JPS62150989A - Digital filter - Google Patents

Abstract

PURPOSE:To obtain a digital filter to be simply realized as a circuit by employing a digital filter designed for a sampling frequency integer times of a color subcarrier frequency and a digital filter identical in constitution and coefficient, to a signal processing circuit for a digital composite color video signal whose sampling frequency is integer times of a horizontal scanning frequency. CONSTITUTION:When the sampling frequency is close to nfsc, the phase difference between the color-width carrier wave component of neighboring two sample values comes to be equal to 2pi/n. Since the phases of the color subcarrier wave components coincide with each other at every n-pieces of sample values in case the sampling frequency in nfsc, color signal processing can be made easy by making use of this phenomenon. Therefore, the constitution and the coefficient that can be easily realized with simple circuitry in case of the frequency of nfsc, are immediately used for the digital filters used in the digital composite video signal processing circuit whose sampling frequency is a frequency integer- times of a horizontal scanning frequency close to nfsc. In such a way, the signal processing circuit complex in constitution hitherto, can be simplified, and a desired characteristic can be obtained in approximation.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to digital signal processing of directly encoded composite color video signals.

Conventional technology Currently, color television systems include NTSC and PAL.
The sampling frequency for digitization is 13.5M, which is an integral multiple of the horizontal scanning frequency fh for all three methods in case of separate coding.
Hz is selected. On the other hand, in the case of direct encoding, if sampling is performed at 13.5 MHz, the sampling frequency will be equal to each color subcarrier frequency fsc in either method.
Since it is not an integer multiple of 3, the phases of the color subcarrier components of each sample value are not aligned. As a result, an example of a one-dimensional YC separation filter will be given below as a conventional example of a digital filter whose configuration is complicated.

As the frequency characteristic H(f') of a one-dimensional YC separation filter, the following three conditions (1) H(f8c) = 1) H'(
Consider the following equation that satisfies f8o)-〇(3)H(o)=C. (Opening [Digital Signal Processing of Images] Expanded Edition Nikkan Kogyo Shimbunsha... (1) However, Z-' = exp (-j2πf/A) --
−(2)θ −2π9. /f8...
...(3) If 9 is 13.5MHz, in the case of NTSC, θ= (35/66)π ...
...(4) In case of PAL θ= (709379
/1080000)π・・・・・・(5)

An example of the configuration of a digital filter having this characteristic is shown in FIG. In the figure, 9 to 12 each represent a delay element of 1 clock period (-1713, 5 MHz), and 19
21 represents a coefficient unit that performs an operation of multiplying the coefficient values written in the figure. Also, 26°27 represents an adder/subtractor,
3o is an input terminal, and 33 is an output terminal. Further, the frequency characteristics of this filter are shown in FIGS. 2 and 4. FIG. 2 shows the case of the NTSC system, and FIG. 4 shows the case of the PAL system. In both figures, the characteristics of this conventional example are shown by broken lines.

The composite color video digital signal applied to the input terminal 30 is expressed as follows.

Xfn=Ym 10cm However, Xe: mth sample value Ym: mth luminance signal component Ce ° mth color subcarrier component The previous sample value and the two sample values xm-1"Ym-10cm-1"・"""・('1'
)5 Vane xm-2""Ym-20cm-2""""'-(8)
is obtained. The coefficient unit 19 gives 2xm- to the 0 point in the figure.
1 is obtained, and by the adder/subtractor 26, at point D in the figure, -xm+2xm-1-xm-2...
-(9) is obtained. Here, both Ym and Ym-1 and both Ym-2・・・・・・・・・・・・(
10) Cm=Ccosθ.・・・・・・
・・・・・・・・・(11) Cnll-1 bow C animals (θ.-
〇) ・・・・・・・・・・・・(12)Cm
-2 Cape C■ (θ.-20) However, C9θ. is a constant
・・・・・・・・・・・・(13) Therefore, -xm +2xm-1-xm-2=-Ym-Cm +2
Ym-1+2cm-1-Ym-2-Cm-2 C (-tei θ.+2 fish (θ.-〇) Part (θ.-20)) = 2 (1 part θ) Cm-1 ・・・・・・・・・・・・・・(14) 6-・-.

becomes. 11.12 and 20.27 in the figure work similarly, and E
At the point, 4(1-cosθ)2cm-2'...
-(15) is obtained and normalized by the coefficient multiplier 21, and only Cm-2, that is, the color subcarrier component, is obtained at the output terminal 33.

In this conventional example, as shown at 20 and 21 in the figure, the coefficients are not integers or fractions of integers, and the digital circuit that implements this becomes complex.

Next, consider a circuit for correcting errors that occur in the transmission path. Generally, errors that occur in the transmission path are corrected by an error correction code, but if an error that exceeds the code's correction ability occurs, a sample value that is close to the sample value containing the error will be corrected on the screen. Correct sample values are estimated by interpolation from the values, and sample values containing errors are replaced with the estimated values (referred to as correction). When sampling at 'fsc, the phase difference between adjacent sampling points is π/2, so this interpolation is easily performed using a digital filter, but when sampling at 13.5MHz, it is 7 bases. .

Since the phase difference between sample points that coincide with each other is (35/66)π, it is difficult to perform similar interpolation using a digital filter.

Problems to be Solved by the Invention As stated above, when a composite color video signal is directly encoded at an integral multiple of the horizontal scanning frequency, the signal processing circuit for the digital composite color video signal may become complicated. There is a problem.

In view of this point, it is an object of the present invention to provide a digital filter that is easy to implement as a circuit in the above-mentioned case.

Means for Solving the Problems In order to solve the above-mentioned problems, the present invention samples a frequency nfsc that is an integer multiple of the fsc between color subcarrier components, which is close to the sampling frequency f8 that is an integer multiple of the horizontal scanning frequency. This is a digital filter that can be easily realized as a circuit assuming that the f8 frequency is the same, and operates at the above f8 without changing the configuration and coefficient values.

If the working sampling frequency is close to fBc, the phase difference between the color subcarrier components of two adjacent sample values will be approximately equal to 2π. If the sampling frequency is n, then the n sample values Since the phases of the color subcarrier components are aligned for each color, color signal processing is easy if this is utilized.Therefore, according to the present invention, the color signal processing is easy.
A configuration of a digital filter for a digital composite video signal processing circuit in which the sampling frequency is an integral multiple of the horizontal scanning frequency close to fBc, with the same configuration and coefficient values that can be realized with a simple circuit when the sampling frequency is nfsc. By using coefficient values as , the conventionally complicated signal processing circuit can be simplified, and desired characteristics can be approximately obtained.

Embodiment FIGS. 1 and 3 show the NTSC system and PA of the present invention.
FIG. 4 shows a block diagram of a YC separation digital filter in a first embodiment and a second embodiment for the L method.

Figure 1 is for the NTSC system, and Figure 3 is for the P
This is for the AL system. In both figures, numerals 1 to 8 each represent delay elements of 1/9 page lock period (=1713.5 MHz), and numerals 13 to 18 each represent a coefficient multiplier that performs an operation of multiplying the coefficients written in the figures. Further, 22 to 25 represent adders and subtracters, 28 and 29 are input terminals of each filter, and 31 and 32 are output terminals of each filter.

Now, in the equations (1), <2), and (3) mentioned above,
Substitute nfsc for 9.

However, in the case of NTSC system, n = 4...
...(16) In the case of PAL method n = 3 ...
...(17). At this time, equation (1) becomes...(18)...(19). In this example, the above equations (8) and (9) are converted to h-13,
Applying to 5MH2, Z-' = exp (-j2r'f8), f,
= 13.5MHz - (2)) 1o Ba.

That is. Equation (18) corresponds to FIGS. 1 and 2, and Equation (19) corresponds to FIGS. 3 and 4.

The color video signal applied to the input terminal is separated into Y and C components by the same operation as in the conventional example, and only the color difference signal components appear at the output terminal. At this time, as can be seen by comparing with Fig. 6, the coefficient of the coefficient multiplier is an integer or a fraction of an integer,
The configuration of the digital circuit that realizes this becomes simple.

Further, the frequency characteristics of the digital filter in this example are shown in FIGS. 2 and 4. Here, Figure 3 is NTS
In the case of C method, the solid line in the figure represents the characteristic of equation (18). In addition, the broken lines in the figure indicate -13,5MHz, AC in equations (1), (2), and (3).
This shows the characteristics of a conventional example using the color subcarrier frequency of the system. As can be seen from the figure, the characteristics between ◯ and fscO are in good agreement.

Center frequency 9. The amplitude value at is approximately 0.991 in this embodiment, compared to 1 in the conventional example, and the difference is small. There is a large amplitude difference between the characteristics of the conventional example and the characteristics of this embodiment near 6.75 MHz, but since the original 11-page NTSC signal has a band of approximately 4 MHz,
The characteristics around this area are not a problem.

On the other hand, Figure 4 shows the case of PAL system, where the solid line represents the characteristic of equation (19), and the broken line represents (1), (2), (3).
In the equation, f8 = 13.5 MHz + f6 (represents the characteristic in the conventional example when 2 is FAI and the color subcarrier frequency of the system.

As can be seen from the figure, the characteristics of this embodiment match well with those of the conventional example. The amplitude at fsc in this example is:
It is about 0.999, and the difference is negligible.

As mentioned above, in the case of the NTSC system, 4f8o, PA
In the case of the L method, by applying a digital filter with the same configuration and coefficients as a simple YC separation digital filter with a sampling frequency of 3fBc, it is possible to achieve YC separation with a simpler configuration and almost equal frequency characteristics compared to the conventional example. A digital filter is obtained.

Next, a third embodiment of the present invention is shown in FIG. This embodiment is a digital filter used for correction when an error occurs in a transmission path. FIG. 4B shows the position on the screen of the sample used for error correction in this embodiment. In this embodiment, the composite color 7 video signal is an NTSC signal, and the sampling frequency is 13.5 MHz. In the figure (b), 63 represents the scanning line in which the sample to be erroneously corrected exists, and 62 represents the scanning line immediately before that. 54 is the sample to be corrected in error, and 55~
61 is a sample point used for error correction. FIG. 5(,) is a block diagram of the digital filter of this embodiment.

In the figure, 34 to 43 represent delay elements of one clock period, and 44 represents a delay element of NH-5 clock periods, where NH is the number of sample points per one scanning line. 46~
Reference numeral 50 represents a coefficient machine that multiplies the coefficients shown in the figure, and 61 represents an adder. 52 is an input terminal of this filter, and 53 is an output terminal.

A sample value sequence containing an incorrect sample value enters this digital filter from the input terminal, and when the incorrect sample value 6413 when Beno reaches point F is x'rn, the delay element 34 .about.44, sample values 65 to 61 appear at the G-M point, and are weighted by coefficient units 45 to 49 and added by adder 51. Further, the value intersection that is normalized by the coefficient unit 5o and appears at the output terminal 53 is expressed as follows.

△ 1 xm= B (Xm-, 3+3XTn-1+3Xm+1
"m+3-2(-Xm-2(-X+2"m-NH-x
m-NH+2) ) (21) where % id represents the kth sample value.

Now, the error-free value of the m-th sample value is X□−Y +
COQSθ. , θ. is a constant ・・・・・・・・・(2
2), then in the NTSC signal, the phase of the color subcarrier is inverted for each scanning line, and 49. Since the phase difference of color subcarrier components between adjacent sample values when sampling with
・---(2s)xm-1 both Y” Ccos (θ
. -H)=Y+C3Ba. −・−・(24) 14 peno xm+1 舛Y” Ccos (θ.+2)=Y−C
s stone θ.・-・-(25)xm+3 Iwa Y+Ccos
(θ.+π)−Y+Cs1n#0 −=−(26)
Xm-NH-2# Y + Ci (%-π-π) = Y+
C cos θ. ...(27)Xm-NH Misaki Y+ Cco
s (θ.-π)-Y-C-θ.・・・・・・(28
)xm-NH+2 #Y+Ccos(190-π+π)
=Y+Ccos#, (29).

Therefore, assuming the sampling frequency is 'fsc', △ Xm unit] “(8Y18Cω1t)-Y1C team d and others answer.

...(3), and the correct sample value - is output terminal 6
Obtained in 3.

Therefore, we changed this filter to 13 without changing the configuration and coefficient values.
．． Applying this to the case of sampling at 5 MHz, the phase difference θ of the color subcarrier component between different sample values is:
As shown in equation (4), θ-(ss/ee)π, so by the same calculation as above, Δxm coyu(8(Y+Ccosθ.crs3θ) + 8
Ccos 66 sin 2θ)= Y+ (cos
fj +snn t) ) Ccosθ.

Cow Y+0.9901 CcOSθ. Co-xm...
(31) 15 vanes are obtained at the output terminal, and a nearly correct value is obtained.

As described above, according to the present invention, interpolation in a correction circuit, which was conventionally difficult at a sampling frequency of 13.5 MHz, can be performed using a simple digital filter, and a more accurate correction value can be obtained.

Effects of the Invention In the digital signal processing of digital composite color video signals, the present invention utilizes a digital filter with a simple circuit implementation designed for a sampling frequency that is an integral multiple of the color subcarrier frequency, and a digital filter with the same configuration and coefficients. By applying the filter to a signal processing circuit for a digital composite color video signal whose sampling frequency is an integer multiple of the horizontal scanning frequency, a digital filter that substantially satisfies desired characteristics and is easy to implement in circuit can be obtained.

[Brief explanation of drawings]

FIG. 1 shows a block diagram of a YC separation digital filter in a first embodiment of the present invention. Further, FIG. 2 is a frequency characteristic diagram thereof, which is shown together with the frequency characteristics of a corresponding conventional digital filter (the vertical axis represents amplitude and the horizontal axis represents frequency). In the figure, the solid line represents the characteristics of the embodiment, and the broken line represents the characteristics of the conventional example. FIG. 3 shows a block diagram of a digital filter in a second embodiment of the invention. Fig. 4 (,) is a frequency characteristic diagram, and Fig. 4 (b) is an enlarged view of the vicinity of the peak, together with the frequency characteristics of the corresponding conventional YC separation digital filter (the vertical axis is the amplitude, and the horizontal axis is the frequency). In this figure, solid lines and broken lines are the same as in FIG. 2. FIG. 5(a) is a block diagram of a digital filter used for interpolation during error correction in the third embodiment of the present invention;
Figure (1)) is a schematic diagram showing the arrangement of sample points on the screen. FIG. 6 shows a block diagram of a conventional YC separation filter corresponding to the first and second embodiments. 1-4, 5-8...Delay element, 13-15°1
6 to 18... Coefficient unit, 22 * 23 t 2
4, 25...Adder. Figure 4 (cL) AL [gate 1-i,) (Toph Figure 5 (OLn (F))

Claims (3)

[Claims] (1) In a signal processing circuit for a digital composite color video signal that directly encodes a composite color video signal using a sampling frequency that is an integral multiple of the horizontal scanning frequency, n
Using the fact that the color subcarrier components of the sample values of the digital composite color video signal have the same phase every n sample values, the frequency nf_s_c, which is n times the color subcarrier frequency f_s_c, is an integer and the sampling frequency is nf_s_c. A digital filter having the same configuration and coefficient values as a digital filter designed as a sampling frequency, and operating with a sampling frequency that is an integral multiple of the horizontal scanning frequency. (2) The sampling frequency, which is an integral multiple of the horizontal scanning frequency, is 13.
2. The digital filter according to claim 1, wherein the composite color video signal is an NTSC signal having a frequency of 5 MHz, and nf_s_c is a frequency 4f_s_c that is four times the color subcarrier frequency f_s_c of the NTSC system. (3) The sampling frequency, which is an integral multiple of the horizontal scanning frequency, is 13.
At 5MHz, the composite color video signal is PAL or S
2. The digital filter according to claim 1, which is an ECAM system, and wherein nf_s_c is a frequency 3f_s_c that is three times the color subcarrier frequency f_s_c of the PAL system.