JPH0325993B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a system for automatically performing registration adjustment for 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 is performed by adjusting the size to match the scanning range of the optical image to be captured, correcting geometric distortion caused by variations in the imaging optical system and image sensor system, and centering adjustment to match the relative position of the scanning range. However, as a registration adjustment method, a manual method as shown in FIG. 1 has conventionally 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 a view finder (or monitor), 2v, 3v are knobs for adjusting the registration, and the operator can adjust the knobs 2v, 2v, and 3v as needed while looking at the image on the view finder 4. 3v was used to adjust the R and B images to match, using the G 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 type of registration tends to shift when the usage conditions of the television camera change, and in particular, large shifts occur due to changes in ambient temperature, requiring readjustment frequently during use. The manual method used did not allow sufficient adjustment, and it was not possible to obtain a video signal of excellent image quality.

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

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

In the figure, 1 to 3 are the same imaging systems for each channel as in Figure 1, 5 is a test pattern for registration, 6 and 7 are preprocessing circuits, 8 and 9 are binarization circuits, and 10 and 11 are intermediate 1 is a processing circuit, 12 is a pulse comparator, 13 is a post-processing circuit, 14 is a control circuit, 1
5 and 16 are registration correction devices, and 17 is a changeover switch.

The test pattern 5 is included in the imaging optical system for the three imaging systems 1 to 3, and the pattern is imaged on the imaging surface of each of the imaging systems 1 to 3 only during registration adjustment.

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.

Intermediate processing circuits 10 and 11 are circuits that perform arithmetic processing necessary for pulse comparison.

The pulse comparator 12 compares the pulses between the input channels, detects the pulse interval, and outputs a signal representing the registration deviation.

The post-processing circuit 13 functions to perform processing such as checking for mistakes.

The control circuit 14 controls the registration correction device 1 in response to a signal representing a registration deviation.
5 or 16, and operates to match the registration 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 registration adjustment between the R channel and the B channel.

Next, to explain the operation, when adjusting the registration, the test pattern 5 is inserted into the optical path of the imaging optical system, and the resulting pattern is imaged on the imaging surface of the imaging systems 1 to 3 of each channel. let

As a result, the video signal based on the test pattern 5 obtained from the G channel imaging system 1 is inputted to the binarization circuit 8 via the preprocessing circuit 6, and is shaped into a pulse by the pulse comparator 12 via the intermediate processing circuit 10. A video signal is supplied.

On the other hand, for example, if the switch 17 is switched to the R channel side as shown in the figure, the video signal of the test pattern 5 from the R channel imaging system 2 will be transferred to the preprocessing circuit 7, the binarization circuit 9, and the intermediate processing circuit 11. The signal is shaped into a pulse via the G channel signal and is supplied to the pulse comparator 12 in the same manner as the G channel signal.

Then, the deviation between the pulsed signals supplied to the pulse comparator 12 corresponds to the deviation of the video signals between the G channel imaging system 1 and the R channel imaging system 2. Therefore, the control circuit 14 takes in the signal from the pulse comparator 12 via the post-processing circuit 13, changes the control signal supplied to the registration correction device 15 of the R channel accordingly, and controls the pulse comparator 12. By making the pulse-like signal from the R channel input to match the pulse-like signal of the G channel, the registration adjustment of the R channel with respect to the G channel is automatically performed.

Then, when the switch 17 is switched to the B channel, the control signal for the B channel registration correction device 16 changes in the same way, and the registration adjustment of the B channel with respect to the G channel is automatically performed. In addition,
If a microcomputer is used as the control circuit 14, all of the above operations can be performed by programs in the microcomputer.

Therefore, according to the auto-registration device shown in FIG. 2, the registration can be readjusted easily and in a short time whenever necessary, and the registration can be maintained accurately at all times. You can obtain a color television signal with excellent image quality.

However, the auto-registration device of the type shown in FIG. 2 requires a special imaging optical system with a built-in test pattern 5, which tends to increase the size of the device and further increases the cost. There was a drawback that it was difficult to use.

An object of the present invention is to provide an auto-registration method that eliminates the drawbacks of the prior art described above and automatically performs accurate registration adjustment while imaging a general subject without using a test pattern. is to provide.

To achieve this objective, the present invention extracts a level difference signal representing a predetermined level difference between video signals from two channels whose registration is to be adjusted, and respective edge signals of these video signals; The present invention is characterized in that the level difference signal and the edge signal are logically operated to obtain a discrimination signal representing the direction of registration deviation, and the registration correction device is controlled in accordance with this discrimination signal.

Embodiments of the autoregistration apparatus according to the present invention will be described below with reference to the drawings.

FIG. 3 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. 3, 18 is a comparator, 19,
20 is a first and second edge signal generation circuit, and 21 is an arithmetic circuit. In addition, in this embodiment, the switch 17 has a 2-circuit, 2-contact configuration instead of a 1-circuit, 2-contact configuration, and the other interlocked switch 1
7' is shown.

A comparator 18 converts a binarized signal G obtained by processing the G channel video signal from the preprocessing circuit 6 into a binarization circuit 8 and a binarization circuit 9 which processes the video signal from the R or B channel from the preprocessing circuit 7. 2 processed with
This is a circuit that generates a level difference signal E by extracting only the portion where the level difference with the digitized signal RorB is more than a predetermined value. You can think of it as an OR circuit.

Edge signal generation circuits 19 and 20 respectively generate binary video signals G and RorB (hereinafter, RorB is referred to as X).
This is a circuit that extracts the edge portion of and outputs edge signals Ge and Xe, and is composed of, for example, a differentiating circuit and a waveform shaping circuit.

The arithmetic circuit 21 receives the level difference signal E and the edge signal.
The following three types of logic using Ge and Xe as input,
, a circuit that generates two types of discrimination signals Y and Z by performing calculations according to one of the following types of logic. In the case of logic, it includes, for example, four AND circuits, one inverter, and two In the case of logic, it is made up of, for example, two AND circuits, and in the case of logic, it is made up of, for example, four 3
It is composed of an input AND circuit, three inverters, and two OR circuits, or one 3-input, 8-output decoder and two OR circuits.

Logic Y=E・Ge+・Xe Z=E・Xe+・Ge Logic Y=E・Ge Z=E・Xe Logic Y=E・Ge・+・Xe Z=E・・Xe+・Ge・In the arithmetic expression, a signal with a bar above a character such as, for example, represents an inverted signal, and therefore, the above signal represents a signal obtained by inverting the polarity of signal E.

Next, the switch 17' controls the registration correction devices 15 and 16 for the R channel and B channel.
This is for switching the control signal for the switch 17 in conjunction with the switching operation of the switch 17.

Next, the operation of this embodiment will be explained with reference to timing charts shown in FIGS. 4 to 6.

Of these timing charts, Fig. 4 shows the case where the arithmetic circuit 21 is configured to perform calculations according to logic, Fig. 5 shows the case according to logic, and Fig. 6 shows the case according to logic. However, since similar operations can be obtained in either case, they will be explained together in the following explanation.

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

Then, at the output of the binarization circuit 9, the video signal from the R channel imaging system 2 is binarized as shown in FIGS. 4, 5, and 6 (hereinafter simply referred to as figures). It appears as a signal R, which is processed by the edge signal generation circuit 20, and an edge signal Re of the R channel appears at its output as shown in the figure.

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 19, and the edge signal Ge of the G channel appears at its output as shown in the figure.

At the same time, the signals G and R obtained by binarizing the G channel and R channel video signals, which are the outputs of these binarization circuits 8 and 9, are also input to the comparator 18, and the levels of these signals G and R are An output signal that is generated only when there is a difference of more than a predetermined value,
That is, the level difference signal E is extracted as shown in the figure.

Therefore, these edge signals Ge and Re and the level difference signal E are supplied to the arithmetic circuit 21, where arithmetic processing is performed according to one of the three types of logic described above, and two types of discrimination signals Y are generated. and Z are taken out as shown in the figures (the logic is shown in FIG. 4, the logic is shown in FIG. 5, and the logic is shown in FIG. 6) and supplied to the control circuit 14. The discrimination signals Y and Z at this time are R for the G channel, as described later.
The control circuit 14 generates a control signal in response to the discrimination signals Y and Z, and outputs the control signal to the registration correction device 1 via the switch 17'.
5 to change the deflection characteristics of the imaging system 2 for the R channel so as to match it with the G channel, thereby automatically correcting the registration deviation.

Once you have finished adjusting the registration of the R channel to the G channel, switch switch 1.
7 and 17' in the opposite direction to that shown in Fig. 3, that is, to the B channel, the registration adjustment of the B channel with respect to the G channel will be performed in the same way as in the case of the R channel described above. Complete.

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

First, centering adjustment of registration will be explained.

1 to 6 in the figure are reference channels, that is, G
This is a case where the compared channel that should be centered with respect to the channel, for example, the R channel, is shifted, and 1 and 2 in the figure are the G channel and R channel.
This shows when the video signal levels of the channels are almost equal and the pulse widths of the respective binarized signals G and R are equal. 3 and 4 are higher than the video signals of the G channel. If the signal level is low and the binarized signal G has a wider pulse than the R channel, then in 5 and 6, the R channel video signal has a higher level than the G channel video signal, and 2 These diagrams show cases in which the pulse width of the quantized signal R is wider than that of the G signal. In either case, the number of pulses Yn of the discrimination signal Y and the number of pulses Zn of the discrimination signal Z are not equal, and the R channel When shifts to the right on the monitor image plane, that is, when 1, 3, and 5 in the figure, Yn>Zn
On the other hand, when there is a shift to the left, that is, at 2, 4, and 6 in the figure, Yn<Zn, and the direction of the registration shift can be determined from the magnitude relationship of the number of pulses of the discrimination signals Y and Z. become.

Also, 7 and 8 in the figure are cases where the R channel matches the G channel, and 7 of these is the case where the G channel matches the G channel.
8 indicates when the level of the video signal of the channel is higher than that of the R channel, and conversely, the level of the video signal of the R channel is higher than that of the G channel, but in either case, the discrimination signal is Number of pulses Yn and Zn appearing as Y, Z
For, Yn=Zn.

Therefore, the control circuit 14 inputs the discrimination signals Y and Z from the arithmetic circuit 21, compares the pulses Yn and Zn, and outputs opposite control signals when Yn>Zn and when Yn<Zn. The centering correction device 15 generates a signal and supplies the signal to the centering correction device 15 via the switch 17'. 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. By performing such control so that Yn=Zn, it is possible to automatically adjust the centering of the R channel with respect to the G channel.

Note that although FIGS. 7 and 8 only show 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, the video of these channels If the signal levels are equal and the pulse widths of signals G and R are equal, the level difference signal E
becomes "0", the edge signals Ge and Re are taken out as they are from the arithmetic circuit 21 as the discrimination signals Y and Z. Therefore, even in this case, the discrimination condition representing the coincidence of centering Yn=Zn is maintained. Needless to say.

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 the figure shows only a part of the video signal of each channel, in reality, only a predetermined range of video signals including the center of the image is extracted from the video signals from each imaging system 1 to 3. Centering may be determined.

Above, we have explained centering adjustment, which is part of registration adjustment, but if you select the video signal extraction range at the periphery of the image plane and make the selected range smaller, deviations due to size, linearity, distortion, etc. will be uniform. Since the deviation can be imitated, the size, linearity, and distortion deviations can be automatically adjusted in the same manner as the centering adjustment described above, and the registration adjustment can be automatically performed.

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 containing a large amount of white and black.

By the way, in the above embodiment, FIGS.
As shown in Figures 11 and 12, when the video signals of the G channel (reference channel) and R channel (compared channel) are completely shifted and there is no overlap between them, and when either Even when one of the video signals is not obtained, the number of pulses Yn obtained in the discrimination signals Y and Z is determined according to logic, except in the cases shown in Fig. 5 11 and 12.
The number of Zn becomes equal and it becomes impossible to distinguish (but
As already explained, this auto-registration method is applied to differences in the operating conditions of color television imaging devices, changes over time, and changes in ambient temperature. Since this is for correcting registration deviations by Unless the subject to be imaged has a very special color tone (for example, in the case of a subject consisting of a single color of red, green, or blue), the video signal of any channel will not be deleted as shown in Figures 11 and 12. This is highly unlikely, and therefore, in practice, there is almost no possibility of inability to distinguish or misjudgment in any case.

Now, we have explained above about registration adjustment in the horizontal direction along the scanning direction of the image plane, but we have provided a memory that can store at least one frame of the video signal, such as RAM, and stored it in this memory. Vertical registration adjustment can be achieved by sequentially reading out video signals in the vertical direction of the image plane at desired portions of the image plane, and performing the same processing as the horizontal registration adjustment described above on the read out video signals. This makes it possible to automatically adjust the registration in both the horizontal and vertical directions. Therefore, according to this embodiment, the registration can be adjusted accurately whenever necessary. registration is automatically obtained, and color television signals of excellent image quality can be easily obtained.

As explained above, according to the present invention, the discrimination signal is obtained by binarizing the two types of video signals used for registration discrimination and then performing unique arithmetic processing. The changing parts (contours) of the signal can be extracted faithfully and accurately, and registration can be performed based on the magnitude relationship of the number of pulses of the discrimination signals Y and Z without being affected by level changes or waveform dullness of the video signal. Therefore, according to the present invention, it is possible to not only detect the displacement but also determine the direction of the displacement.
It is possible to easily and automatically adjust the registration as needed while imaging a general subject without using an imaging optical system with a built-in special test pattern, which is superior to conventional technology. Aside from its shortcomings, it can always maintain accurate registration even in imaging devices that have various operating conditions and are prone to misregistration, such as not only studio type cameras but also portable television cameras. An auto-registration device capable of obtaining color television signals can be provided in a small size and at low cost.

[Brief explanation of the drawing]

Fig. 1 is an explanatory diagram showing an example of a registration adjustment method in a conventional color television camera, Fig. 2 is a block diagram showing a conventional example of an auto-registration method, and Fig. 3 is an explanatory diagram showing an example of an auto-registration method according to the present invention. A block diagram showing one embodiment of the system, FIGS. 4, 5 and 6 are timing charts for explaining the operation. 1 to 3...imaging system, 14...control circuit, 15,
16... Registration correction device, 18...
Comparator (window comparator or exclusive OR circuit), 19, 20...edge signal generation circuit.

Claims (1)

[Scope of Claims] 1. In a television camera that is equipped with a plurality of image sensors and obtains a color television signal by combining video signals of a plurality of systems, two systems of video images for which image registration adjustment is to be performed. The signals are each binarized, and a level difference signal E which shows a logical value of 1 in a portion where the level difference between these binarized video signals G and X is equal to or greater than a predetermined value is obtained, and First and second edge signals Ge and Xe having a logic value of 1 are generated at the edge portions of video signals G and X, respectively, and these level difference signals E and the first and second edge signals Ge and Xe are generated. The following calculation characteristics,,, i.e., Y=E・Ge+・Xe Z=E・Xe+・Ge Y=E・Ge Z=E・Xe Y=E・Ge・+・・Xe Z=E・・Xe+・By processing with an arithmetic circuit having one of Ge and 1, the direction of registration deviation between the above two systems of video signals can be determined by the number of pulses of the discrimination signals Y and Z.
An auto-registration method for a color television camera, characterized in that it is configured to discriminate based on the difference between Yn and Zn.