microcomputer or the like.

The switch 17 is a switch for switching the centering adjustment between the R channel and the B channel.

The switch 17' is used to switch control signals for the R channel and B channel centering correction devices 15 and 16 in conjunction with the switching operation of the switch 17.

Next, the operation of this conventional example will be explained with reference to timing charts shown in FIGS. 3a to 3c.

First, switch 17, 17' is set to R as shown in the figure.
Assume that the channel has been switched.

Then, the video signal from the R channel imaging system 2 is binarized and appears at the output of the binarization circuit 9 as shown in FIG. 3 (signal R), which is processed by the edge signal generation circuit 12. , R channel edge signal Re appears at its output as shown in FIG.

On the other hand, regardless of the switching state of the switches 17 and 17', the output of the binarization circuit 8 is a signal G obtained by binarizing the video signal from the imaging system 1 of the G channel, as shown in FIG. It is processed by the edge signal generation circuit 11, and the edge signal Ge of the G channel appears at its output as shown in FIG.

At the same time, signals G and R, which are the outputs of these binarization circuits 8 and 9, which are the binarized video signals of the G channel and R channel, respectively, are also input to the exclusive OR circuit 10, and the signals G and R are input to the exclusive OR circuit 10. An output signal, that is, an exclusive OR signal E, which is generated only when there is a difference in level of a predetermined value or more, is extracted as shown in FIG. 3.

Therefore, these edge signals Ge, Re and exclusive OR signal E are supplied to the arithmetic circuit 13, where the above-mentioned logical arithmetic processing is performed and the two types of discrimination signals Y and Z are generated as shown in FIGS. 3a to 3c. It is taken out and supplied to the control circuit 14. The discrimination signals Y and Z at this time, as described later, represent the direction of centering deviation of the R channel with respect to the G channel, so the control circuit 14
generates a control signal in response to the discrimination signals Y and Z, and controls the centering correction device 15 via the switch 17'.
The deflection center of the imaging system 2 of the R channel is changed to match the G channel, and the centering deviation is automatically corrected.

After completing the centering adjustment of the R channel with respect to the G channel, it is time to turn switches 17 and 1.
7' is switched in the opposite direction to that shown in FIG. 2, that is, to the B channel, the centering adjustment of the B channel with respect to the G channel is performed in the same manner as in the case of the R channel described above, and the centering is completed.

Next, the operation of determining the centering deviation using the determination signals Y and Z will be explained in more detail according to the state of the centering deviation.

1 to 6 in Figs. 3a to 3c are cases where the compared channel, for example, the R channel, which should be centered with respect to the reference channel, that is, the G channel, is out of alignment.
1 and 2 of c indicate when the video signal levels of the G channel and R channel are almost equal, and the pulse widths of the respective binarized signals G and R are equal, and 3 and 4 are If the level of the R channel video signal is lower than the G channel video signal, and the binarized signal G has a wider pulse than the R channel video signal, then in 5 and 6, the R channel video signal is higher than the G channel video signal. The level is higher than the video signal of , and the pulse width of the binarized signal R is wider than that of G. In either case, the number of pulses Yn of the discrimination signal Y and the pulse width of the discrimination signal Z are The number of pulses Zn is not equal, and when the R channel shifts to the right on the monitor image plane, that is, at 1, 3, and 5 in Figure 3 a to c, Yn > Zn, and conversely shifts to the left. Sometimes Yn<Zn, and as a result, the direction of the registration shift can be determined from the relationship in magnitude between the number of pulses of the determination signals Y and Z.

Next, 7 and 8 in Fig. 3 a to c are cases where the R channel matches the G channel, and 7 is when the level of the video signal of the G channel is higher than that of the R channel. On the contrary, it shows the case when the level of the video signal of the R channel is higher than that of the G channel, but in any case, the number of pulses Yn and Zn appearing as the discrimination signals Y and Z is Yn = Zn. becomes.

Therefore, the control circuit 14 inputs the discrimination signals Y and Z from the arithmetic circuit 13, compares the number of pulses Yn and Zn, and outputs opposite control signals when Yn>Zn and when Yn<Zn. is generated and the signal is sent to the centering correction device 15 via the switch 17'.
supply to. Then, by changing the scanning range of the imaging system 2 of the R channel, when Yn>Zn, the signal R is controlled to move to the left, and conversely, when Yn<Zn, the signal R is moved to the right. If control is performed so that Yn=Zn, R for the G channel will be automatically adjusted.
Channel centering adjustment can be performed.

In addition, in 7 and 8 of this Figure 3 a to c,
Only the case where there is a level difference between the video signals of the G channel and the R channel and therefore the pulse widths of the signals G and R are different is shown. If the pulse widths of are equal, the exclusive OR signal E will be "0", so
The edge signals Ge and Re are taken out from the arithmetic circuit 21 as they are as the discrimination signals Y and Z, and therefore, it goes without saying that the discrimination condition of Yn=Zn, which represents the coincidence of centering, is maintained even in this case.

Once the centering adjustment of the R channel with respect to the G channel has been completed, next switch 17, 17' to the B channel side, and the G
Centering adjustment of the B channel with respect to the channel is automatically performed, and the centering adjustment is completed.

Although only a part of the video signal of each channel is shown in FIGS. 3a to 3c, in reality, the signal in a predetermined range including the center of the image among the video signals from each imaging system 1 to 3 is shown. It suffices to extract only this to determine centering.

Further, at this time, it goes without saying that the video signals from each of the imaging systems 1 to 3 may be from any subject, but it is most desirable that the video signals be from a subject that contains a large amount of white and black.

By the way, in this conventional example, FIG.
As shown in 9 and 10 of c, G channel (reference channel) and R channel (compared channel)
11 and 11 in Figures 3a to 3c, when the video signals of
As shown in 12, even when one of the video signals is not obtained, the number of pulses Yn and Zn obtained in the discrimination signals Y and Z become equal, making discrimination impossible, but as already explained, This auto-centering method is applied to correct centering deviations due to differences in operating conditions or aging of color television imaging devices, so as shown in 9 and 10 in Figure 3 a to c. It is highly unlikely that the video signals of each channel would have such a shift that there would be no overlap, and unless the subject to be imaged has a very special color tone (for example, red, green, or blue). In the case of a monochromatic object), it is highly unlikely that the video signal of any channel will be missing, as shown in 11 and 12 in Figure 3 a to c, and therefore, in practice, it is impossible to distinguish in any case. There is little risk of this happening.

In addition, as shown at 11 and 12 in Figure 3b, if the second logic Y=E・Ge, Z=E・Xe is applied when the video signal of any channel is deleted, an error occurs. A judgment signal appears, but if you choose an object with a sharp black and white contrast as the subject,
Cases like 11 and 12 can be excluded,
There is no problem in practical use.

Note that when a microcomputer is used as the control circuit 14, the above operations can be processed by a program of the microcomputer.

Therefore, according to the auto-centering device shown in Fig. 2, centering can be easily adjusted in a short time as and when necessary, and excellent image quality with always accurate registration can be achieved. You can get a color television signal.

However, with this conventional autocentering device, although it is expected that the horizontal centering of the image quality will be sufficiently accurate,
The drawback was that vertical centering was not very accurate.

The reason for this will be explained below.

In order to perform vertical centering with the conventional device described above, the video signal in a predetermined range at the center of the image is sampled along an extraction line assumed in the vertical direction. We must try to obtain G and X. However, since most television signals are field interlaced,
The number of scanning lines for each field is extremely small, being one integer fraction of one frame.

As a result, in the above-mentioned conventional apparatus, the centering operation is performed in each field where the scanning line spacing is considerably wide, making it impossible to obtain sufficient accuracy.

SUMMARY OF THE INVENTION An object of the present invention is to provide an auto-centering device which eliminates the above-mentioned drawbacks of the prior art and allows vertical centering to be performed with sufficient accuracy even in a field interlaced television image pickup device.

In order to achieve this object, the present invention uses a memory capable of storing video signals for multiple fields, and extracts signals necessary for vertical centering from a video signal for one frame. It is characterized by the following points.

Embodiments of the autocentering device according to the present invention will be described below with reference to the drawings.

FIG. 4 shows an embodiment of the present invention, in which the same or equivalent parts as in the conventional example shown in FIG. 2 are given the same reference numerals.
A detailed explanation has been omitted.

In FIG. 4, 18 and 19 are random access memories (hereinafter referred to as RAM), 20 is a memory control circuit, and the rest is the same as the conventional example shown in FIG.

The RAMs 18 and 19 function to store the binarized video signal over one frame, that is, over two fields, so that it can be read out at will.

The memory control circuit 20 controls the RAMs 18 and 19 to sequentially sample binarized video signals from a predetermined range of horizontal scanning lines including the central portion of the image in correspondence with a predetermined vertical centering extraction line.
Two fields are written to RAM18 and 19, and then the written binary video signal is interpolated with the second field signal into the first field signal in order according to the vertical centering extraction line. It serves to read out signals G and X in this way.

As a result, the signals G and X read out from the RAMs 18 and 19 respectively are video signals for one frame in which the horizontal scanning line of the second field is interpolated between the horizontal scanning lines of the first field and the fields are interlaced. becomes the same as

This will be explained with reference to FIG.

In Fig. 5, A represents the state in the first field, for example an odd field, B represents the state in the second field, for example an even field, C represents the state in which one frame is completed after field interlacing, and A represents the state in the second field, for example an even field. A predetermined range set in the center of the image, l is at least one extraction line set in the vertical direction, l _H1 is a horizontal scanning line in an odd field, and l _H2 is a horizontal scanning line in an even field. are shown respectively.

Note that the number of extraction lines l at this time may be arbitrary, but increasing this number can improve the centering accuracy accordingly, but on the other hand, RAM1
Since a large capacity of 8 and 19 is required, a predetermined number may be determined taking these into account.

Therefore, in the RAMs 18 and 19, the binary video signal indicated by the black circle in Figure 5A is written to a predetermined address during the odd field period, and the binary video signal indicated by the white circle in Figure 5B is written during the even field period. The binarized video signal of that portion is written to a predetermined address.

Next, as shown in FIG. 5C, from the RAMs 18 and 19, the signals in the odd fields indicated by black circles and the signals in the even fields indicated by white circles are alternately transmitted along the vertical extraction line l, that is, on one side. Sequentially from each address, with the signal of the other field interpolated to the signal of the field of
It is extracted as binary video signals G and X.

Therefore, as is clear from this Figure 5 (c),
The signals G and X read from RAM18 and 19 are
It is extracted from a video signal with one frame worth of scanning lines.

Thereafter, these signals G and X are sent to an exclusive OR circuit 10 and edge signal generation circuits 11 and 12.
The centering operation is performed in the same way as in the conventional example shown in FIG.

As a result, according to the above embodiment, the binarized video signal for vertical centering is extracted from the video signal for one frame, and it is possible to prevent a decrease in centering accuracy due to an increase in the scanning line interval. can.

In the above embodiment, two fields' worth of binary video signals are written in the RAMs 18 and 19, and the signals are interpolated by reading them out alternately for each field to generate one frame's worth of signals. When I finish writing the binary video signal of one field, I write the 2nd field of the next field.
As the digitized video signal, the signals directly obtained from the binarization circuits 8 and 9 may be used, and the signals read from the RAMs 18 and 19 may be taken out alternately. In this way, the storage capacity of the RAMs 18 and 19 can be reduced.

As explained above, according to the present invention, the signals necessary for vertical centering can be obtained with sufficient density, and in addition, two types of video signals used for registration determination can be binarized. Because unique arithmetic processing is applied to the video signal after processing, changes in the video signal (contours) can be extracted faithfully and accurately, and the discrimination signal can be extracted without being affected by video signal level changes or waveform dullness. Y and Z can be reliably extracted, and as a result, not only can registration deviations be detected from the magnitude relationship of their pulse numbers, but also the direction of the deviation can be reliably determined. According to the present invention, there is provided an auto-centering device such as a color television device capable of obtaining a color television signal of excellent image quality in which the registration is always maintained accurately, while eliminating the drawbacks of the prior art. be able to.

[Brief explanation of drawings]

FIG. 1 is an explanatory diagram showing an example of a centering adjustment method in a conventional color television camera.
Fig. 2 is a block diagram showing a conventional example of an autocentering device, Figs. 3a, b, and c are timing charts for explaining its operation, and Fig. 4 is a block diagram showing an embodiment of an autocentering device according to the present invention. , FIGS. 5A to 5C are explanatory diagrams showing the operation. 1 to 3...imaging system, 18,19...random
Access memory (RAM), 20... memory control circuit.

Claims (1)

[Scope of Claims] 1. In a television imaging device that is equipped with a plurality of image sensors and obtains a field-interlaced color television signal by combining video signals of a plurality of channels, the two channels to be centered are It is possible to binarize each video signal of a portion along at least one predetermined vertical extraction line within a predetermined range of the corresponding image, and extract and store these binarized video signals over at least one field. a memory, and at least two image signals stored in the memory.
a memory control means for reading each of the two channels of signals independently while interpolating between fields; and a level difference between the video signals G and X read from the memory by the memory control means to a predetermined value. an exclusive OR logic circuit that generates a level difference signal E that exhibits a logical value of 1 at a certain portion; first and second edge signals Ge that exhibit a logical value of 1 at respective edge portions of the video signals G and X;
first and second edge extraction circuits that generate Xe;
Using these level difference signal E and the first and second edge signals Ge, Xe as input, the following calculation characteristics are obtained.
, that is, Y=E・Ge+・Xe Z=E・Xe+・Ge Y=E・Ge Z=E・Xe Y=E・Ge・+・Xe Z=E・・Xe+・Ge・1 and generates the discrimination signals Y and Z.
An auto-centering device characterized in that it is configured to determine the direction of vertical centering deviation between the video signals of the two channels based on the difference between the number of pulses Yn and Zn.