JPH06266529A - Pattern compression system - Google Patents

Abstract

PURPOSE:To provide the high compression rate for any front pattern by providing a code-numbered sub matrix group arranged in the order of code number and code number indicators. CONSTITUTION:A pointer section storing the address of the fixed length (for example, 3 bytes) against characters, font table section including printing data to be referred by a pointer and code number (index numbers), code number indicator showing the index number position in the font table, and code umber sub matrix to be referred to by the index number are in a CGROM. In this case, the slice to be referred to by the index in the font table appears more than two times in the entire font. In general, the one which appears more than N times can be referred to by index. The 1 bit is allocated for each slice to the code number indicator and the logic '1' is stored for the slice to be referred to by index and the logic '0' is stored in the slice referring to the real data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a compression system for character font patterns output by a display device or a printing device.

[0002]

2. Description of the Related Art In the conventional so-called dictionary (coded) compression method, a character group in which one character is a bit pattern of M rows × N columns is converted into a sub matrix of M1 rows × N1 columns for each character. The sub-matrix is divided and a code number (index addition) is executed for all the sub-matrices (the same number is assigned to the sub-matrices having the same dot pattern), and
By expressing the characters divided into rows × N1 columns by the index numbers corresponding to each sub-matrix, the M1 × N1 bits of the sub-matrix are compressed to the number of bits used in the index number (usually a multiple of 8). Was there.

However, as actual compressed data, a dot pattern table (dictionary) arranged in index order is required in addition to the compressed data represented by index numbers.

[0004]

In the prior art, since the sub-matrix that appears only once in all the sub-matrices is also compressed by executing the code number, such sub-matrix is assigned an index number. The number of used bits (index bits) required more storage locations than the original data (because the size of the M1 × 1N-bit sub-matrix stored in the dot pattern table is constant regardless of the number of reference times). .

Further, since one number is assigned to the sub-matrix that is referred to only once, a large number of numbers are required for the code numbers, and the index bit becomes large. For this reason, the compression efficiency was very poor for the compression of patterns containing many such sub-matrices.

Further, even for a sub-matrix which is referred to only once at the time of decompression, it is necessary to refer to an actual sub-matrix pattern from the index number like the sub-matrix which is referred to twice or more.

[0007]

In order to solve this problem, the present invention uses N times or more (N> =) in all sub-matrices.
2) An unnumbered sub-matrix group in which the numbers appearing are executed and arranged in numerical order, and a number indicator for identifying the M1 × N1 dot pattern sub-matrix and the unnumbered sub-matrix.

[0008]

With the above structure, it is possible to perform pattern compression with higher compression efficiency.

[0009]

【Example】

(Embodiment 1) The present invention will be described in detail below with reference to the drawings. FIG. 1 is a block diagram of the first embodiment. In the figure, 1 is a central processing unit (CP) that controls a number of processes.
In U), the control procedure shown in FIG. 4 is executed. Reference numeral 2 is a random access memory (RAM), which is an area FADDR, I
Including NDFLG, SCOUNTER, INO, IADDR, 3 is a read only memory (ROM) including the control procedure shown in FIG. 4, 4 is a printing device, 5 is a printing device 4
A print buffer for holding print data of the character generator, 6 is a character generator read only memory (CGROM) for storing the compressed font and the number indicator of the present invention, and finally 7 is between the CPU and other peripheral devices. It is a common bus (SB) for exchanging data with.

Next, the structure of the CGROM 6 will be described with reference to FIG. In the CGROM, a pointer portion for storing a fixed-length (3 bytes in this embodiment) address for a character, a font table portion including print data and a reference number (index number) referred to by the pointer, and a font table There is a number indicator that indicates the index number position of, and finally, a number submatrix referenced by the index number.

As shown in FIG. 3, the font is composed of a 48 × 48 dot pattern, and the dots to be printed are assigned a logic “1” and the blank dots are assigned a logic “0”. Further, in the present embodiment, the original image is divided into sub-matrices of 48 dots in the vertical direction × 1 dot in the horizontal direction (hereinafter referred to as slices) and compressed, but the applicable range of the present embodiment is not limited to this and any size may be used. Those skilled in the art will easily understand that the font pattern and the sub-matrix of any size can be similarly processed.

Here, in this embodiment, the slice which is the index reference in the font table is 2 in all fonts.
Although it may appear more than once, it is needless to say that the applicable range of the present invention is not limited to two times, and in general, it is possible to refer to what appears N times or more.
Further, a 1-bit number is allocated to each slice of the department number indicator, and a logical "1" is assigned to a slice which is an index reference and a logical "0" is assigned to a slice which is an actual data reference.
Is stored. Therefore, the code number indicator is composed of 48 bits (6 bytes) for each character (see FIG. 3 number indicator).

The numbers with "h" added to the end of the numbers represent hexadecimal numbers, and the numbers with "b" added represent binary numbers. No decimal numbers are added unless otherwise specified.

First, the overall control of the CPU 1 having the above-mentioned configuration and executing the control procedure shown in FIG.

First, in step S100, the CPU 1 calculates the character code of the font to be output.times.3 + the pointer start address, and stores the address of the 3-byte actual font data stored at that address in FADDR. (Items enclosed in parentheses in the flowchart indicate stored contents.)
Next, in step S101, the character code × 6 bytes + code number indicator start address is calculated, and the content stored at that address is copied to INDFLG for 6 bytes. Furthermore S
In 102, SCOUNTER should be processed for 48 slices.
Substitute 48 as the initial value of Next, in step S103, IND
It is determined from the FLG whether or not the code number indicator of the corresponding slice is logical "1". If the determination is true, the contents of the address indicated by FADDR in 2 bytes are IN
Copy to O. Next, S105 calculates the INOX6 + code sub-matrix start address and stores it in IADDR. In the next S106, 6 bytes of data from the IADDR are expanded in the print buffer 5 as print data. Further, the value of FADDR is increased by 2 in S107. Next, in S108, the value of SCOUNT is decremented by 1. Next S1
In S09, it is determined whether or not SCOUNT has become “0”. If the determination is false, the process returns to S103, and if true, the process ends.

On the other hand, if the determination in S103 is false, S1
In 10 the data of 6 bytes from FADDR is directly expanded into the print buffer 5 as print data, in S111 FADDR is increased by 6 and the process proceeds to S108.

Next, an actual example will be described.

The internal character code for the character "A" is 14
10 and the pointer and font table are as shown in FIG. First, 1410 × 3 is calculated in S100, and the content of the pointer indicates the first data of “A”. For convenience, the content of the pointer is assumed to be 10000h. (FADD
R = 10000h. ) Then in s101 1410 ×
6 is calculated, the start address of the code number indicator is added to this, and the code number indicator of the corresponding character is substituted into INDFLG. Next, 48 is substituted into SCOUNTER in s102. Next, in s103, it is determined whether or not the code number indicator of the corresponding slice is 1, but in this case, since the code number indicator for the first slice is "1", the determination is "true". In s104, 2-byte data is fetched from the address indicated by FADDR and INO is substituted as an index number (INO = 0). Calculation of s105 is INO
Since = 0, the leading address of the code number sub-matrix is stored in IADDR. Next, in s106, the data is fetched as 6-byte print data from the address indicated by IADDR (in this case, 00 at the beginning of the code number sub-matrix).
h, 00h, 00h, 00h, 00h, 0ch are fetched as print data. ).

Next, in s107, FADDR is increased by 2 to 10002h. Scount is 47 in s108
However, since it is not 0, the determination in s109 becomes “false” and the process returns to s103.

Similarly, the index number is 0001h for 2, 3, and 4 slices (in the font table of FIG. 5, 1 word is Lo according to the specifications of the CPU of Intel Corporation).
Since it is entered in the order of w and High), 18h, 00h, 00h, 00h, from address 6 of code number 00h,
00h and 0ch are extended. In this way, the original image of "A" is reproduced until count == 4.

When scount == 4, the code number indicator for this slice becomes "0", so s103
Is "false" and the data of 6 bytes from FADDR is processed as print data in s110 (78h, 01h, 80h, 00h, 0h from the font table this time).
The data of 0h and 7ch is expanded as print data. ).

Similarly, "A" is extended to the end.

As described above, according to the present invention, since a slice that appears only once in all fonts exists in the font table, unnecessary indexing can be omitted and index reference can be performed with a small number of bytes.

Actually, a case of actually compressing data for 8120 characters of Mincho typeface in this embodiment will be described. The slice of all characters is 8120 × 48 = about 390,000. Among them, there are about 160,000 different patterns, and about 130,000 of these patterns appear only once.

Since 3 bytes are required as an index to simply index this as in the conventional case, 3 ×
A storage area of 130,000 = 390,000 bytes is required as an index for this pattern that appears only once.

On the other hand, according to the present invention, the number required as an index is 160,000-130,000 = about 30,000 patterns, and it can be easily understood that the index can be referenced with 2 bytes.

Therefore, according to the present invention, the difference (3-2) between the index of 390,000 bytes and the number of indexes (39
(3-1) x (390,000-130,000)
No need for bytes. (390,000 + (3-2) x (39
10,000-130,000) = 650,000). Where the number indicator is 6
Since x8120 = about 480,000 bytes is required, an excellent effect that it can be realized with a capacity of 65−48 = 17 (10,000) bytes less can be expected.

Furthermore, in the prior art, an index of 3 bytes was required, but in the case of this embodiment, it is significantly easier to process because it is 2 bytes, and index calculation can be omitted for slices that are referenced only once. Excellent effect can be expected.

(Second Embodiment) FIG. 6 shows a CGROM of the second embodiment.
It is a block diagram of FIG. In the second embodiment, the code number indicator itself is also provided in the font table for further speeding up. In this embodiment, a code number indicator for 6 bytes is provided immediately before the font data, and s101 of the first embodiment is not necessary, but the details are the same as in the first embodiment and will be omitted.

(Third Embodiment) FIG. 7 shows a CGRO of the third embodiment.
It is a block diagram of M6. In the third embodiment, in order to further increase the compression rate, the character number data is stored in the font table to which the code number sub-matrix is first referenced, and the code number sub-matrix holds the address. The corresponding code number indicator of the code number sub-matrix stored in the font table is set to logical "0", and the reference number indicator is set to logical "1" for the second and subsequent references.
The ROM is configured.

The present embodiment is exactly the same as the first embodiment except that instead of s106 in the first embodiment, 3 bytes are taken from the IADDR as a pointer and the process is executed as 6-byte print data from the address indicated by the pointer. Since it is available, details are omitted.

In the case of this embodiment, (6-3) is further added.
× It is possible to save the memory by the number of sub-matrices.

(Embodiment 4) In Embodiment 4, a blank indicator showing a blank slice in which the slice does not include any printing dots in order to further increase the compression rate, and an indicator on the left side showing that the whole slice is the same as the immediately preceding slice. To provide. FIG. 8 is a diagram showing a slice attribute indicator in which a code number indicator, a blank indicator and a left indicator are realized by 2 bits. That is, as shown in FIG. 8, attributes are assigned to each slice by 2 bits, and an attribute indicator of 2 × 48 bits (12 bytes) is provided for one character.

Next, the control procedure executed by the CPU will be described in detail with reference to FIG. s200 is the same as s100 of the first embodiment. Pointer FADD in s201
Copy 12-byte data from R to ATTRFG. Further, the pointer FADDR is set to 12 so as to point to the beginning of the font data. Next, in S202, 4 is set as the initial value of SCOUNTER in order to process 48 slices.
Substitute 8 Next, in S203, it is determined from ATTRFLG whether or not the attribute indicator of the relevant slice (2 bits of the relevant slice) is logical '00'. When the determination is true, 6 bytes of 00h are processed as print data in S204, and the process proceeds to S214.

On the other hand, if the determination is "false", then in S205, A
From TTRFLG, it is determined whether the attribute indicator of the corresponding slice (2 bits of the corresponding slice) is logical “11”. If the determination is "true", the data immediately before 6 bytes is processed as print data in S206, and the process proceeds to S214.
If the determination in S205 is "false", the ATT is determined in S207.
It is determined from the RFLG whether the attribute indicator of the corresponding slice (2 bits of the corresponding slice) is logical "01". If the determination is “true”, the contents of the address indicated by FADDR are copied into INO for 2 bytes in S208. Next, in step S209, the INO × 6 + code number sub-matrix start address is calculated and stored in IADDR. Next, in S210, 6 bytes of data from the IADDR are expanded in the print buffer 5 as print data. Furthermore, in S211, F
Increase the value of ADDR by 2. Next, in S214, SCOUN
Decrement the value of T by 1. Next, in S215, it is determined whether or not SCOUNT is not "0", and if the determination is false, S is determined.
Returning to step 203, if true, the processing ends.

On the other hand, if the determination in S203 is false, S2
In 12 the data of 6 bytes from FADDR is directly expanded into the print buffer 5 as print data, and in S213 FADDR is increased by 6 and the process proceeds to S214.

Next, an actual example will be described.

The processing for the character "A" will be described, but only the portions different from the first embodiment will be described.

The first and second slices are the same as in the first embodiment. Since the same data as the second slice is used for the third and fourth slices, the attribute indicator stores "11" indicating the same left side. Therefore, the judgment at s205 becomes "true", and at s206, the data is taken in from the second slice and processed as the third and fourth slices. In the same manner, the character "A" is expanded.

Here, since the font data for the same left slice such as the third and fourth slices is unnecessary and is not included in the font table, the compression rate is further increased.

Similarly, since the font data is unnecessary for the blank slice and is not included in the font table, the compression rate is further increased.

(Fifth Embodiment) In the fifth embodiment, the original diagram is subjected to an exclusive OR (XOR) process in order to further increase the compression rate.

FIG. 11 is a diagram showing an exclusive OR process, in which the second slice takes an exclusive OR with the first slice. Similarly, the Nth slice and the N-1th slice are exclusive-ORed (N> = 2). The contour figure thus differentiated is newly regarded as an original figure, and the same processing as that of the fifth embodiment is performed.
It is easy to see that this process can reduce the total number of different patterns in all character slices and increase the blank slices.

In the case of the present embodiment, as shown in FIG. 12, instead of the left slice indicator, an all-black diagram slice indicator showing that the entire slice is a print dot is provided. This means that the same slice as the previous slice is already X
This is because the left slice indicator is wasted because it is a blank slice due to the OR processing.

Therefore, in order to further increase the compression rate, an all-black indicator in which all dots are print data is employed instead.

FIG. 12 is a flow chart of the present embodiment. The difference from the fifth embodiment is that instead of processing the immediately preceding slice data as print data in s206, f for 6 bytes is used.
fh is processed as print data.

Further, the XOR for reproducing the original image is performed by expanding the outline font shown in FIG.
May be used to reproduce the original image, or each slice may be expanded while taking the XOR with the immediately preceding reproduced slice.

(Sixth Embodiment) FIG. 13 is a block diagram of a sixth embodiment in which a dot refresh type display device is adopted instead of the printing device of the first embodiment, but the font output destination is changed from the print buffer to the VRAM. Since it is the same as the first embodiment, the details will be omitted.

[0049]

As described above, the effect of the present invention is great, and the present invention enables a high compression rate for any font pattern.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a first embodiment of the present invention.

FIG. 2 is a configuration diagram of a CGROM according to the first embodiment.

FIG. 3 is a diagram showing a bit configuration of a character “A” and a reference number indicator.

FIG. 4 is a flowchart of control according to the first embodiment.

FIG. 5 is a diagram showing a data structure of a character “A” in CGROM.

FIG. 6 is a CGROM configuration diagram of the second embodiment.

FIG. 7 is a diagram showing an example of a slice attribute indicator according to the fourth embodiment.

FIG. 8 is a diagram showing an example of a slice attribute indicator according to the fourth embodiment.

FIG. 9 is a flowchart of control according to the fourth embodiment.

FIG. 10 is a diagram showing a bit configuration and an attribute indicator of the character “A” according to the fourth embodiment.

FIG. 11 is a diagram showing an XOR process and an attribute indicator of the fifth embodiment.

FIG. 12 is a diagram showing an example of a slice attribute indicator according to the fifth embodiment.

FIG. 13 is a flowchart of control according to the fifth embodiment.

FIG. 14 is a configuration diagram of a sixth embodiment.

[Explanation of symbols]

 1, 11 CPU 2, 12 RAM 3, 13 ROM 4 Printing device 5 Printing buffer 6, 16 CGROM 7 System bus 14 Display device 15 VRAM

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Internal reference number FI technical display location G09G 5/22 8121-5G

Claims (8)

[Claims] 1. A character group in which one character is composed of a bit pattern of M rows × N columns is divided into submatrices of M1 rows × N1 columns for each character, and N times (N
Is a number of 2 or more), and a number of coded sub-matrixes that number the ones that appear more than once, and M rows divided into M1 rows × N1 columns ×
A pattern compression method characterized by having a code number indicator showing a coded sub-matrix in a bitmap of N columns. 2. The pattern compression method according to claim 1, wherein the code number indicator stores one of binary codes when the sub-matrix is coded and the other when the sub-matrix is not coded. . 3. The pattern compression scheme of claim 1, wherein the sub-matrix comprises a blank indicator that indicates that it contains no printed dots. 4. The pattern compression method according to claim 1, wherein the sub-matrix includes an all-black indicator indicating that all dots are print dots. 5. The pattern compression method according to claim 1, wherein the sub-matrix includes an indicator on the left side that indicates the same pattern as the immediately preceding sub-matrix. 6. The pattern compression method according to claim 1, wherein, prior to the submatrix formation, the M1 row × N1 column submatrix is subjected to an exclusive OR operation with an adjacent submatrix and differentiated. 7. The pattern compression system according to claim 1, further comprising a printing device for printing the font pattern compressed and expanded by the pattern compression system. 8. The pattern compression system according to claim 1, further comprising a display device for printing the font pattern compressed and expanded by the pattern compression system.