JP3938256B2 - Television camera device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a television camera device that generates discretely sampled video signals, such as deterioration of frequency characteristics of luminance signals, moire phenomenon caused by sampling aliasing distortion components, or generation of false color signals of subject outlines. The present invention relates to an improvement of a television camera device for minimizing it.
[0002]
[Prior art]
In a television camera using a solid-state imaging device such as a CCD (Charge Coupled Device), aliasing occurs in the sampled video signal when sampling the video signal from the solid-state imaging device. Here, when the sampling frequency fs is not sufficiently high, aliasing distortion enters the video signal band, and moire is generated in the reproduced image to deteriorate the image quality.
[0003]
Conventionally, as a method for reducing the aliasing distortion, a so-called matrix operation function in a television camera apparatus having a solid-state image sensor that respectively generates G (green), R (red), and B (blue) video signals is used. A method called “spatial pixel shifting” is known, for example, as disclosed in Japanese Patent Publication No. 55-19553. In this method, as shown in FIG. 3, for example, each pixel of the light receiving unit (imaging unit) of the G CCD is arranged in the horizontal scanning direction with respect to each pixel of the light receiving unit (imaging unit) of the R and B CCDs. The pixels are arranged so as to be relatively shifted by a half of the pixel interval Px in the horizontal scanning direction.
[0004]
Thus, by arranging each CCD, the phase of the R and B video signals is shifted by 180 degrees with respect to the G video signal, and the phase of the aliasing distortion that occurs is also shown by the broken line and the alternate long and short dash line in FIG. As shown in FIG. Here, the luminance signal Y obtained by performing the matrix operation on the G, R, and B signals is represented by the following equation, for example.
Y = 0.59G + 0.30R + 0.11B
G, R, and B in the above formulas indicate G, R, and B signal levels, respectively. The level ratio (white level ratio) of the white levels (white reference level) of G, R, and B is 1: 1.
[0005]
Therefore, in the luminance signal Y obtained by this matrix operation, the G aliasing distortion component (indicated by the broken line waveform in FIG. 2) and the anti-phase R and B aliasing distortion components (indicated by the one-dot chain line in FIG. 2). (Indicated by the waveform) is partially canceled, and the aliasing distortion component is reduced. However, the aliasing distortion component Yd of the luminance signal Y shown by the two-dot chain line waveform in FIG.
Yd = GRB = 0.59-0.30-0.11 = 0.18
As shown, about 18% of the luminance signal Y remains.
[0006]
For this reason, there is a aliasing distortion component almost equal to the video signal component in the frequency band near ¾ of the sampling frequency fs (near 0.75 fs), and the aliasing distortion component in the frequency band of 0.75 fs or higher. As a result, there is a problem in the video quality due to the S / N ratio, and the video signal cannot be put into practical use, and the video signal cannot be made higher in resolution.
[0007]
Therefore, as a conventional method for increasing the resolution of the video signal, as disclosed in Japanese Patent Laid-Open No. 8-223590, spatial pixel shift is performed between the R and B image sensors and the G image sensor. Is used to switch the R video signal and G video signal whose sampling phases are reversed by the spatial pixel shift, with the opposite phase of the sampling frequency to generate a signal that alternately switches between the R signal and the G signal. Further, after the band is limited by a band pass filter, it is combined with the luminance signal output from the matrix circuit to improve the deterioration of the frequency characteristic of the luminance signal. This improvement method is called digital sampling aperture (DSA).
[0008]
In this DSA, the R signal and the G signal are sampled in such a manner that they are alternately switched every cycle of a frequency twice as high as the sampling frequency of the solid-state imaging device, and the white level ratio between the R signal and the G signal is set to 1. Switch as pair 1. By doing so, an auxiliary luminance signal is generated. Therefore, this auxiliary luminance signal has a color balance characteristic with respect to the luminance signal output from the matrix circuit based on the G, R, and B signals obtained by sampling at a frequency twice the sampling frequency. Although it is slightly different, the frequency characteristics are almost the same signal. The auxiliary luminance signal is switched by setting the white level ratio between the R signal and the G signal to 1: 1, and the phases thereof are different from each other by 180 degrees and have opposite phases. Therefore, in the auxiliary luminance signal, the aliasing distortion component of the R signal and the aliasing distortion component of the G signal are completely canceled and do not include the aliasing distortion component. The auxiliary luminance signal is band-limited by a bandpass filter. Therefore, by adding such an auxiliary luminance signal to the luminance signal output from the matrix circuit, the SN ratio of the finally obtained luminance signal is improved to a high frequency close to the sampling frequency.
[0009]
Furthermore, as described in the specification of 1998 Patent Application No. 159001, spatial pixel shifting is performed between the R and B imaging elements and the G imaging element, and the sampling phase is mutually changed by the spatial pixel shifting. The inverted R signal and G signal are switched at the opposite phase of the sampling frequency to generate a signal in which the R signal and the G signal are alternately switched, and the R signal that cancels the aliasing distortion component of the luminance signal output from the matrix circuit After mixing, the band was further limited by a bandpass filter, and then combined with the luminance signal output from the matrix circuit to achieve both the correction of the frequency characteristic degradation of the luminance signal and the reduction of the aliasing distortion component of the luminance signal. .
[0010]
[Problems to be solved by the invention]
The method of adding the auxiliary luminance signal to the luminance signal output from the matrix circuit to increase the resolution of the video signal includes a means for sampling the G signal and the R and B signals that are 180 degrees out of phase. A sampling means is required, and an auxiliary luminance signal and an auxiliary R signal are generated and added after the phase and the band are set to a predetermined value. With increased power and costs.
[0011]
In addition, the method of replacing the high frequencies of all video signal outputs with signals that are alternately switched between the R signal and the G signal increases the circuit scale in proportion to the number of horizontal pixels, resulting in an increase in power consumption and cost.
[0012]
For this reason, in a small and low-priced television camera device or a television camera device using a solid-state imaging device having a large number of pixels so that the number of horizontal pixels is about 1200, correction of the aliasing distortion of the luminance signal is performed so much. There is nothing. Also, it is not often performed to process the signal by shifting the sampling phase of the G signal and the sampling phases of the R and B signals by 180 degrees.
[0013]
[Means for Solving the Problems]
In order to solve the above-mentioned problems, the present invention
The phases for sampling the R, G, and B signals are the same.
Also, generation of auxiliary luminance signals and auxiliary R signals, band limitation, delay, and replacement of high-frequency signals are not performed.
Further, the luminance signal (Y) is generated from the R, G, B signals by a matrix calculation of double speed sampling every half of the sampling period (1 / fs) of the solid-state imaging device, and is output by double speed sampling.
Further, before the matrix calculation, the high frequencies of the R and B signals are amplified from the ratings, and the phase of the G signal is delayed by approximately half of the sampling period (1 / fs) from the phase of the R and B signals.
[0014]
As shown in FIG. 5, in the luminance signal output from the conventional matrix circuit, the aliasing distortion component Yd of the luminance signal Y indicated by the waveform of the two-dot chain line in FIG. 5 is Yd = GRB = 0.59-0. .30-0.11 = 0.18
18% of the luminance signal Y remains.
[0015]
On the other hand, in the method of the present invention, when the high frequency band of G is attenuated and the high frequency bands of R and B are amplified, if the signal levels of the high frequency bands G, R, and B are GH, RH, and BH, the luminance The high frequency range YH of the signal Y is expressed by the following equation, for example.
YH = 0.5GH + 0.375RH + 0.125BH
Here, the aliasing distortion component YdH of the high frequency luminance signal Y is
YdH = GH-RH-BH = 0.5-0.375-0.125 = 0
Thus, the aliasing distortion component is canceled in the high frequency range YH of the luminance signal Y. However, a little aliasing distortion component of the luminance signal Y is left only in the low frequency range.
[0016]
G is constant, and by increasing the high frequency amplification of R and B, that is,
YH = 0.59G + 0.432R + 0.158B
Then, the deterioration of the frequency characteristic of the luminance signal can be improved, and the aliasing distortion component YdH of the high frequency luminance signal Y is
YdH = GH-RH-BH = 0.59-0.432-0.158 = 0
Thus, the aliasing distortion component is canceled in the high frequency range YH of the luminance signal Y. However, a slight distortion component of the luminance signal Y ′ remains only in the low frequency range.
[0017]
Here, since the analog / digital conversion is performed by double-speed sampling and output, deterioration of the frequency characteristic of the luminance signal due to the sampling aperture of analog / digital conversion can be improved.
[0018]
In addition to the horizontal shift of half a pixel, the high-frequency phase of the G signal is delayed from the phase of the R and B signals by approximately half of the sampling period (1 / fs), so that the contour of the subject whose video signal level changes suddenly The false color signal generated in is canceled out.
[0019]
As a result, the level of the final output luminance signal becomes higher than the aliasing distortion component up to the sampling frequency, the SN ratio is improved, the resolution of the television camera device is improved to the sampling frequency, and the false color signal generated in the contour It is possible to achieve both reductions in the circuit while suppressing an increase in circuit scale.
[0020]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.
FIG. 1 is a block diagram showing a configuration of an embodiment of a television camera apparatus according to the present invention, and FIG. 6 is a timing chart showing signal waveforms of respective parts for explaining the operation of FIG.
[0021]
In FIG. 1, R, G, and B CCDs 30, 32, and 34 are IT (Interline Transfer) type CCDs in this example, and are imaging units (light receiving units) 30a, 32a, and 34a and readout units 30b and 32b, respectively. , 34b. As the CCD, a FIT (Frame Interline Transfer) or FT (Frame Transfer) type CCD, or a CMOS image sensor may be used.
[0022]
Charges obtained by photoelectric conversion in the pixels of the imaging units 30a, 32a, and 34a are sequentially read out to the reading units 30b, 32b, and 34b, respectively. Here, the pixels of the imaging unit 32a of the G CCD 32 are pixels in the horizontal scanning direction in the horizontal scanning direction with respect to the pixels of the imaging units 30a and 34a of the R and B CCDs 30 and 34, as in FIG. They are shifted by a half of the interval Px.
[0023]
The sampling pulse generator 40 includes a sampling pulse signal S1 (FIG. 6A) and S2 (FIG. 6B) having a predetermined period (frequency fs) and a clock signal S3 having a frequency 2fs (FIG. 6G). A sampling pulse signal S4 ((h) in FIG. 6) having a frequency of 2 fs is generated.
[0024]
The sampling pulse signal S1 is supplied to sampling hold circuits (SH circuits) 4, 5, and 6, respectively. The sampling hold circuits 4, 5, and 6 include the readout unit 30b of the R CCD 30 and the readout unit 32b of the G CCD 32. The video signals output from the reading unit 34b of the B CCD 34 are sampled and held.
[0025]
Low-pass filter circuits (LPF circuits) 7, 8, and 9 limit the output video signals Vr, Vg, and Vb of the sampling hold circuits 4, 5, and 6 to a predetermined band, respectively.
[0026]
Reference numerals 10, 11, and 12 denote R, G, and B variable gain video amplifier circuits (AGC circuits) that can perform shading correction, white correction, and black correction, and reference numeral 13 denotes variable gain video amplifier circuits 10, 11, and 12. A pre-gamma amplification circuit that performs pre-gamma (pre-γ) processing on the output signal of R, G, and B. Reference numerals 24 and 26 denote high-frequency amplifier circuits for amplifying the high-frequency portions of the R and B video signals output from the pre-gamma amplifier circuit. Reference numeral 25 denotes a delay circuit for delaying the G video signal output from the pre-gamma amplifier circuit. (Delay
Line: DL).
[0027]
Reference numeral 14 denotes an A / D converter for analog / digital conversion (A / D conversion) of each R, G, B output signal of the pre-gamma amplifier 13 by the sampling pulse signal S2. Reference numeral 15 denotes various digital signal processing, such as gamma correction processing, and matrix processing for generating the luminance signal Y and the color signals Cr and Cb from the R, G, and B output signals (see reference numeral 23 in FIG. 1). A digital signal processing circuit, i.e., a DSP (Digital Signal Processor) circuit.
[0028]
Here, the matrix processing of the luminance signal Y is performed at the sampling rate of the frequency 2fs, but the other signal processing is performed at the frequency fs, whereby the circuit scale can be further reduced.
[0029]
Reference numeral 16 denotes a D / A converter that performs digital / analog conversion (D / A conversion) on the output of the DSP 15 with a sampling pulse signal S4 having a frequency of 2 fs.
[0030]
Here, each pixel of the imaging unit 32a of the G CCD 32 is ½ of the pixel interval Px in the horizontal scanning direction in the horizontal scanning direction with respect to each pixel of the imaging units 30a and 34a of the R and B CCDs 30 and 34. Corresponding to the fact that the phase of the G video signal is shifted relative to the phase of the sampling pulse signal S1, the delay circuit (Delay Line: DL) 25 adjusts the phase of the high frequency region of the G video signal as described above. It is delayed by a half cycle and becomes VG ′ ((e) in FIG. 6).
[0031]
As a result, the false color signal of the subject outline is canceled. The matrix circuit 23 of the DSP 15 uses a clock signal S3 having a frequency of 2 fs to perform a double speed sampling matrix every half of the sampling period (1 / fs) of the solid-state imaging device from the input R, G, B signals. A luminance signal (Y) is generated by calculation.
[0032]
As described above, the spatial arrangement of the R and B CCDs and the G CCD is shifted every half of the horizontal pixels, so that the signal output from the matrix circuit portion 23 of the DSP is simulated for each pixel of the CCD. The luminance signal Y ′ has a switching period equivalent to the switching period of the luminance signal sampled at the frequency 2fs that is twice the output frequency of the output signal.
[0033]
Further, in the case of the circuit configuration not including the LPFs 21 and 22 and the modulation circuit 27 shown in FIG. 1, the Y / Cr / Cb signal output of the D / A converter 16 is input, and the clock component at the time of D / A conversion is input. And a low-pass filter circuit (LPF) 17 to 19 for removing the luminance signal Y ′ is output as a luminance signal of the television camera device. In addition, by incorporating the LPFs 21 and 22 and the modulation circuit 27, the color signal (C) can be directly generated.
[0034]
Through the above operation, the relative gain of the luminance signal Y ′ obtained by amplifying the high frequencies of the R and B signals and matrix-calculated has improved frequency characteristics as compared with the original luminance signal Y as shown in FIG. The aliasing distortion is reduced only in the low frequency, and the occurrence of the moire phenomenon due to the aliasing distortion is further reduced, and the relative gain of the luminance signal is larger than the aliasing distortion relative gain up to near fs, the SN ratio is improved, and the limit resolution is improved. Can improve.
[0035]
With respect to the black and white subject having a width of 1.5 pixels in FIG. 6 (o), the G signal in FIG. 6 (c) becomes 2 pixels wide by shifting the horizontal pixels. 6 (d) R and B signals are 1 pixel wide, and the delayed G signal of FIG. 6 (e) and the R and B signals of FIG. The luminance signal Y in (k) of FIG. 6 is also 1.5 pixels wide. Accordingly, since the occurrence of aliasing distortion is eliminated, the limit resolution can be up to about fs.
[0036]
Further, the high-frequency amplification of the R and B signals can be realized by changing the high-frequency compensation capacity of the amplifier 13, and this embodiment is a π-type filter (one inductor) as a G signal delay circuit in the form of a conventional television camera. And it becomes a simple structure only adding 2 capacity.
[0037]
In the configuration described above, high-frequency amplification is performed for the R and B video signals, and the level of the amplified high-frequency portion is relatively increased compared to the high-frequency portion of the G video signal. It is made like. However, although not shown, it does not have a circuit for amplifying the R and B video signals at high frequencies, but instead, by adding a high frequency attenuation circuit for attenuating the high frequency part of the G video signals, It goes without saying that the level of the high frequency part of the R and B video signals can be relatively increased as compared with the high frequency part of the G video signal, as in the above-described operation. Further, by using both a high-frequency amplification circuit that amplifies the high-frequency portion of the R and B video signals and a high-frequency attenuation circuit that attenuates the high-frequency portion of the G video signal, the high frequency of the G video signal is obtained. The level of the high frequency part of the R and B video signals may be relatively increased as compared with the part.
[0038]
FIG. 7 is a block diagram showing another embodiment of the television camera apparatus according to the present invention, and FIG. 8 is a timing chart showing signal waveforms of respective parts for explaining the operation of the embodiment of FIG. 7, components having the same functions as those in the embodiment of FIG.
[0039]
In this embodiment, the sampling hold circuits 4, 5, 6, the variable gain amplifier circuits 10, 11, 12 and the A / D converter 14 are integrated in one ASP (Analog Signal Processor) IC 28. Although not shown here, an ASP / IC in which the pre-gamma amplifier circuit 13 is also integrated may be used.
[0040]
Further, in the present embodiment shown in FIG. 7, output is performed from the high-frequency amplifier circuits 24 and 26 that amplify the high-frequency signal of the R and B video signals, the delay circuit 25 that delays the G video signal, and the matrix operation circuit. The delay circuit 20 for delaying the luminance signal and the LPF circuits 21 and 22 integrated so as to be realized by a signal processing circuit in the DSP 15 are used.
[0041]
Therefore, the matrix circuit 23 obtains the same signal Y ′ as in the embodiment of FIG. 1, and the luminance signal Y of FIG. 8 (k) is also 1 for the monochrome pixel of 1.5 pixel width of FIG. 8 (0). .5 pixel width, which indicates that there is no aliasing distortion and the limit resolution is close to fs.
[0042]
Further, when the delay circuit 20, the LPF circuits 21 and 22 and the modulator 27 are incorporated, the chroma signal C and the composite synchronizing signal VBS can be directly generated.
[0043]
【The invention's effect】
As a result, the circuit scale of the small digital three-plate color camera is reduced, and it is easy to reduce the size, reduce the power consumption, and reduce the price.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of an embodiment of a television camera device according to the present invention. FIG. 2 is a diagram showing frequency characteristics of relative gains of a video signal and aliasing distortion. FIG. 4 is a diagram illustrating a method of spatially shifting pixels between imaging units of an image sensor. FIG. 4 is a diagram illustrating a frequency characteristic of a relative gain of a luminance signal and aliasing distortion related to the present invention. FIG. 6 is a timing chart showing signal waveforms of respective parts for explaining the operation of FIG. 1. FIG. 7 is another embodiment of the television camera apparatus according to the present invention. FIG. 8 is a block diagram showing an example configuration. FIG. 8 is a timing chart showing signal waveforms at various parts for explaining the operation of FIG.
4, 5, 6: Sampling circuit (SH circuit) 7, 8, 9, 17, 18, 19, 21, 22: Low-pass filter circuit (LPF circuit) 10, 11, 12: Shading correction, white correction, Variable amplification circuit including black correction, 13: pre-gamma amplifier, 14: A / D converter, 15: DSP, 16: D / A converter, 20, 25: delay circuit, 23: matrix circuit, 27: modulator, 24, 26: High-frequency amplifier circuit, 28: ASP, 30, 32, 34: Solid-state imaging device such as CCD, 40: Sampling pulse generator.

Claims (3)

G (green), R (red), B (blue) has a first, a second, a third solid-state image sensor for generating video signals, respectively, each light-receiving unit of each solid-state image sensor, A plurality of light receiving pixels arranged at predetermined intervals in the horizontal scanning direction and the vertical scanning direction; In the television camera device, the light receiving pixels of the light receiving unit are arranged to be spatially shifted in the horizontal scanning direction by 1/2 of the predetermined interval of the light receiving pixels.
Sampling means for sampling the video signal from the solid-state image sensor at a predetermined period (1 / fs), and a predetermined frequency of two of the sampled G, R, and B video signals High-frequency amplification means for amplifying (high-frequency amplification) signals in the above frequency bands,
Of the sampled G, R, and B video signals, a video signal other than the two video signals subjected to the high frequency amplification and the two high frequency amplified video signals of the first sampling means. A matrix operation circuit that performs a predetermined matrix operation for digital signal processing at a half period (1/2 fs) of the sampling period (1 / fs) and generates a luminance signal;
Means for delaying the phase of the G video signal input to the matrix arithmetic circuit by a half of the sampling period (1 / fs) from the phase of the R and B video signals input to the matrix arithmetic circuit When
A television camera device comprising: G (green), R (red), B (blue) has a first, a second, a third solid-state image sensor for generating video signals, respectively, each light-receiving unit of each solid-state image sensor, A plurality of light receiving pixels arranged at predetermined intervals in the horizontal scanning direction and the vertical scanning direction; In the television camera apparatus, the light receiving pixels of the light receiving unit are arranged in the horizontal scanning direction so as to be spatially shifted by ½ of the predetermined interval of the light receiving pixels in the horizontal scanning direction.
Sampling means for sampling the video signal from the solid-state imaging device at a predetermined cycle (1 / fs), and a predetermined frequency of one of the sampled G, R, and B video signals High-frequency attenuation means for attenuating signals in the above frequency band (high-frequency attenuation);
Of the sampled G, R, and B video signals, two video signals other than the one video signal subjected to the high-frequency attenuation and the high-frequency attenuated video signal are used for the first sampling means. A matrix operation circuit that performs a predetermined matrix operation for digital signal processing at a half period (1/2 fs) of the sampling period (1 / fs) and generates a luminance signal;
Means for delaying the phase of the G video signal input to the matrix arithmetic circuit by a half of the sampling period (1 / fs) from the phase of the R and B video signals input to the matrix arithmetic circuit When
A television camera device comprising: G (green), R (red), B (blue) has a first, a second, a third solid-state image sensor for generating video signals, respectively, each light-receiving unit of each solid-state image sensor, A plurality of light receiving pixels arranged at predetermined intervals in the horizontal scanning direction and the vertical scanning direction, respectively, and the second and third solid-state image sensors relative to the light-receiving pixels of the light-receiving unit of the first solid-state image sensor. In the television camera device, the light receiving pixels of the light receiving unit are arranged to be spatially shifted in the horizontal scanning direction by 1/2 of the predetermined interval of the light receiving pixels.
Sampling means for sampling the video signal from the solid-state image sensor at a predetermined period (1 / fs), and a predetermined frequency of two of the sampled G, R, and B video signals High-frequency amplification means for amplifying (high-frequency amplification) signals in the above frequency bands,
Of the sampled G, R, and B video signals, a video signal other than the two video signals subjected to the high frequency amplification and the two high frequency amplified video signals of the first sampling means. A matrix operation circuit that performs a predetermined matrix operation for digital signal processing at a half period (1/2 fs) of the sampling period (1 / fs) and generates a luminance signal;
Means for delaying the phase of the G video signal input to the matrix arithmetic circuit by a half of the sampling period (1 / fs) from the phase of the R and B video signals input to the matrix arithmetic circuit When
High frequency attenuation means for attenuating (high frequency attenuation) video signals other than the two video signals subjected to the high frequency amplification,
The television camera apparatus according to claim 1, wherein the matrix calculation means generates a luminance signal by performing a predetermined matrix calculation from the two high-frequency amplified video signals and the high-frequency attenuated video signal.

Images