JPH0767140A - Encoding transmitting method - Google Patents

Abstract

PURPOSE:To prevent a code word disadvantageous for the extraction or identificial reproduction of a clock from being allocated to a code word for video data by adopting a specified generating polynomial. CONSTITUTION:A signal B(X) provided by dividing a data word D(X) composed of a digital composite video signal with a generation polynomial G(X) and further dividing it with a polynomial X+1 is transmitted, on the reception side, the X+1 is multiplied to this signal B(X) and the generation polynomial G(X) is multiplied so as to recover the data word D(X). In such a encoding transmitting system, the generation polynomial G(X) is defined as a 20th-degree polynomial G(X)=X<20>+X<3>+1. By selecting the generation polynomial G(X) to the 20th degree, the data word D(X) to generate the worst pattern in transmissive data concerning the video data of a 20-bit cycle can become a code word containing any inhibit code. Therefore, bit sequence independence characteristics can be improved concerning a flat field to sppear sometimes on practical use.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of encoding and transmitting a composite video signal, and more particularly to a method of determining a generator polynomial in that case.

[0002]

2. Description of the Related Art In the current video data transmission system, SCR NRZI (Scrambled Non-Return to Zero Inverse) is used for channel coding, as specified by the serial transmission standard SMPTE 259M. The polynomial is 9th order.

FIG. 1 is a block diagram schematically showing this method. If this system is simply invented with reference to the figure, the serial video data D (X) supplied to the input terminal 1 is divided by the generator polynomial G (X) in the division circuit 2 to obtain its quotient A (X). It is supplied to the next division circuit 3.

The division circuit 3 further adds the quotient A (X) to (X +
The result is divided into 1) and the result is sent out on the transmission path 4.

On the receiving side, in the multiplier 5, the transmission line 4
The signal sent to is multiplied by (X + 1) and the product is supplied to the multiplier 6. The multiplier 6 further multiplies the product signal by the generator polynomial G
Multiply (X) to recover the original serial video data D (X).

In the above calculation, the quotient A (X) is (X +
It is divided by 1) because A (X) is represented by an NRZ (non-return to zero) signal.
This is for converting into a ZI (non-return to zero inverse) signal.

To aid in understanding the description of the invention below,
Here, the relationship between the NRZ signal A (X) and the NRZI signal B (X) will be briefly described.

For example, the above A (X) is A (X) = X ⁷ +
If X ⁴ + X ³ + X, then B (X) = A (X) (X + 1) ⁻¹ = (X ⁷ + X ⁴ + X ³ + X) (X + 1) ⁻¹ = X ⁶ + X ⁵ + X ⁴ + X ² + X. This is a 9-bit digital signal 010011
010 is converted to 001110110.

Considering the meanings of these digital signals, since the former signal A (X) corresponds to NRZ, a positive pulse for the code symbol "1" and a code symbol "0". A negative (or zero) pulse is associated with "", and when "1" continues, it does not return to zero and is represented as a continuous pulse of "1". This state is shown in FIG.

On the other hand, since the latter signal is NRZI, it is a signal indicating that the polarity of the pulse is inverted when the code symbol is "1" and is not inverted when the code symbol is "0" and the previous state is maintained. is there.

The above-mentioned digital signal 001110110
, The first two bits will retain the previous state,
A signal that inverts at the next 1 bit, inverts the next bit, and inverts at the next bit is formed.

In order to explain the difference between NRZ and NRZI in an easy-to-understand manner, the digital signal 0 described above for NRZ is used.
The same signal as 10011010 is considered as an NRZI signal and a waveform diagram is shown in the second row of FIG.

As is apparent from FIG. 2, in the NRZ signal, the code symbols "1" and "0" of the digital signal correspond to the presence or absence of the pulse, so that this is synchronized with the clock pulse shown in FIG. If read out, it can be read out as a pulse signal as shown in FIG.

On the other hand, in order to convert the NRZI signal into a digital signal, the rising and falling edges of the pulse waveform shown in FIG. 7B are read in synchronization with the clock pulse shown in FIG. The digital signal as shown in (c) is read.

In the system of FIG. 1, the video data is, for example, a coded signal digitized by a 10-bit AD converter, and one word is composed of 10 bits.

When the 10-bit expression is used,
There are 2 ¹⁰ = 1024 types of words, which are selected and assigned to video data, control signals and the like.

In this system, video data is transmitted serially, and the system operates in synchronization with clock pulses. Therefore, the digital signal is expressed as "1" when there is a pulse at a predetermined time and "0" when there is no pulse.

The synchronizing signal is sent together with the video data as a composite video signal to the receiving side through the transmission line 4, and the receiving side separates the synchronizing signal from the composite video signal to form a clock signal.

[0019]

According to the above-mentioned current method, a 9th-order polynomial G (X) = X ⁹ + X ⁴ +1 is used as a generator polynomial. In this case, a 20-bit period image often appears in practice. When the data (flat field) is input, it becomes easy to judge whether the data reproduction is good or bad.
Transmission data having a problem with SI (Bit Sequence independence) characteristics may occur.

For example, if the contents of the register of the divider are all 0 when the video data as shown in Tables 1 and 3 of 20-bit cycle of the same color on one screen are input, as shown in FIGS. Generate a transmission pattern.

The transmission pattern shown in FIG. 3 is a pattern in which n-bit "1" is followed by m-bit "0" and n and m have substantially the same magnitude, that is, a pulse waveform whose duty is close to 50%. Is.

Such a waveform has a minimum clock signal component contained in the signal waveform and is a disadvantageous pattern for extracting the clock signal. The words that produce such a transmission pattern are shown in Table 1 for a generator polynomial of degree 9. The ones shown in this table and the exchange of the first and second words shown therein are all of the code words that give rise to this kind of transmission pattern.

The transmission pattern shown in FIG. 5 is a repetition of a signal in which only one bit out of n bits is "1" and the rest are all "0".

Since this type of transmission pattern has a small ratio of "1" (mark rate), it is a signal which is disadvantageous for identification and reproduction. Table 3 shows video data that may cause such a worst pattern when the generator polynomial is of the 9th order.

The code word shown in Table 3 and its first
Swapping the word and the second word is all of the code words that can cause this worst pattern.

In Tables 1 and 3, (H) indicates hexadecimal notation, and the X mark is a code word whose use as a code word of video data is prohibited. For example, 000,
001, 002, 003, 3FC, 3FD, 3FE, 3
There are FF etc.

The transmission of codewords is done serially, in an order starting from the leftmost bit of the table and ending with the rightmost bit. Therefore, when the same word is repeated, the bit string is transmitted in the direction from the left end of the first word to the right end of the second word, and then again, the bit string is advanced from the left end of the first word to the right end of the second word.

In view of the above points, the present invention has an object of defining a generator polynomial so that a code word that is disadvantageous for clock extraction or a code word that is disadvantageous for identification reproduction is not assigned to a code word assigned to video data. And

[0029]

According to the present invention, a data polynomial G for generating a data word D (X) consisting of a digital composite video signal is provided.
The signal B (X) obtained by dividing by (X) and further by the polynomial (X + 1) is transmitted through the transmission path, and the signal B (X) is multiplied by (X + 1) at the receiving side, and Generator polynomial G
In the coded transmission method in which (X) is multiplied to recover the data word D (X), the generator polynomial G (X) is set to 2
There is provided a coded transmission method characterized in that a 0th-order polynomial G (X) = X ²⁰ + X ³ +1 is set.

Further, according to the present invention, a signal B (X) obtained by dividing a data word D (X) consisting of a digital composite video signal by a generator polynomial G (X) and further by a polynomial (X + 1).
Is transmitted via the transmission line, and this signal B (X) is received at the receiving side.
Is multiplied by (X + 1) and is multiplied by a generator polynomial G (X) to recover the data word D (X). Code words that are disadvantageous for clock extraction, identification reproduction, etc. in the signal B (X) are defined, and the generator polynomial G (X) is determined based on these code words and the data word D (X).
A method for determining a generator polynomial for determining is provided.

[0031]

According to the above-mentioned structure of the present invention, all the data words which generate the worst code pattern such as the video data word which is disadvantageous for the clock extraction and the video data word which is disadvantageous for the identification reproduction are prohibited from being used for the video data. Since it is a code, good BSI characteristics can be obtained.

[0032]

Embodiments of the present invention will be described with reference to FIGS.

As a signal representing the image data, there is an NRZI signal as shown in FIG. 3 as one of the signals (this is called the worst pattern) which is not preferable to be given to the transmission line or the recording medium of FIG.

This NRZI signal is, for example, 0011111.
1111100000000 ... The code symbols “1” and “0” are long continuous signals. If this signal is written as a code polynomial, B (X) = X ^n-3 + X ^n-4 +
It becomes like + X ^n-12 .

As is clear from FIG. 1, B (X) is A
Since it is a value obtained by dividing (X) by (X + 1), A (X) = B (X) (X + 1) to obtain A (X) in reverse.
Should be calculated.

According to the above example, A (X) = (X ^n-3 + X
^n-4 ＋ ・ ・ ・ ＋ X ^n-12 ) (X + 1) = (X ^n-2 + X ^n-12 ).
Then, 0100000100000000000000 ...
It is given by a bit string signal such as. FIG. 4 shows a signal having such a pattern.

In other words, when the value obtained by dividing the video data word by the generator polynomial has the pattern shown in FIG. 4, it is the worst pattern.

As another worst pattern, a code pattern as shown in FIG. 5 can be considered. If the video data signal given to the transmission line or the recording medium of FIG. 1 has a pattern as shown in FIG. 5, the mark ratio fluctuates greatly in the AC coupling system, which causes waviness of the baseline and is disadvantageous for identification and reproduction. is there.

For example, the signal of FIG.
When expressed by 1000000000, this is represented by a code polynomial as B (X) = X ¹⁹ + X ⁹ , and when A (X) is obtained from the above, A (X) = B (X) (X + 1) =
(X ¹⁹ + X ⁹ ) (X + 1) = X ²⁰ + X ¹⁹ + X ¹⁰ + X ⁹ That is, the code symbol “1” is continuously present in 110000000101100000000 for 2 bits and then a long “0” is present.
Is a continuous signal. This is shown in FIG.

Signal A quoted in the above description
(X) is a generator polynomial G (X) of the video data signal D (X)
Since it is divided by
(X) can be obtained.

That is, it is given by D (X) = A (X) G (X). Using this relationship, the code pattern of D (X) when A (X) takes the worst pattern shown in FIGS. 4 and 6 can be calculated backward.

Conventionally, in this type of code transmission system, a 9th-order polynomial G (X) = X ⁹ + X ⁴ +1 was used as a generator polynomial, so D (X) = A (X) (X ⁹ + X ⁴ +1) Can be used to find D (X).

By substituting the worst pattern as described above as the value of A (X) to obtain D (X), an unfavorable pattern can be found as a code word representing video data. As mentioned above, Tables 1 and 3 show such codewords.

The inventor of the present invention has obtained very good results by using the polynomial of the 20th degree G (X) = X ²⁰ + X ³ +1 as the generator polynomial.

Similarly to the above, D (X) = A (X)
Since the relationship of (X ²⁰ + X ³ +1) is established, the value of D (X) can be obtained by using this formula.

Referring again to FIG. 3, considering a code pattern which is disadvantageous for clock extraction, in general, when the data is sufficiently long,

[Equation 1] The relationship is established.

For example, the coefficient of F (X) is 10101100.
Then the complement of F (X) is 1 in 10101100.
Since (1111111) is added, (10101100) × (11) in (Equation 1) is {(1010110
0) + (11111111)} × (11), and the second item of this expression is (11111111)
Since x (11) = 100000001 and the first "1" is a carry, 00000001 is obtained, which is almost equal to zero. This means that you can safely ignore the second item. Therefore, (Equation 1) is established.

Since the code pattern of FIG. 3 is a code pattern in which "1" is continuous in the first half and "0" is continuous in the second half, the first half and the second half are in a complement relation, and the binary number 1
The value multiplied by 1 (code polynomial X + 1) is 20 bits (10
One code symbol "1" is included in 2 words (1 bit is one word).

Thus, in the present invention, the generator polynomial is selected as G (X) = X ²⁰ + X ³ +1. When this is expressed as a binary bit string, it is 10,000,000.
Since the repetition cycle of "1" is long since it is 10,000,1001, D (X) becomes almost the same value as A (X).

Table 2 shows the values actually calculated from the relationship of D (X) = A (X) (X ²⁰ + X ³ +1).
When all the values shown in this table and all the values obtained by replacing the first word and the second word in this table are combined, all the video data words that are disadvantageous for clock extraction are covered.

Looking closely at Table 2, all of the video data words disadvantageous for the clock extraction shown here are 000 in either one of the two words (first word and second word) constituting each code word. Since the prohibition codes are (H), 001 (H), etc., the code patterns shown in this table are not originally used for encoding video data.

Therefore, with the same code word as in the current method,
Only by making the generator polynomial into the twentieth order, it is possible to prevent generation of the data word which is disadvantageous for the above clock extraction.

In the same manner, the video data word D (X) which generates the code pattern which is disadvantageous for the identification reproduction shown in FIG. 5 can be obtained.

When the code pattern of FIG. 5 is represented by B (X), A (X) is A (X) = B (X) (X + 1),
It is given by the code pattern shown in FIG.

The video data word is D (X) = A (X) G
It is given by (X), but G (X) is "1" as described above.
Is sufficiently long, D (X) has almost the same pattern as A (X).

Table 4 shows the patterns of the video data words that are actually calculated and are disadvantageous to the identification reproduction. As is clear from the table, each code word includes a prohibition code which is prohibited as a video data word in the first word or the second word, and therefore it can be identified by using a twentieth-order generator polynomial as in the present invention. No data word that is disadvantageous for reproduction is generated.

In summary, when the generator polynomial is of the 9th order as in the conventional encoding, the worst pattern exists for the video data of 20-bit period.

That is, when video data of a certain 20-bit cycle is input, there is a possibility that transmission data, which is disadvantageous for clock extraction and identification reproduction, as shown in Tables 1 and 3 is generated.

Therefore, in practice, a 20-bit period flat field often created by CG (computer graphics), that is, an image having the same one screen (or one line) is likely to cause practical problems.

According to the present invention, the generator polynomial is selected to the twentieth order so that the data word that causes the worst pattern in the transmission data with respect to the video data of the 20-bit period becomes the code word including the prohibition code. Therefore, it is possible to improve the BSI characteristic with respect to a flat field that often appears in practical use.

Although the twentieth-order generator polynomial has been described in the above description, the present invention is not limited thereto, and it goes without saying that various modifications are included without departing from the technical idea of the present invention. is there.

[0062]

According to the above configuration of the present invention, it is prohibited to use all the data words which generate the worst code pattern such as the video data word which is disadvantageous for the clock extraction and the video data word which is disadvantageous for the identification reproduction, for the video data. Since it is a code that has been set, good BSI characteristics can be obtained.

[Table 1]

[Table 2]

[Table 3]

[Table 4]

[Brief description of drawings]

FIG. 1 is a diagram showing an example of an image communication system to which the present invention is applied.

FIG. 2 is an explanatory diagram showing a comparison of signal forms.

FIG. 3 is a waveform diagram showing an NRZI code pattern of video data which is disadvantageous for clock extraction.

FIG. 4 is a waveform diagram showing an NRZ code pattern of video data which is disadvantageous for clock extraction.

FIG. 5 is a waveform diagram showing an NRZI code pattern of video data that is disadvantageous for identification reproduction.

FIG. 6 is a waveform diagram showing an NRZ code pattern of video data which is disadvantageous for identification reproduction.

[Explanation of symbols]

 1 Input Terminal 2 Generator Polynomial Divider 3 Code Converter {(X + 1) Divider} 4 Transmission Line or Recording Medium 5 Code Converter {(X + 1) Multiplier} 6 Generator Polynomial Multiplier 7 Output Terminal

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical indication H04N 7/24

Claims (2)

[Claims] 1. A signal B (X) obtained by dividing a data word D (X) consisting of a digital composite video signal by a generator polynomial G (X) and further by a polynomial (X + 1) via a transmission line. A coding transmission method for transmitting, and multiplying this signal B (X) by (X + 1), and multiplying by a generator polynomial G (X) on the receiving side to recover the data word D (X), Let G (X) be a 20th degree polynomial G (X) = X
A coded transmission method characterized in that it is set to ²⁰ + X ³ +1. 2. A signal B (X) obtained by dividing a data word D (X) consisting of a digital composite video signal by a generator polynomial G (X) and further dividing by a polynomial (X + 1) via a transmission line. The generator polynomial G (X in the coded transmission which is transmitted and which multiplies the signal B (X) by (X + 1) and the generator polynomial G (X) at the receiving side to recover the data word D (X). ) Determining the signal B on the transmission line.
The code words which are disadvantageous for clock extraction, identification reproduction, etc. in (X) are defined, and these code words and the data word D
A method for determining a generator polynomial for determining the generator polynomial G (X) based on (X).