JPS61118034A - Method and apparatus for code communication - Google Patents

Abstract

PURPOSE:To limit the range of an error within transmitted information even if a code error takes place in the transmission line by decreasing the rate where the boundary of information is made not clear in code transmission and reception through a transmission line where error might be caused. CONSTITUTION:A code given to a terminal 501 is inputted to a flag detection circuit 502, and when a flag F is detected, a reset pulse is outputted to a line 503 and initializes a mode discrimination circuit 504. Since the next input code is a mode identification code without fail, an A switch 505 selects an output line 506. A reset pulse 503 cases a B switch to transmit a 1-bit mode identification code to a register 508. The mode discrimination circuit 504 discriminates the modes S, W and when picture information is inputted to the A switch 505, the picture information is inputted to a register 511. The mode identification code on the line 506 is inputted to a correction code generating circuit 512 to generate a correction code. The mode discrimination circuit 504 connects a C switch 515 to a flag generating circuit 510, the mode identification code register 508 and a picture information code register 511.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a coded communication method and apparatus for coding and transmitting continuous notifications.

Conventional configuration and its problems In a conventional communication system that transmits information consisting of multiple reports using codes of different encoding systems with different numbers of codes per report, when a code error occurs on the transmission path, This has the disadvantage that the boundaries between notifications become unclear, and once the boundaries become unclear, the meaning of subsequent codes becomes unclear, and the codes can no longer convey information.

Conventional code transmission will be described below with reference to the drawings.

In the following explanation, the original image information obtained by sampling or quantizing a multi-tone photographic image is divided into blocks of P1i pixels in the main scanning line direction and P2 pixels in the sub-scanning direction. The unit of report is the encoded original image information.

FIG. 1 shows the relationship between each pixel set 1o and blocks. In FIG. 1, a grid of thin lines represents each pixel. Arrow 11 indicates the main scanning direction, and arrow 12 indicates the sub-scanning direction. Thick line 13A.

13B, 13C, 13D, 13E, and 13F indicate block boundaries.

Next, regarding the encoding of image information within the Glock, block encoding will be described in terms of a tari. As an example, boundaries 13A, 13B
, 13D, and 13E will be described below. There are 4×4=16 pixels in the block,
When the luminance information of each pixel is quantized with 8 bits, there are 16×8=128 bits of original image code in one block. Give each pixel in the block a number K corresponding to its position. In the above case, -0,...15
), and its pixel value is p(ic). Here, the block encoding method will be explained step by step.

(1) Find the average value U of pixels P (g) in the block.

(2) Difference E between the average value U and each pixel value in the block (6)
seek.

]:(K)=P(K)-UE Compare the absolute value of E(K) with constant C. E(K) l E(K) l <C for all pixels in the block,
(K=o.

..., 15), this block defines the mode to be S. The value of C above is the pixel value P(6
) is expressed in 8 bits and ranges from δ to 255, the value 2 in the 2nd to 16th place, for example 4 (C=4), is selected. When the mode is S, this block has few changes in pixel values within the block.

The average value U is substituted for the pixel value of each pixel. Therefore, when an S mode block is encoded, the number of codes consists of a mode identification code, for example, 1 bit, and an average value code, for example, 8 bits, resulting in 9 bits per block. On the other hand, within the block l E(K) l>C, (K=O, -・
-.-, 15) (7) When a pixel exists, mode W is defined. When the mode is W, each pixel is compared with the average value, pixels larger than the average value are set as H group, those smaller than the average value are set as L group, and the average of the pixel values of H group is the pixel value of H and L groups. The average of
shall be. When the mode is W, the number of codes per block is 1 bit for the mode identification code, 8 bits for the H group average value H, 8 bits for the L group average value, etc.
There are a total of 33 bits, 16 bits (this is referred to as T), one bit per pixel, indicating whether each pixel belongs to the H group or the L group. Therefore, when consecutive blocks are sequentially encoded using the above-mentioned encoding method, the number of codes per block will be irregularly arranged, such as those with 9 bits and those with 3.3 pits. . This is shown in the second ward. In FIG. 2, 20 is a series of code strings. 21 is the code of mode S, 21A is the mode identification code, 21B is the code indicating the average value, 22 is the code of mode W, 22A is the mode identification code, 22B is H,
It is the sum of L and T.

In the code string shown in Figure 2, if a code that is not an identification code in a block causes an error on the code transmission path for some reason, the pixel values in that block will not be reproduced correctly, but the effect will be on the block. Stay within. However, if the mode identification code is incorrect, the effect affects all subsequent codes, and from then on, pixel values become incoherent, and the original image can no longer be reproduced at all. That is, 21
If an error occurs in the mode S identification code 21A, the receiving device determines that the following 32 bits are one code and that the next block's identification code comes next, but there is no identification code. I haven't. In this case, once the identification code is incorrect, it is no longer possible to decode it.

Purpose of the Invention The present invention reduces code errors on the transmission path by reducing the rate at which the boundaries of notifications become unclear when transmitting and receiving codes using a transmission path where errors may occur. Even if an error occurs, the effect of the error is limited to the relevant report and is prevented from spreading to subsequent reports, thereby reducing the deterioration of the transmitted information.

Structure of the Invention The present invention forms a notification using at least an identification code and a transmission information code as one unit, and for code information consisting of a plurality of consecutive notifications, the identification code and the transmission information code in at least two or more notifications are The above object is achieved by transmitting or receiving the transmission information codes separately.

DESCRIPTION OF THE EMBODIMENT A method according to this embodiment for reducing the probability that an identification code is mistaken relative to other codes will be described below.

FIG. 3 shows a code string 3o as one processing unit used in a code communication device according to an embodiment of the present invention. A set of multiple blocks can be considered as a processing unit. In Figure 3, 31 is 21A in Figure 2.
, 22A, and a set of codes used to reduce errors. 32 is a set of codes related to pixels, such as 21B and 22B in FIG. In FIG. 3, the set 31 of codes is before the set 32 of the same codes, but 31 may be after 32, or 31 may be divided and placed within the divided 32. It's okay.

Further, the processing unit 31 may be divided and distributed to other processing units.

FIG. 4(,) is for explaining in more detail the identification code and the set 31 of codes for reducing errors shown in FIG. 41 is an identification code, and 42 is a parity code including known ones used for error correction, a BCH code, a fire code, etc. 4(b) and (b'), in order to make it more robust against errors, two codes of FIG. 4(,) are prepared and placed on both sides of the set of pixel values 310 shown in FIG. 3. 46.49 corresponds to 41, and 4a, 50 corresponds to 42.

Next, using Fig. 5, we will discuss a code communication device that collects mode identification codes, adds error correction codes, arranges image code strings following this, and restores the original code string from these. I will explain in detail.

In FIG. S, a symbol as shown in No. 21 is inputted to a terminal 501. This code will be explained in detail again using FIG. FIG. 6(a) shows the code flow, 61
f'i is a code string indicating a fixed position of one data, for example, the beginning (this is called a flag and is indicated by the symbol F).

62 is a code string of one block in S mode (denoted by symbol S).

63 is a code string of one block in W mode (this is indicated by the symbol W). Fig. 6(b) shows an example of the flag F, and when expressed in hexadecimal,
7 F ) 1s, and is a code string whose frequency of appearance other than flags can be considered to be almost zero.

FIG. 6(C) is a detailed explanation of the codes of one block in S mode.
There is 1 bit of (symbol SM), followed by 8 bits of image information 65 (symbol SP).

FIG. 6(d) shows the code for one block in W mode, where the mode identification code 66 (symbol WM) is 1 at the beginning of the block.
This is followed by image information 67 (symbol wp) of 32 focus.

The code input to the terminal 5CM shown in FIG. 5 is input to the flag detection circuit 602. If 7 lag F is detected here, a reset pulse is output on line 503. As a result, the mode discrimination circuit S4 is set to the initial state. In this state, the code input by humans is always the mode identification code, regardless of whether the mode is S or W.
A switch 505 selects output line 606.

Since the input line 506 is selected by the reset pulse 503, the B switch 607 sends the 1-bit mode identification code input following the flag to the mode identification code register 508. The code input to the terminal 601 is also input to the mode discrimination circuit 504, so if the 1-bit mode identification code is S mode, the following 8 bits are image information, and if it is W mode. Continued<
It can be seen that 32 bits are image information.

Therefore, when the image information is input to the A switch 505, the A switch 505 is connected to the line 509, and the image information is input to the image information code register 511. Since the number of code blocks (or the number of codes) is determined in advance, when the predetermined number of mode identification codes are input to the B switch 507, the B switch 507 changes the input line to the output of the correction code generation line 512. line 5
Switch to 13. Note that the mode identification code on the line 506 is also input to the correction code generation circuit 612, and the correction code generation circuit 512 creates a correction code based on this.

For example, a 32-bit fire code is used as the correction code, but other correction codes can also be used. A 32-bit Fire code can correct burst errors of up to 11 bits. Therefore, the correction code is stored in the mode identification code register 508 following the series of mode identification codes. The mode discrimination circuit 504
4, the C switch 515 is connected to the flag generation circuit 510. Next, the flag generation circuit 610 passes the flag code through the C switch 515 and outputs the flag code to the line 519. After the correction code is stored in the mode identification code register 608, the register drive circuit 518 reads the contents of this register 608, that is, the set of the mode identification code and the correction code, once (FIG. 4(,)) or twice. (FIG. 4(b)) and sends it to the line 619.Then, the register drive circuit 518 outputs a signal,
The input line 517 of the C switch 615 is switched to output the contents of the image information code register 511 to the line 519. line 5
The code string output at 19 is restored through the reverse process described above. That is, a correct mode identification code is stored in the mode identification code register 608K using the flag, and image information is stored in the information register 511.

By first reading out one mode identification code, reading out the number of image information indicated by this code, and then reading out the mode identification code, the codes shown in FIG. 6 (&) are reproduced.

That is, when decoding using the transmitted code, the identification code is correctly decoded before the block in question, the mode of the block is determined using it, the number of codes assigned to the block is extracted, and the code number assigned to the block is extracted. By decoding the value of each pixel using , deterioration of transmitted information can be prevented.

Although the present embodiment has been described using a block encoding method as an example, the present invention is applicable to all code communication devices in which the code length per report is not constant and has a code indicating the code length of the vehicle-side report.

In addition, although the type of code string was not specified in this example,
Examples of notifications in which the number of codes that can be used by the device of the present invention are not constant include the following.

^ Bandwidth compression encoding of information for one or multiple scanning lines of a facsimile image, (B) Facsimile I
Coded as a unit of area (also called block) of multiple pixels in the main scanning direction and multiple scanning lines in several scanning directions of J pixels, (q Audio waveform divided and encoded at fixed time intervals, (To) Numerical values whose number of digits changes, (2) Encoded values of each pixel of multi-gradation pixels such as photographs and television images, (G) Others.

Effects of the Invention According to the present invention, when a report is composed of images in blocks, the rate at which the boundaries of the report become unclear is reduced, and even if there is an error in the line, image quality deterioration can be reduced. can.

[Brief explanation of the drawing]

FIG. 1 is a conceptual diagram explaining an example of an image arrangement, FIG. 2 is a conceptual diagram explaining a code string in normal transmission, and FIG. 3 is a conceptual diagram explaining a code sequence used in a code communication device in an embodiment of the present invention. 4 is a conceptual diagram illustrating the arrangement of identification codes and correction codes used in the device. FIG. 5 is a block wiring diagram of a code communication device according to an embodiment of the present invention. FIG. FIG. 2 is a conceptual diagram illustrating a code string used in the device. 501...Terminal, 502...Flag detection circuit, 504...Mode discrimination circuit, 506...
・A switch, 506・ ・B switch, 508...
Mode identification code register, 610... Flag generation circuit, 511... Image information code register, 512... Correction code generation circuit, 515...
・C switch, 518...Register drive circuit. Name of agent: Patent attorney Toshio Nakao and 1 other person No. 1
Figure U 2nd rI! J Box Figure 3 Figure 4 Cause Figure 5 Figure 6

Claims (2)

[Claims] (1) At least an identification code and a transmission information code are used as one unit to form a notification, and for code information consisting of a plurality of consecutive notifications, the identification code and the transmission information in at least two or more notifications. A code communication method in which each code is transmitted or received separately. (2) A notification is formed by using at least an identification code and a transmission information code as one unit, and a distinction is made between the identification code and the transmission information code in the notification for code information consisting of a plurality of consecutive notifications. means, a first storage means for sequentially storing only the identification codes in the plurality of notifications by the discrimination means, a second storage means for storing only the transmission information codes in the plurality of notifications, and a second storage means for storing only the transmission information codes in the plurality of notifications; , and sending means for mutually sending each piece of information stored in the second storage means.