JPS5838071A - Vertical and horizontal conversion data compressing system - Google Patents

Abstract

PURPOSE:To eliminate dependence to scanning direction by comparing original picture data with vertical-horizontal conversion data in data compression of picture data and transmitting and restoring using data of smaller data quantity and vertical-horizontal conversion designating information. CONSTITUTION:After making vertical-horizontal conversion on a picture data memory 2 so that picture data disposed in main scanning direction X and subsidiary scanning direction Y are disposed in subsidiary scanning direction Y and main scanning direction X, compressed data are compared with original picture data which are not converted vertically and horizontally. A sending side 30 transmits data of smaller data quantity from a compressed data memory 7 using information that designates compressed data of smaller data quantity and designates whether vertical-horizontal convertion is executed or not. A receiving side 50 outputs the restored picture data on a printing device 37. Thus, two pictures are formed by rotating by 90 deg. main scanning direction X without rotating physically the direction and smaller compressed data are used. Accordingly, dependence to the main scanning direction X is eliminated.

Description

[Detailed description of the invention] This invention @Hiro Ryozo Deme data compression), picture data o
This relates to a data compression method in which ti = prime array t - vertical and horizontal conversion is performed, the amount of data before and after the conversion is reduced, and the smaller i is used as vhh = 5.

When compressed data is created using a conventional 77 comb device, the compression ratio of the compressed data differs depending on the relationship between the writing direction of the document text and the main scanning direction. That is, the compression ratio of compressed data depends on the main scanning direction of the image data, and the compression ratio becomes larger when the main scanning direction is parallel to the book frame direction and when it is perpendicular.

Figure 1 (!l), (b) Hiroko's relationship! Each one is written vertically! ! Manuscript and horizontal writing 11J [1g4 manuscript
Shown clearly.

In the same figure (1), one vertically written character string 20t to 20s is arranged on a document 20, and the main scanning direction X# is perpendicular to this, and all
If l scanning direction difference is swiss, then main scanning line jl*jl
. . . In the area section ■~@, @~O2@~@ in the aY direction, the change in iii speed (■-■, etc.) is shown in Table 1, and it becomes p.

In contrast, in the first print (-), horizontal writing - character lines 201 to 20'
If s is arranged, similar main scanning III Jt e
Im .

In the 1s 2 table data compression method, encoding is usually performed by focusing on changed pixels in the main scanning line, so the smaller the number of changed pixels, the higher the compression ratio.

Comparing Table 1 and Table 112, even though the original patterns are the same as shown in Figure 111 (a) and φ), Table 2 has fewer pixels that change in the main scanning direction. It can be seen that if you combine this with the direction of *C, the compression ratio becomes very hot. For example, when compressed using the MR method, the compressed data in FIG.

Therefore, it would be better to read the original by turning it 90 degrees as shown in Figure 1 (the same EΦ in the case of the → original), but in the case of a normal 77-comb device, the reading pt size and recording of the original Since there are restrictions on the size of the image, it is restricted to physically rotate it, and it is limited to objects that are smaller than the maximum overcurrent reading size. For example, in a facsimile machine that can read up to A4, the short side of an A4 document is taken in the main scanning direction.
It is difficult to spread the image with this long side in the main scanning direction.

Fi that can be read with the long side in the main scanning direction
The size will be A5 or smaller.

In the present invention, two images are created by rotating the main scanning direction by 90 degrees using memory without using a rotation wheel, and the i image with less compressed data is used. Dependency disappears.

An object of the present invention is to provide a data compression method in which p-compressed data does not depend on the main scanning direction by vertical-horizontal conversion of image data.

In order to achieve the above objectives, the image data arranged in the main scanning direction and the sub-scanning direction are arranged in the main scanning direction and in the main scanning direction, respectively, using the horizontal and vertical conversion data compression method of the present invention. Comparing the compressed data after vertical and horizontal conversion and the compressed data obtained by compressing the original image data after vertical and horizontal conversion, the amount of data is small h70EIII
It is transmitted using data and information specifying whether or not it is 6h after vertical/horizontal conversion.

The feature is that it performs restoration.

The present invention will be described in detail below.

No. 2511. The 3rd ml was filled with Il1, which shows the composition of Example 011 of the present invention. 211 is a transmitting section, and FIG. 3 is a receiving section 6j, 114111g211. Main part of Russia Detail 1
14WA, Iris 5flIQ, and Rattan 6kaku are the explanation of the operation in Figures 2 and 5 glya hh0 11 The transmission part oyu ivy diagram when the image data for one sheet of 2W11a manuscript is held in Kanabe Memine. %I! at Toh0 transmitter At 311140, the original 20 is converted into image data by a CCD (charge coupled device) 21''t'im*n.

When the drive circuit 22 sends the Ii image data to the data EIIa*SO, it adds exactly one pulse of 1jii elements to the address variable 1AiiiI as an m** period signal. Address change @il bribe 1 is 114IIO detailed IAEllc tyo 5
Address control to count 1c, 9iy@/12
, an address cliff ypls that counts 1 lie/medium O bulk addresses, a comparison circle 14 that compares the number of lies y with the final lie:/* 1B, and a ratio soft that compares the number of pixels with the final number of pixels 17. circuit 16 and address control (
9) Route 11 follows. ζ Here, the manuscript is to be printed using the short side of the 40 enlarged 4th edition manuscript as the scanning line. scanning**
Is your degree CCIT? %fh passed), main scanning line 172
8 pixels/line (8 pixels/-), sub-scanning: 112,174 twins/m4 originals.

The address change* circuit 1 writes the image data of the autumn equinox of the document 1 into the image data memo 92, and the image data reader 2 operates the address to change the main scanning of the image data from the short side of the document to the long side. (This conversion will hereinafter be referred to as vertical/horizontal conversion).

ζHere, write the image data 412, I! Regarding the output [I will quote Figure 4 and Figure 1/s5 for an explanation.

First, when writing image data file 4v20, III4mr
ic>n, the address control circuit 11 writes the image data pixel by pixel into the sequential image data 412 each time a pixel synchronization signal is input to the drive circuit 22. The address value bit of both image data is specified by the address counter 12, and the lower bit is specified by the address input/reference 1S. The capacity of the image data memo 12 is 5JI5KB (kilopite), which is 468MW (megaword) x 1 bit O'III.

Address cuff/ri 15f nio kunda tsubu and vh heat,
Artless counter 130 content is maximum pixel count 172
8, the ratio IR circuit 16 notifies the address control circuit 11 of this. The address control circuit 1611 accepts the request and selects the address 2F? Counter 12t-1 is counted and the address counter 13 is cleared.

Hereafter, according to the above-mentioned procedure, add 150 kakuy to the address counter.
Continue touring operation. And Address Power Kun I 1
When the content of 2 softens to the final line a 2576, the ratio *
1ljl14hiro This is notified to the address control circuit 11. The address control circuit 11 receives the ζO end signal and the above-mentioned final element and signal.
e 1J 2 HetD*ig41 Autumn equinox D@ Image data po storage ends.

Hiro image day at gsFjA! A state in which the image data is written to the image data memo 2 is graphically shown. That is, the line number and pixel address are written on the pot pixel, and 16. The left half of the diagram shows the image data according to the address - image data amount - &'i12 for $
11? It is written in layers in the heart.

Next: The amount of data for both images is $1411 for Mori 20.
Address control circuit 11 Hiroto 2m control circuit 5
Receives the O read signal from the address counter 12 and sets it to 1.
The image data is revealed one pixel at a time, one by one. Address power #: Count up 12 and i
and the content of the address counter 12 is the final line number 2s7.
6 and 1 hour, soft circuit 14 mine address wmtm road 1
Let 1 know. In response to this, the address control circuit 11
The address counter 11 is counted up by one, and the address ocunter 12 is incremented by t/V. Thereafter, the count-up operation of the address counter 12 is continued according to the instructions given.

Then, the address control circuit 11 finally receives the twin aS and final jII elementary signals and outputs the image data for one document from the image data memory 2.

The state of generating the 2nd and 4th b image data 5i1 is shown in the right half of the figure. In other words, the first tweet yo1 number ao* and the second tweet yo1 qiao.

It is read out like the first one of ll5rai/.

As described above, by writing into the its data memory 20 and outputting Il, the image data is converted into an aspect ratio converter.

112 rcL), the control circuit 5 converts the image data from the drive circuit 22 into image data! While storing it in Megori 2,
The multiplexer S is switched to input power to the data compression circuit 4. The data pressure llA circuit 4 compresses this image data and transmits it through the six-wavelength converter 6t switched by the control circuit 5.
Memo the compressed data (hereinafter referred to as normal compressed data) v7-1
write to. At the same time, the control circuit 5 controls the amount of compressed denim data.
- inform.

Next, the control circuit 5 uses the image data memo v2), the vertically and horizontally converted image data (hereinafter referred to as converted image data) t-,
The multi-blephrase black is switched and inputted to the data compression circuit 4. The data compression circuit 4 compresses the converted and double image data and writes the compressed data c or less (referred to as conversion pressure - deme) into the memory 7-2 via the demultiplexer 6 switched by the control circuit 5. Control circuit 5 simultaneously controls the amount of converted and re-compressed data.
Let me know eC.

The control circuit 5 compares the data amount of the original compressed data and the converted compressed data, and determines whether the data amount is small i! The h side is memorized and transmitted from 7 through multiplexer 8. Note that data that distinguishes which compressed data is to be sent (hereinafter referred to as identification data)
is transmitted prior to sending the compressed data OfR.

Thereafter, the transmitter reads the next document) and repeats the same operation.

The memo v7 is divided into three parts b). Compressed data is read from one part and during transmission, normal compressed data and converted compressed data are written to the other two parts.

FIG. 5 is a block diagram of a receiving section corresponding to the transmitting section of FIG. 2.

In the same figure, data restoration @5oo11-circuit s2 receives either normal compressed data or converted compressed data from the identification data of the transmission data IO. When the compressed data is restored and the data is normally compressed data, the de!multiplexer 5s is switched to - of the !lunarplexer 56, and the restored image data is output to the printing device 37 via the !multiplexer 56.

In the case of converted compressed data, the control circuit 52 switches the demultiplexer 33 to the image data memory 35 side. Then, using the address conversion circuit 34 and the image data amount 55, the restored image data t-1 is subjected to the above-mentioned vertical/horizontal conversion.

FIG. 6 shows the state of vertical and horizontal inversion.

The inverse vertical/horizontal conversion is performed by the reverse operation of the horizontal/horizontal conversion shown in FIG. Address conversion circuit 54
is the 411th address conversion circuit 1 of j12iiiJ.
It is served with the following composition. Since it is performed in the same way as the vertical and horizontal inverse conversion, the explanation 114 will not be expanded upon.

The control circuit s2 converts the converted image data into the original 1III data by vertical and horizontal inverse conversion! It is outputted to the print t57 via the multi-retardant 56.

Above, we have described the method of storing all of the original 1 Autumn Equinox III data in the image data memory, but this method a (1)
The image data amount 417 requires a large capacity; (2) the image data memory stores image data for one document in ms;
! Rounding time is required for horizontal conversion.

Therefore, the question is whether to improve performance by dividing the image data of one page of original and storing it in the image data memory.

Rattan 7wA (荀, 伽) is the case when one page of manuscript is divided.
The theory of principle that compares to the case of not dividing (the case where it is not divided) is shown in the figure.

In (4) of the same figure, Jt t js e Is m ””e
jm 0iii image data is transformed into l'l by vertical/horizontal conversion
e I'm e I's # "e I'B OIf
It becomes l image data. In the same figure (b), a case is shown in which the image data is divided into N parts and stored in a message. In this case, Hiro 1st division I
Le11#・*@, Ja◆! , second division 1.1+1
m 41m44@ ”'e jam, 5th division~jlI
N division, total 1. j, 綽1 m ”'e ldl ('
The image data is transformed into j11 minutes mj each by vertical and horizontal conversion.
1°4%..., I-1 second division l-1 caple, axis,
4. Le Kaboku-~Nth division/Ill+1. J-*+m
, . . . indicates that the image data is I'40 image data.

When the image data is divided as shown in FIG. 2(b), the amount of compressed data becomes slightly larger than that shown in FIG. 2(a). When the above-mentioned M4 original is compressed using the MB method, the increment of compressed data per divided machine is 214B (not including j20L). Compressed data is usually 20KB to 50-B''
Since there is
Since the values are output, even if both image data are divided and stored in memory, the effect of vertical/horizontal conversion can be fully obtained.

If you have manuscript 1 Autumn Equinox data in memory, Iris 2 Figure 1
jII 5 ItteU 585KB (1) WAAll
f -Me/ -%V #Although it is rarely necessary, by dividing both image data, the capacity of both image data memory can be reduced.

FIGS. 8 and 9E are explanatory diagrams showing the configuration of other embodiments of the present invention. 8th (Fig. 9 is a transmitting unit, a receiving unit is shown in Fig. 9, and an irises 101a) to (d) are explanatory diagrams showing the operating procedure of Fig. 8.

FIG. 8 is a block diagram of a transmitter that divides the blank image data and sends it to both images. The transmitter 50' in the figure is different from the transmitter 50 in Figure 2 in that the amount of Q image data is now 2 (2-1-2-2)% (2) Data compression circuit (4-1, 4-2), and the compressed data memory has decreased by 4 Kt (7-1 to 7-4).

By having a single point of difference in this way, it is possible to divide the original document, read it continuously, compress it, and transmit it. Since there are two address change w & garden path 1Fi image data memories, the circuit shown in 1/E4m is for 2m path. However, the final number of lines in the iris 4 diagram is the number of twins of the divided image data.

Next, the divided image data is compressed in the receiving section.

The transmission procedure is shown in Figure 10.

Each division of %A bottle, iii image data, normal compressed data, and converted compressed data shall be indicated by a code as shown in Table 5.

In the configuration shown in ) to (b) in the same figure, two image data memories 2-1 and 2-2 are used for printing and writing, respectively, and four compressed data memories 7-1 to 7-4 are used for printing and writing. side) No.
'5I! For writing positive compression data, (2) For waiting for positive pressure stripe data, (
(4) is used to write the converted compressed data, and (4) to output the converted compressed data for transmission. The compressed data is obtained continuously with a delay of TItIIiL time for dividing the image data into two.

That is, in FIG. Pi1 (1), when PD1 is input, it is written to the private data memory 2-1, and CDI is written to the compressed data memory 7-1.

Next, in the same figure (b), when PD2 is input, PD2 is written from writing 2-1 to the image data memory 2-2, and CDI' is written to the compressed data memory 7-2. , CD2 is written to compressed data memory 7-4, and CDI 7-1 is moved to standby.

1ii3 vA(e) and dPD5 outputs image data.
-1 & CPD1, the writing and reading of the compressed data memory is advanced by one stage, CD2 of 7-4 is moved to standby, and CDI 9 is of CDI'. Small compressed data is output. In the same figure (d), PD
4 writes data to the image data memory 2-2 as a substitute for PD2, the pressure data memory advances the writing and exposure of the same figure (C) by one step, and the CDS 7-3 is set aside for standby. jl, CD
In case of 2 or 9, the smaller compressed data of CD2' is output.

The compressed data divided in this way is output continuously.

FIG. 9 is a block diagram of a receiver that divides image data and stores it in memory. The difference between the data restoration unit 50' shown in the figure and the data restoration 1iA50 shown in Fig. 186 is that there are two image data memories 11 (351, 55-2), and the restored image data is always stored in the image data memory. It is. Since two image data memories are provided, the address conversion circuit 544 has 21 circuits of 4 m length. However, the final number of lines is divided into iii*day! The number of lines is the same as the number of lines.

The operation of the data restoration section 50' is as follows. As in FIG. 5, the control circuit 32 first receives identification data and learns whether normal compressed data or converted compressed data is being sent. Next, the control circuit s2 causes the data restoration circuit 31 to restore the slightly compressed data. Then, in the case of normal compression data, the image data restored via the demultiplexer 33 is stored in the original image data memo 1Ji5, and in the case of converted compression data, it is restored and the good image data is reversely converted vertically and horizontally. ' is stored in the image data memory 35. The two image data memories 35-1 and 55-2 are used for writing and for ejecting the bladder, respectively, and are arranged in a two-byte rough format. Control 1 circuit 5
'J55 in both image data file 4 of image data for 1 division of 2 companies
-1.35-2 After writing is completed, this JII image data! Select with multiplex tsh and ii! Birth/birth printing device 57] Prints.

According to the embodiment of Fig. 8 and 9th wA, if one and both image data are divided and stored in the memory, for each portion of 1IIIliiii image data,
Since the capacity of the normal compressed data and the converted compressed data are compared and the smaller one is transmitted, the dependence of the compressed data on the scanning direction for each divided image data is eliminated. This is more advantageous than the implementation shown in FIG.

Furthermore, in each of the embodiments described above, in the embodiment where the data of both parties is divided and stored in the memory, the i! Obviously, this may be an unequal division according to a particular algorithm.

Takata, as in the embodiment, the same effect as the present invention can be obtained even if the -fiLvL original is mechanically rotated by 90 degrees and read again without electrically changing the vertical/horizontal direction. can be realized. However, in the case of Jin, the receiving and transmitting sections must be able to handle two document sizes for one document.

As explained above, according to the present invention, it is possible to eliminate dependence on the scanning direction of the original, so the compressed data is the same regardless of whether the original is written vertically or horizontally. It is natural to obtain a compression ratio of .

[Brief explanation of drawings]

Figure j11 ←), 佽) are vertical writing, explanatory diagrams of the relationship between horizontal 11 angles and data compression, Figure jg2, and 5 wJ are detailed explanatory diagrams of the present invention. Detailed explanation of main parts @ diagram,
jls diagram, 86WA is the operation explanatory diagram of Figure 2 and Figure 5 respectively, jIz diagram ℃), 佽) xm of vertical and horizontal conversion by ore division
Figures 8 and 9 are explanatory diagrams of other embodiments of the present invention; , '34 is an address conversion circuit, 2
．． 35 is both image data memory,! S, 8.34 is a multiplexing, 4-wire data compression circuit, 5.52 is a control circuit,
6.55 is a demultiplexer, 7 is a compressed data memory, 20 is a manuscript, 21 is a CCD quinsenna for disfigurement, 22
Reference numeral 31 indicates a drive circuit, 31 a data restoration circuit, and 57 a printing device. Patent issuer Fujikoshi Co., Ltd. Sub-agent Patent attorney 1) Zen Saka 1 Figure 1 (a)

Claims (1)

[Claims] The image data arranged in the final scanning direction and the sub-scanning direction is arranged in the sub-scanning direction and the main scanning direction respectively. The unconverted original image data is compared with the compressed data, and the compressed data with the smaller amount of data is used to transmit and restore using the information that determines whether or not vertical/horizontal conversion has been performed. A data compression method that performs vertical and horizontal conversion.