JPS6126745B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a compression method for binary graphic data with a high degree of compression.

When transmitting binary graphic (character) data by facsimile, etc., data compression is performed to reduce the amount of information to be transmitted, but conventional compression methods only correlate each pixel in one or two rows. Therefore, there was a limit to the degree of compression. The present invention aims to compress three lines at once, improve the data compression rate, and realize high-density encoding.

The data compression method of the present invention converts binary graphic data into 3
The graphic data of the three rows are extracted row by row and divided into blocks so as to correspond to one of the graphic data groups of a specific pattern prepared in advance.
The feature is that the code is represented by a code indicating the length until a change in value information occurs, a label code indicating black and white of the block, and a black and white order change code indicating a change in the black and white order. This will be explained in detail with reference to.

FIG. 1 is a flowchart showing the conditions for dividing graphic data such as characters to be transmitted into blocks, and FIG. 2 is a diagram showing the types of graphic data blocks of specific patterns used in the present invention. In the present invention, graphic data is extracted in three rows at a time, and the data is further divided into blocks of 11 types from 0 to 10 shown in FIG. Here, FIG. 2 0 is a white block in which the pixels or figure data in 3 rows and n columns are all white, and FIG. 2 1, 2, and 3 each have 3 rows, and the first, second, and 3 rows, columns i to n, columns j to n,
It is a block in which columns k to n are black and the rest are white.
Note that although the binary image information is white and black here, it may be any color. Also, Figure 2 4,
5 and 6 still have 3 rows, and the 1st row, 2nd row
A block in which columns 1 to i, columns 1 to j, and columns 1 to k of the third row are black and the rest is white;
8, 9, 10 also have three rows, and columns i-n of the first row and columns j-n of the second row are black, columns j-n of the second row and columns i-n of the third row. is black, columns j to n of the first row and columns i to n of the second row are black, columns i to n of the second row and columns j to n of the third row are black, and the remaining blocks are white. It is. Here n, i,
j and k are integers, and as is clear from the drawing, n>k>j>i. The graphic data block of 3 rows and n columns can be any of the 11 types of blocks shown in FIG. 2, and the procedure for forming the block will be explained below with reference to FIG. 1.

The graphic data output by a facsimile machine is retrieved three rows at a time from memory, etc., and the three rows of the i-th column are compared with the three rows of the adjacent i+1 column. ), memorize the length a up to that point. Then, the boundary judgment condition 1, ``Has a second information change occurred in the same line within the block?'' is checked, and if YES, there is a block change, that is, the area up to that point is considered to be one block. If NO, the next boundary judgment condition 2, ``Has an information change occurred in two or more lines within the block?'' is checked, and if NO, it is determined that there is no change in the block, that is, the change is an intra-block change. YES
If so, a monotonous judgment is made as to whether only one line changes and the lines that change to a different pattern are adjacent to each other. If YES, it is assumed that there is no block change and the length b up to that point is memorized, and if NO, it is assumed that there is a block change. do. Here, the change to a different pattern is shown in Figure 2, 7.
Arrow F in 8 refers to a change in _one direction, not a change in the opposite direction. When there is a block change, the length c up to that point is memorized, but as is clear from Figure 2, the length b
There are also cases where there is no , or there is no a and it becomes c directly.

By performing such processing, any binary figure can be divided into any one of 0 to 10 in FIG. 2. For example, a block to which a black portion shown by a dotted line in FIG. 2 is added is judged to have a block change in judgment 2, and up to the nth column becomes one block, and the next block becomes a black block. Although black blocks are not shown in FIG. 2, there are also blocks whose black and white are reversed, and these are identified by labels. Furthermore, in the pattern shown in Fig. 3, when the nth column from the left is reached, the information changes twice in the second row, so it is determined that there is a block change based on judgment 1, and the next column is the first column of the next block. becomes. In this row, a white block is followed by another white block, so please refer to Figure 2 1 as described below.
The black and white order change code shown in 1 is attached. The "number within a block" in FIG. 2 indicates how many parts the block is divided into. The signal for transmitting such graphic information is, for example, the one shown in FIG. "Length sign"
It can be composed of four types of codes, and a label code indicating whether the block is black or white.

For the label, look at the first three lines of a block and choose the color (label) that occupies two of the black and white lines. The label is attached to the beginning of the block, and in the present invention, the screen is divided by black-and-white order reversal parts, so it is normal for black and white blocks to be lined up alternately, with a black block following a white block, and a white block after that. If the black-and-white order change code shown in FIG. 2 is attached to the head of the block whose order has been disrupted, it is not necessary to attach the black-white identification code to the head of each block.

In decoding, after identifying the block code, the lengths are decoded as many times as the number in the block. FIG. 4 shows a block diagram of the encoding circuit, and FIG. 5 shows a block diagram of the decoding circuit. In these figures, 11 is a column change detection circuit, i, i+1,
i+2 Detects black and white changes in each line and uses a counter
CTR ₁ , CTR ₂ , and CTR ₃ detect the aforementioned lengths a, b, and c. 12 is a block change detection circuit;
Block changes are detected according to the flow shown in FIG.
13 is an encoding circuit which encodes the block code, encodes the lengths a, b, and c, and sends them out as data compression codes D, C, and C. In the decoding circuit, upon receiving this code DCC, the block identification circuit 14 learns the type of block, and the distance code end detection circuit 15 decodes the lengths a, b, c, etc.
Restore each row of i, i+1, and i+2.

As is clear from the above explanation, in the present invention, when transmitting character image information by facsimile etc.
The screen is divided into three lines, and the three lines are further subdivided in the line direction according to predetermined conditions to create a block with a specific pattern. Since it is a compression code, a high degree of information compression can be obtained. For example, a blank area can be represented by a single white block as shown in FIG. 2, and the degree of compression can be extremely high. Further, by utilizing the feature of continuous blocks, additional information such as white and black label information can be omitted, and the amount of transmitted information can be reduced.

[Brief explanation of the drawing]

Figure 1 is a flowchart showing the procedure for creating blocks, Figure 2
FIG. 3 is an explanatory diagram showing the types of blocks, FIG. 3 is an explanatory diagram showing an example of blocking, FIG. 4 is a block diagram of an encoding circuit, and FIG. 5 is a block diagram of a decoding circuit. In the drawings, 0 to 10 in FIG. 2 indicate graphic data blocks of a specific pattern.

Claims (1)

[Claims] 1. Extract the binary graphic data in three rows at a time, divide the three rows of graphic data into blocks corresponding to one of the graphic data groups of a specific pattern prepared in advance, and divide the blocks into blocks using a block identification code, Data characterized in that it is represented and encoded by a code indicating the length until a binary information change occurs in each line in a block, a label code indicating black and white of the block, and a black and white order change code indicating a change in the black and white order. compression method.