JP2003134524A - Method and apparatus for correcting registration and television camera - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an apparatus for correcting registration of a television camera that can easily obtain registration adjustment of sufficient accuracy and easily maintain the set registration. SOLUTION: The apparatus for correcting the registration is provided with interpolation processing circuits 61R, 61G, 61B providing a delay time of one pixel interval or below and unit variable delay circuits 62R, 62G, 62B providing a delay time of one pixel interval. The interpolation processing circuits 61R, 61G, 61B finely adjust deviation in the registration and the unit variable delay circuits 62R, 62G, 62B coarsely adjust the deviation in the registration.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal processing technique of a video signal processing device including a color television camera,
In particular, the present invention relates to a registration deviation correction device suitable for a television camera of a type in which an image sensor is joined to a color separation optical system, a correction method therefor, and a television camera having a registration correction function.

[0002]

2. Description of the Related Art Conventionally, as a television camera used for applications requiring relatively high performance standards such as broadcasting, an incident image is separated into RGB three primary color images by a prism, and each color image is divided into each CCD. A so-called three-plate color television camera, which is a system of converting an electric signal by a solid-state image sensor such as the above, is in the mainstream.

FIG. 7 shows an example of such a three-plate type television camera, in which a subject image by an image pickup lens (not shown) is separated into RGB three primary color images by a color separation prism 1 and solid-state image pickup of each color channel. Elements 2R, 2G,
After being converted into an electric signal by 2B, it is supplied to the analog signal processing circuit 3.

The electric signal processed by the analog signal processing circuit 3 such as amplification is converted into an AD conversion circuit 4.
Is converted into a digital signal by the digital signal processing circuit 5, and then the digital signal processing circuit 5 performs process processing such as gamma correction and contour enhancement to obtain an output signal.

By the way, in such a three-panel television camera, usually, a method of semi-permanently fixing each of the solid-state image pickup devices 2R, 2G, and 2B to the color separation prism 1 by an adhesive or the like is adopted. But in this case 3
If the optical arrangement of the one solid-state image pickup element is not performed correctly, the position of the element cannot be adjusted to eliminate the shift.

Then, in this case, the images of the three color signals do not exactly match at the time of reproduction, that is, a so-called misregistration state, which causes deterioration of resolution, generation of false color signals, and the like. Since the image quality is significantly deteriorated, high precision is required for the alignment in the bonding of the solid-state image pickup device.

Moreover, in recent years, with the advent of the high definition television era,
Since the number of pixels of the image sensor has increased dramatically, the required alignment accuracy is also extremely high, which requires a high level of skill and a great deal of time for the alignment work, and even with slight deformation. Since it has a great influence, it is extremely difficult to maintain the performance after manufacturing.

The fact that the delay characteristics in the signal transmission system such as the signal circuit system or the cable of the image pickup device differ among R, G, and B also causes the registration deviation. Since it is necessary to perform horizontal alignment, it becomes more and more difficult to obtain high precision.

Therefore, there has been known a method in which the suppression of the registration deviation depends not only on the assembling accuracy of the optical system but is electrically corrected by the processing of the image signal. For example, in the related art, there is a method of correcting the registration deviation by adjusting the delay amount between the color signals by an analog delay line before AD converting the color signals.

[0010]

The above-mentioned prior art does not take into consideration the effect of using an analog delay line, and has problems in terms of cost increase and performance retention. That is, since an analog delay line is required to be added, extra wiring and work are required, which requires a lot of work, resulting in an increase in cost.

Further, since the delay characteristic by the analog delay line is constant, it is not possible to flexibly cope with various deviation amounts and it is also impossible to cope with the change of the characteristic, so that a problem occurs in maintaining the performance. Japanese Patent No. 2744577 discloses that when a color signal from each CCD element of R, G and B is read out, a delay circuit adjusts the phase of a reference horizontal transfer pulse based on a registration error to perform registration. A method of making a correction is disclosed. In this method, since the horizontal transfer pulse having a plurality of different frequencies is produced, the whole circuit configuration becomes complicated and the adjustment is troublesome. In addition, JP-A-6-3036
Japanese Unexamined Patent Publication (Kokai) No. 22 discloses the use of a two-dimensional filter mainly for the purpose of removing color shift due to lens aberration. The color misregistration correction using the two-dimensional filter increases the circuit scale and increases the cost.

An object of the present invention is to provide a registration correction device and a correction method and a registration correction function for a video signal processing device in which registration registration with sufficient accuracy can be easily obtained and a set registration registration state can be easily maintained. To provide a television camera having.

To achieve the above object, a registration correction apparatus according to the present invention comprises a digital sampling obtained by sampling continuous pixel signals from at least one of a plurality of color signal channels. It has an interpolation circuit that receives a pixel signal and generates a digital pixel signal at an arbitrary position between two adjacent digital sampling pixel signals by interpolation processing, and a control circuit according to a registration deviation between a plurality of color signal channels. To set an arbitrary position between two adjacent digital sampling pixel signals. Since the signal between the adjacent pixel signals obtained by the interpolation circuit is a signal delayed by the time corresponding to the registration deviation, the horizontal registration with a time accuracy smaller than the adjacent pixel interval, that is, the sampling period. The deviation can be corrected. The registration correction apparatus of the present invention is a horizontal registration shift caused by a placement error of a plurality of solid-state image pickup elements on a color separation prism, and a horizontal shift caused by a delay time difference between signal paths of a plurality of color signal channels in a video signal circuit. The misregistration in the direction can be corrected accurately and easily. further,
In the registration correction device according to the embodiment of the present invention, as the interpolating circuit, a digital sampling signal is input, and an FIR filter having a predetermined tap number is used, and the tap coefficient value of the FIR filter is set according to the registration deviation. As a result, the output of the FIR filter becomes a digital pixel signal obtained by interpolation processing. Furthermore, the registration correction apparatus according to the embodiment of the present invention has a variable delay circuit that changes the delay time stepwise in units of one sample period from zero to a predetermined time, together with the interpolation circuit, The variable delay circuit performs registration correction (coarse adjustment) in units of sample periods longer than one sample period, and the interpolator performs registration correction (fine adjustment) with a delay time shorter than one sample period. Highly accurate registration correction can be easily performed by appropriately combining the rough adjustment and the fine adjustment.

A further object of the present invention is to provide a television camera, which uses a color separation optical system and three image pickup devices, and performs registration adjustment by adjusting a signal delay time between channels. A unit variable delay unit that can be selected at intervals and an interpolation processing unit that can change the delay time within a period of one pixel period or less are provided, and the signal delay time is adjusted by the unit variable delay unit and the interpolation process. It is achieved by fine adjustment by means.

According to the above means, in a three-panel television camera or the like, a horizontal registration deviation caused by the time when the solid-state image pickup device is bonded to the optical system such as the prism, the subsequent aging, and the variation of the circuit or the like. It is possible to easily correct.

[0015]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, a registration correction device for a television camera according to the present invention will be described in detail with reference to an embodiment shown in the drawings. FIG. 1 shows an embodiment of the present invention. In FIG. 1, reference numeral 6 denotes a registration correction circuit, and in addition, a color separation prism 1, a solid-state image sensor 2R, 2G, 2B for each color channel, and an analog signal processing circuit 3 are provided. 7, the AD conversion circuit 4 and the digital signal processing circuit 5 are provided so that an image signal of a subject image is obtained as an output signal from the digital signal processing circuit 5 in the case of the conventional technique described in FIG. Is the same as.

Here, the registration correction circuit 6
Is unique to this embodiment, and this embodiment is different from the prior art in FIG. 7 in that the registration correction circuit 6 unique to this embodiment is provided between the AD conversion circuit 4 and the digital signal processing circuit 5. There is a point.

The registration correction circuit 6 has interpolation processing circuits 61R and 61R as shown in FIG.
G, 61B, variable delay circuits 62R, 62G, 62B, changeover switch circuits 63R, 63G, 63B, and a registration correction control circuit 64.

First, the interpolation processing circuits 61R, 61
G and 61B can interpolate the phase shift for each of the R, G, and B signal channels in units of time smaller than the interval between adjacent pixels, and provide data with a time delay corresponding to the transport shift (registration shift). It works to generate. Here, the term “adjacent pixel interval” refers to a row of pixels in the horizontal direction
It is the time (phase) interval between the signals of two adjacent pixels. When the signal for each pixel output from each of the solid-state image pickup devices 2R, 2G, 2B such as CCD is converted into a digital signal by the A / D conversion circuit 4, a digital image (color) of each pixel of the CCD The signal is acquired every sample time. That is, the adjacent pixel interval corresponds to one sample period of the A / D-converted digital signal. The registration correction device according to the present invention can perform registration correction with time (phase) accuracy smaller than this one sample period.

Next, the variable delay circuits 62R, 62G and 62
Similarly, B operates with a delay time individually set for each signal channel in adjacent pixel interval time units, outputs an input signal after the set delay time elapses, and phase-shifts the phase in adjacent pixel time units. It works to generate the data with the time delay required to match the deviation.

The changeover switch circuits 63R, 63G,
63B is a variable delay circuit 62R, 62G for each channel,
62B input, on the other hand, interpolation processing circuits 61R and 6R, respectively.
The output is switched between 1G and 61B, and on the other hand, the interpolation processing circuit 6
It bypasses 1R, 61G, 61B and switches to the original signal input from the input terminals R, G, B.

Further, the registration correction control circuit 6
4 is an interpolation processing circuit 61R to a unit variable delay circuit 62R.
, And to each of the changeover switch circuits 63R, a predetermined control signal N, S, E is supplied, whereby the processing necessary for correcting the registration deviation is executed.

Therefore, although not shown, the registration correction control circuit 64 is provided with a predetermined input device such as a keyboard and a control panel 65 for a man-machine interface for adjustment work. A predetermined operation input is given by an operator in registration registration.

Here, the interpolation processing circuits 61R, 61G, 6
1B and the variable delay circuits 62R, 62G, and 62B will be described in more detail. First, the interpolation processing circuit 61R,
One of the R, G, and B color signal channels will be described with respect to the principle of generating the phase shift interpolation signal with a resolution smaller than the interval between adjacent pixels by 61G and 61B.

It is assumed that a registration deviation of another color signal to be corrected with respect to a certain reference color signal corresponds to a phase deviation amount tm smaller than the interval between adjacent pixels. In this case, it is necessary to generate a correction signal with a delay time corresponding to this tm from the actual sample signal. In the registration correction device of the present invention, this registration correction signal is generated by interpolation processing using a digital sampling pixel signal sandwiching a correction signal having a delay time corresponding to tm and a digital sampling pixel signal in the vicinity thereof. The principle of generating the interpolation signal at this time will be described. For example, FIG. 4 shows an A / D conversion circuit 4 for one color signal channel.
3 shows a portion of a continuous stream of discrete sampled signals output from. As shown in FIG. 4, a plurality of data in the original pixel data string (Xn) ... Xn-1, Xn, Xn + 1,
From Xn + 2, ..., the pixel data string (Yn) that appears with a deviation time tm corresponding to a registration deviation that is less than the adjacent pixel interval for these pixel data strings (Xn) = ...
・ Yn-1, Yn, Yn + 1, Yn + 2, ... are obtained by interpolation processing. A method for obtaining a specific one data Yn in the pixel data string (Yn) by interpolation will be described below.

Therefore, first, the pixel data string Yn sampled at a phase shifted by 1 / N pitch from the pixel data string (Xn), which is a discrete signal string having the sampling frequency f1 (sampling period T = 1 / f1), has a numerical value m. Regarding the numerical value n, when m is within a predetermined integer range including n, it is generally obtained by the following formulas (1) and (1 ′).

[0026]

[Equation 1] Here, tm is the sampling point Yn (refers to the data Yn in FIG. 4)
And the phase difference of the data Xn in the vicinity of the data Yn is represented by time. Next, K (t) is an interpolation function. When t is a time parameter, the following equation ( The function represented by 2) is used. K (t) = Sin (πf1t) / πf1t (2) Here, this equation (2) is an ideal expression that generally has a frequency band of 1/2 of the frequency f1. It represents the impulse response of a simple low-pass filter and is called a Sinc function.

The above expression (1) can be expressed by the following expression (3) for the range of the data string shown in FIG.

[0028]

[Equation 2] Then, as shown in this equation (3), a finite number (K
In order to predict the data Yn from the data sequence (Xn) of each number) and obtain this data Yn by interpolation, as shown in FIG. 5, a circuit called an m-th order FIR (Finite Impulse Response) filter is used. It turns out that it should be used.

Therefore, in this embodiment, the interpolation processing circuits 61R, 61G and 61B are configured by the circuit shown in FIG. As shown in the figure, this is because (m-1) unit delay elements 611-1 to 611- (m-1) connected in series having the same delay time as the sampling period T and m multipliers 612. −1 to 612-m, and an integrator 613 that cumulatively adds the outputs of these multipliers.

Here, each of the multipliers 612-1 to 612-m
Is a coefficient K0 consisting of numerical values between 0 and 1, respectively.
Km−1 is set, and each unit delay element 611-1 to 61-1
The input data and output data of 1- (m-1) are weighted and supplied to the integrator 613.

The coefficients K1 to Km at this time are called filter coefficients or tap coefficients and are values obtained by the above-mentioned equation (2). Each value of these coefficients K0 to Km-1 is a numerical value N.
The interpolated signal according to the shift time tm shown in FIG. 4 is obtained by changing it according to the above, and a data sequence (Yn) having an arbitrary delay time can be obtained within the interval of adjacent pixels. Thus, it is possible to match the registration deviation smaller than the adjacent pixel interval (1 sampling interval T).

At this time, as is clear from the above equation (2), in this case, for any phase shift theoretically, the value of the time tm corresponding to the phase shift is given by the equation (2). It can be seen that the coefficient Ki obtained by substituting for can be applied, and the value of the time tm can be continuously applied.

Next, the variable delay circuits 62R, 62G, 62.
Describing B, these are as shown in FIG.
Unit delay elements 621-1 to 621-p (p is a positive integer) of delay time T connected in series from input to output, and input data and output data of these unit delay elements 621-1 to 621-p. It can be easily configured by the selector circuit 622 which selects and outputs one of them.

Therefore, by selecting the necessary data by the selector circuit 622, a data string (Yn) having an arbitrary delay time that changes discontinuously for each pixel unit with respect to the data string (Yx) can be obtained. Can be done, and this is 1
It is possible to match the registration deviation for each pixel from pixel to p.

Then, the registration matching in the television camera of the embodiment of FIG. 1 provided with the registration correction circuit 6 of FIG. 2 will be described. Registration is generally performed by an operator (a person who performs adjustment work) monitoring an image. Therefore, first, an image signal output from the digital processing circuit 5 is displayed on a predetermined monitor display 66. Supply,
An image is displayed, and registration is performed by operating a predetermined input device such as a control panel 65 such as a keyboard provided in the above registration correction control circuit 64.

At this time, first, the changeover switch circuits 63R, 63G, 63B are switched to the upper contacts in FIG. 2 by the control signal S, and in this state, the monitor display 66.
The registration deviation amount of each color of RGB is observed from the image displayed in FIG. As an observing method, for example, an image having a strong contrast of light and dark is picked up, and the presence or absence of a color shift which may occur at the light and dark switching position and its degree are examined. While observing as such, the control panel 65 is operated to control the output value of the control signal E output from the registration correction control circuit 64. The control signal E causes variable delay circuits 62R, 62G, 62
The delay amounts of B are independently switched and selected. The output value of the control signal E is adjusted so that the width of the registration shift that changes depending on the selection is minimized on the monitor display 66. In this case, for example, using the R channel signal as a reference signal, the registration deviation of the G channel and B channel color signals with respect to the R channel signal may be corrected by adjusting the delay amounts of the variable delay circuits 62G and 62B. Good. In this case, the reference color signal channel may be any of R, G, and B. The registration correction control circuit 64 is realized by a microcomputer and a memory device. The microcomputer incorporates a program for generating control signals N, S, E according to the input signal from the control panel 65.

Here, if the registration between the RGB images can be adjusted to a satisfactory state, the registration may be completed while fixing the state as it is.

By the way, the adjustment of the registration at this time is performed by changing the delay amount stepwise for each unit of the adjacent pixel interval (T). Accurate misalignment is possible for an integer multiple of this time T, that is, a delay corresponding to delay times 2T, 3T, ..., PT, but when it is not an integral multiple but a fractional multiple, rough registration is performed. I can only get the coordination.

Moreover, in practice, the probability that the registration shift amount will be exactly an integral multiple of the interval between adjacent pixels is quite low. Therefore, when the variable delay circuit alone cannot be used for accurate correction, after this, each changeover switch circuit 63
R, 63G, 63B are switched to the lower contacts in FIG. 2, and this time, the control panel 65 is operated to control the output value of the control signal N output from the registration circuit 64. Interpolation processing circuits 61R, 61
By changing the delay amounts tm of G and 61B independently of each other, it becomes possible to make an accurate registration registration state, that is, a state in which the registration deviation is not confirmed on the monitor display 66.

At this time, the amount of delay by the interpolation processing circuits 61R to 61R is a value smaller than the interval between adjacent pixels and can be adjusted to a substantially continuous value according to the control signal N. Therefore, according to this embodiment,
Finally, registration registration can be obtained in a sufficiently accurate state.

After the registration is completed in this way, the respective states of the control signals N, S and E output from the registration correction control circuit 64 at this time are stored in a predetermined memory (nonvolatile memory) not shown. To do.

After that, the registration correction control circuit 64 controls the control signals N, S and E stored in this memory when the television camera is put into operation.
Is read out and supplied to each of the interpolation processing circuits 61R-, each of the unit variable delay circuits 62R-, and each of the changeover switch circuits 63R-. The R, G, and B video signals that have been read out in this way and subjected to registration correction in an accurate state according to the supplied control signals N, S, and E are input to the digital processing circuit 5. As described above, the digital processing circuit 5 performs process processing such as gamma correction and edge enhancement on each video signal. In this process processing and the video processing circuit after the digital processing circuit 5, the R, G and B video signals are processed in synchronization with each other and processed at the same timing. However, as described above, since the delay and the interpolation processing for the registration correction are already performed before being input to the digital processing circuit 5, each video signal is synchronized with the video signal after the digital processing circuit 5. Even if the signal is processed, the registration correction remains valid.

Therefore, according to this embodiment, the registration can be performed easily and with sufficient accuracy only by operating the input device. Moreover, once the registration is performed, the television can be registered. Degradation of resolution,
It is possible to reliably obtain a color image signal without the risk of image quality deterioration due to the generation of a false color signal.

Further, according to this embodiment, even if the registration shift appears for some reason after performing the registration alignment, the registration alignment is repeated only by operating the input device 65. Therefore, it is possible to easily maintain the performance of the television camera.

Next, another embodiment of the present invention will be described. By the way, in the embodiment described above, each multiplier 612-1 of each interpolation processing circuit 61R, 61G, 61B.
The coefficient to be set to ˜612-m can be controlled, and theoretically, the correction of the horizontal registration deviation can be performed with any accuracy. However, as a television camera, there are cases where the registration accuracy is not so high in practical use.

Therefore, another embodiment of the present invention corresponding to such a case will be described below with reference to FIG.
Shows the registration correction circuit 6A in this embodiment, and the configuration other than this registration correction circuit 6A is the same as that of the embodiment described in FIG. When the registration correction circuit 6A is applied to a television camera, the registration correction circuit 6 shown in FIG.
Is a replacement for.

The registration correction circuit 6A shown in FIG. 3 differs from the registration correction circuit 6 shown in FIG. 2 in that the registration correction control circuit 6 is used.
4 does not supply the control signal N to the interpolation processing circuits 61R, 61G, and 61B.
The point is that the delay times tm of R ′, 61G ′, and 61B ′ are fixed.

Here, each interpolation processing circuit 61 at this time
Each delay time tm due to R ', 61G', 61B '
Is set to be, for example, T / 2 hours so that fine adjustment can be obtained for each 1/2 of the adjacent pixel interval.

Therefore, according to the embodiment of FIG. 3, the accuracy of registration alignment is ½ of the interval between adjacent pixels.
However, since the configuration of the interpolation processing circuits 61R, 61G, and 61B is simplified, there is an effect that the cost can be reduced.

By the way, in the embodiment described above,
In each case, a delay unit including an interpolation processing circuit and a variable delay circuit is provided in each of the RGB channels, but in many cases, the registration may be relatively matched between the R, G, and B channels in practice. Therefore, no delay means is provided for a specific one channel such as the G channel, or a fixed delay is provided for the other channel,
For example, even if the delay means is provided only for the R and B channels, the same effect can be obtained. The embodiment of the present invention described above, the solid-state imaging device has been described as an example of the case of three three-plate type television camera, the registration correction apparatus of the present invention is not limited to the three-plate type, four solid. It can also be applied to a television camera using an image sensor. In addition, the present invention is not limited to the registration deviation due to the spatial error caused by the inaccurate placement of the solid-state image pickup device, and the difference in the delay amount between the plurality of different color signal channels in the signal transmission path in the video signal circuit. Since it is possible to deal with registration deviation due to electrical error due to
It can be used to perform registration correction in video signal processing devices other than television cameras. Further, even when there are a plurality of different color signal channels, if registration correction is required only for a specific color signal channel, only the specific color signal channel is provided with the interpolation processing circuit and the variable delay circuit according to the present invention. A registration correction device may be provided.

[0051]

According to the present invention, it is possible to easily perform registration with high accuracy regardless of the bonding state of the image pickup device to the color separation optical system. As a result,
Since high precision is not required for joining the image pickup element to the color separation optical system, it is possible to easily obtain a large reduction in the manufacturing cost of the television camera.

Further, according to the present invention, registration registration can be easily repeated, and as a result, long-term maintenance of high performance can be easily realized.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a registration correction device for a television camera according to the present invention.

FIG. 2 is a block diagram showing details of a registration correction circuit according to an embodiment of the present invention.

FIG. 3 is a block diagram showing details of another example of the registration correction circuit according to the embodiment of the present invention.

FIG. 4 is an explanatory diagram of a principle of creating an interpolation signal in an embodiment of the registration correction device of the present invention.

FIG. 5 is a block diagram showing an embodiment of an interpolation processing circuit of the registration correction device according to the present invention.

FIG. 6 is a block diagram showing an embodiment of a unit variable delay circuit of the registration correction device according to the present invention.

FIG. 7 is a block diagram showing an example of a conventional color television camera.

[Explanation of symbols]

1 Color Separation Prism 2R Solid-state image sensor (CCD) for R channel 2G Solid-state image sensor (CCD) for G-channel 2B Solid-state image sensor (CCD) for B-channel 3 Analog processing circuit 4 AD conversion circuit 5 Digital processing circuit 6, 6A Resist Correction circuits 61R, 61G, 61B 61R ', 61G', 61
B'Interpolation processing circuits 62R, 62G, 62B Unit variable delay circuits 63R, 63G, 63B Changeover switch circuit 64 64 'Registration correction control circuits 611-1 to 611 (m-1) Unit delay elements 612-1 to 612-m Multiplier 613 Accumulators 621-1 to 621-p Unit delay element 622 Selector circuit

Claims (7)

[Claims] 1. A registration correction device for a video signal processing circuit for inputting a color signal having a registration deviation, wherein the registration correction circuit corrects the color signal having the registration deviation, and a monitor for displaying the color signal. An input control device for adjusting the registration deviation on the monitor; and the registration correction circuit for correcting the registration deviation displayed on the monitor based on the output of the input control device. And a control circuit that controls the registration correction device. 2. The registration correction device according to claim 1, wherein the registration correction circuit further includes a plurality of variable delay circuits corresponding to the color signals for delaying the color signals by at least one sample time. And a plurality of interpolation circuits corresponding to the color signals for interpolating the color signals sampled in the range of zero to the one sampling time, the control circuit including the variable delay circuit. The registration correction device is characterized in that it controls the delay time and the interpolation range of the interpolation circuit. 3. In a television camera, a resolving optical prism for resolving incident light into a plurality of color lights, and a plurality of image pickup devices for respectively receiving the color lights in order to output a color signal corresponding to the color lights, An A / D converter that samples the color signal and converts it into a digital sampling signal, a registration correction circuit that corrects the color signal having the registration deviation, a monitor that displays the color signal, and a monitor on the monitor. An input control device for adjusting the registration deviation, and a control circuit for controlling the registration correction circuit for correcting the registration deviation displayed on the monitor based on the output of the input control device. And a registration correction device. 4. The television camera according to claim 3, wherein the registration correction circuit further includes a plurality of variable delay circuits corresponding to the color signals for delaying the color signals by at least one sample time. A plurality of interpolation circuits corresponding to the color signals for interpolating the color signals sampled in the range of zero to the one sampling time from the color signal, and the control circuit includes the variable delay circuit. A television camera which controls a delay time and an interpolation range of the interpolation circuit. 5. A registration correction method for a video signal processing circuit having a registration correction circuit for inputting a color signal having a registration shift, wherein the color signal corresponding to a predetermined video pattern is input to the registration correction circuit. And a step of displaying the color signal that has passed through the registration correction circuit on a monitor, and a step of adjusting the registration deviation of the color signal displayed on the monitor based on an operation of an input control device. And a step of controlling the registration correction circuit to correct the registration deviation displayed on the monitor based on the output of the input control device. 6. A color separation optical system and three image pickup devices are used,
In a television camera adapted to perform registration adjustment by adjusting a signal delay time between channels, a unit variable delay means capable of selecting a delay time at 1 pixel period intervals and a delay time varying at a time of 1 pixel period or less An interpolation processing unit is provided, and the signal delay time can be adjusted by rough adjustment by the unit variable delay unit and fine adjustment by the interpolation processing unit. Correction device. 7. The registration correction device for a television camera according to claim 6, wherein the interpolation processing means is composed of an FIR filter.