JPH056831B2 - - Google Patents

Description

[Detailed Description of the Invention] (Industrial Application Field) This invention is a COM (Computer Output
This invention relates to a method for compressing form information used in microfilmer devices.

(Prior Art) A COM device is a device that combines fluctuating information such as income and expenditure items entered in a commercial ledger with fixed information such as a form such as the ruled lines of a commercial ledger and records it on a recording medium. There are (for example, literature:
Micrograph '83-June to September issue, Japanese Patent Application Laid-Open No. 1983-67722).

In this recording method, variable information recorded on a recording medium such as a tape is temporarily displayed on a cathode ray tube display screen through computer processing, and this display image and fixed information recorded on a form slide are optically read. A method of superimposing a form slide image and recording it on a single frame of microfilm, or an image signal of variable information obtained by computer processing and an image signal of fixed information obtained by scanning a form slide with a laser beam. There is a method of synthesizing the images and recording them on a recording medium.

However, since these methods read the form slide optically, they have the disadvantage that positional deviation of the recorded image, distortion of the image, etc. are likely to occur due to the influence of external or internal vibrations or other causes.

As a way to avoid these drawbacks, instead of reading the information directly from the form slide when recording with a COM device, form information is stored in memory as an electrical signal in advance, and when recording with a COM device, the form information is stored in memory as an electrical signal. One possible method is to read fixed information.

(Problem to be Solved by the Invention) However, the recorded image used in a COM device is usually at least approximately 2000x per frame.
Since it has more than 2500 resolution points, it is difficult to prepare a memory with memory locations corresponding to this number of resolution points, and if memory is provided at all corresponding locations, the device will become large and the structure will be complicated. It has the disadvantage of becoming

By the way, form information generally consists of only framework information such as ruled lines and other fixed information such as customer name and code number that should be recorded in fixed locations, so the information that should be recorded is There is little information, and most of it is redundant information (blank information). Therefore, when storing this form information in memory, the information is compressed and stored, and these memories are used in a COM device to expand the information and record a composite image of the above-mentioned variation information and form information. For example, since a small capacity memory can be used, it is expected that many advantages will be obtained in terms of miniaturization of the COM device and manufacturing cost.

As methods for compressing this information, there are run-length methods, plane coding methods, change-point address coding methods, and other methods that have been commonly used in the communication technology field, but all of them are complicated and require processing. The disadvantage is that it is slow. In particular, in the case of COM devices, it is necessary to process each pixel in a maximum of about 30 ns, but with the run-length method, there is enough time to compress and decompress within such a short time. The disadvantage is that there is no

In view of the shortcomings of the conventional compression methods, an object of the present invention is to provide a method that allows compression processing to be performed simply and quickly.
Moreover, it is an object of the present invention to provide a form information compression method that compresses compressed information in a form that allows for easy decompression of compressed information in a short time.

(Means for Solving the Problems) In order to achieve the object of the present invention, according to the present invention, a raster address is given to each raster in the raster scanning order, and the raster is Compare the information patterns, and if they match, change the raster address of the current raster to the raster address of the matching higher raster, and if they do not match, store the raster address of the current raster in the raster address memory, and further if they do not match. In this case, the raster information pattern of the current raster is stored in the raster information memory, the aforementioned raster information pattern from the raster information memory is divided into multiple units each consisting of multiple bits, a unit address is given to each unit, and the unit is connected to the raster information pattern. The unit information pattern is compared with the previous unit, and if they match, the matching code is stored in the unit memory, and if they do not match, the mismatching code is stored in the unit memory, and if they do not match, the unit information pattern of the current unit is executed. Store it in the actual pattern memory as a pattern.

The present invention provides a form information compression method characterized by the following.

(Operation) The operation of the method of the present invention will be explained with reference to FIGS. 2 to 4. First, raster addresses A, B, C, ... J, ... Z... are assigned from the higher raster (Figure 2), and according to this raster address order, a certain raster (this raster) and the higher raster are assigned. Compare the raster information patterns with the raster, and if there is a matching (same) pattern, change the raster address of this raster to the same upper raster address, and then change the raster address of the raster to the same upper raster address, and then If the raster addresses are not the same, the raster addresses of the raster are stored as they are in the respective raster address memories (FIG. 3A), and redundancy in the raster ranking direction is compressed. In this way, by specifying a raster address, if the specified raster matches the raster information pattern of the upper raster, the matching raster address will be called, and if they do not match, the address of the current raster will be changed to the raster address. Can be recalled from memory.

Next, the raster addresses A, B, D, . . . are stored as mismatched rasters in the raster address memory.
For each raster of J, .
1, U2, U3, U4, etc. (Figure 3B). Next, the unit information patterns of each unit and the unit immediately before it are compared, and if the information pattern of this unit matches the information pattern of the immediately previous unit, the information pattern of this unit is replaced with the matching code S. Then, if they do not match, they are replaced with the mismatch code N (FIG. 3C), and these results are stored in the unit memory. Then, the units assigned the mismatch code N (in this example, unit addresses U1, U4, U
Only the actual patterns of the unit information patterns (5, . . . ) are stored in the actual pattern memory (FIG. 4).
In this way, redundancy can be compressed in the unit order direction within the same raster.

As described above, according to the method of the present invention, redundancy compression in the raster order direction and unit order direction can be performed more simply and easily than in the past.

In addition, when decompressing information using the obtained raster address memory, unit address memory, and actual pattern memory, when a raster address is specified, the raster address or the upper raster address with the same raster information pattern is called from the raster address memory. Then, according to this raster address, a raster with compressed unit information corresponding to that address is called from the unit memory, and an actual pattern is called from the actual pattern memory according to the mismatch code N of this unit information. In the case of a matching code S, the immediately preceding real pattern can be output as it is as the real pattern of the unit. Therefore, the decompression process can be performed in a shorter time than conventionally.

(Example) An example of the present invention will be described below with reference to the drawings.

FIG. 1 is an operational flowchart for explaining the method of the present invention, FIG. 5 is a diagram showing an example of an apparatus for carrying out the invention, and FIG. 6 is a diagram for explaining the invention. , FIG. 7 is a diagram for explaining an example of a device for decompressing information compressed according to the present invention.

It should be noted that each of the above-mentioned drawings shows the main flow and means, and it goes without saying that the present invention is not limited to this example.

First, this method is started in step 10, and the form information for each raster of one screen (frame) 50 formed by, for example, 1980 rasters and 2596 dots is programmed in advance. input by operating the keyboard 52 (step 12).

The raster address assigning means 56 sequentially assigns raster addresses A, B, C, . . .
……Give Z…… (Figure 2 and Step 16).

Next, information on these rasters is stored in the line memory 58.
The raster data is temporarily stored in the line memory 58, and the rasters are sequentially read from the upper to lower rasters, and the raster information pattern of the lower raster is compared with the raster information pattern of the higher raster. In this case, the raster address assignment process and the raster information pattern comparison process can be performed sequentially or in parallel.

As an example, this comparison is performed as follows. First, address raster information of the first address A is read out in parallel from the line memory 58 and temporarily stored in the raster information memory 62 via the first coincidence detection means 60, and the address A is read out in the address reading means 64. Then, it is read out and stored in the raster address memory 66.

Next, the raster information at address B is read from the line memory 58, and the raster information memory 62
The line information of the address A is read from the address A, and the patterns of both line information are compared to see if they match (are the same) in the first match detection means 60 (step
20). If they are the same, the address changing means 6
8, the raster address B of the raster is changed to the upper raster address A (step 22) and the raster information pattern is erased, and then the address reading means 64 writes an address to the location of the raster address B in the raster address memory 66. Memorize A (step 26). If both raster information patterns do not match (are not the same), the first match detection means 6
Address B is read from 0 by the address reading means 64 and stored in the raster address memory 66 (step 24), and the raster information pattern is stored as it is in the raster information memory 62 (step 27). In this embodiment, as shown in FIG. 3A, the case is shown in which they are not the same.

Next, the information of the address C is read from the line memory 58, and the first coincidence detection means 60 compares it with the information pattern of the rasters of the upper addresses A and B stored in the raster information memory 62. For example, the information pattern of address A
If the information pattern does not match the information pattern of address B but matches the information pattern of address B, the address C is changed to the same upper raster address B and stored in the raster address memory 66 in the same manner as described above. If it does not match the information pattern of all the upper addresses, the address is stored in the raster address memory 66, and the raster information pattern is stored as it is in the raster information memory 62 (step
18-26).

In this way, the lower raster information is sequentially read out from the line memory 58 and the information patterns are compared between each of these rasters and the upper raster stored in the raster information memory 62, and if they match, then The raster address memory 66 stores the changed address of the raster, and if they do not match, the raster addresses of the raster are stored sequentially, and if they do not match, the raster information pattern of the raster is stored in the raster information memory 66. For example, sequential addresses as shown in FIG. 3A are stored in the raster address memory 66 (steps 18 to 26). In the embodiment shown in FIG. 3A, the raster information patterns of addresses A, B, D, G, J, .
indicates that the raster information pattern matches the raster information pattern of address D, address F matches address B, and addresses H and I match address G.

Next, for each raster information pattern stored in the raster information memory 62, the unit address assigning means 70 assigns it to n bits, for example.
The data is divided into a plurality of units by 16 bits each, and unit addresses U1, U2, U3, U4, etc. are sequentially assigned to each unit (step 28). Each raster address A, B, . . . J, . . . in which the raster information pattern is stored in FIG.
The unit address of each unit in the raster is shown as an example.

The unit information pattern to which these raster addresses and unit addresses are assigned is temporarily stored in the unit information memory 72 (step 30).

Subsequently, unit information patterns are sequentially read out from the unit information memory 72 for each raster starting from the upper unit, and the second coincidence detection means 74 reads out the unit information patterns for each raster from the unit in question and the unit immediately before it.
The unit information patterns are compared to detect whether or not the unit information patterns match (step 32).

An example of this comparison of unit information patterns will be described. First, each unit information pattern for each raster stored in the unit memory 72 is sent to the second coincidence detection means 74 for comparing unit information patterns in the order from the upper unit address to the lower unit address. The coincidence detecting means 74 compares the information pattern of the unit address sent one after another with the information pattern of the unit address sent immediately before. In this case, the coincidence detection means 74 always adds a mismatch code N, for example "1", as compressed information to the first unit information pattern of each raster, that is, the unit address U1 (see FIG. 3C). ), this is stored in the corresponding address location of the unit memory 76, and
For example, the actual pattern is as shown in Figure 4.
0011100110000111 is stored in the actual pattern memory 78 (steps 34, 38, 41, 42).

Next, when the unit information pattern of the second address U2 is sent, the second coincidence detection means 7
4, the information pattern of this unit is stored at the address U stored in the actual pattern memory 78.
Compare with information pattern 1. As a result, if the two patterns match (step 34), a match code S, for example "0", is assigned to the unit as compression information (step 36), and if they do not match, a match code "1" is assigned to the unit. (Step 38), these "0" or "1" information are stored in the unit memory 76.
If there is a mismatch, the actual pattern of the unit is stored in the actual pattern memory 78.

In this way, the information patterns of the unit sent sequentially and the unit immediately before it are compared, and the match or mismatch codes are stored in the unit memory 76 from the upper unit address U1 to the lower unit addresses U2 and U3. , . . . and stored in the corresponding positions (steps 34 to 40). This sign “1”
For example, FIG. 6 shows a state in which "0" is added. In this example, the unit addresses U1, U
The information patterns of 4, U5, . .

FIG. 4 shows an example of an actual pattern in the case of mismatch.

In this way, the above-described processing is performed on all units of all the raster information stored in the raster information memory 62, and the compression processing is completed (step 44).

Note that the above-mentioned processing of storing the raster information pattern in the raster information memory 62, processing of assigning a unit address in the unit address assigning means 70, processing of detecting a match/mismatch in the second coincidence detecting means 74, and processing in the unit memory 76 and Actual pattern memory 7
8 can be performed in parallel.

In addition, each of the above-mentioned means 56, 60, 64, 68,
70, 72, 74 as well as memories 58, 6
2, 66, 76, 78 and other components can be easily constructed using commonly used electronic elements and parts.

Next, a case in which compressed form slide information is expanded using the raster address memory 66, unit memory 76, and actual pattern memory 78 formed in this manner will be described with reference to FIG.

FIG. 7 is a diagram schematically showing a circuit configuration of a portion that decompresses information used when actually writing form information on a recording medium such as a microfiche in a COM device. Each memory 66, 76, 78 is provided with read means 80, 82, 84 for accessing the respective read signal.
For example, if the raster address input of each raster is supplied to the reading means 80 in synchronization with the reading of variable information from the computer, the raster address of the concerned raster called up by the reading means 80 or the raster that has been changed can be The address is recalled from memory 66. The new raster address read from this memory 66 is supplied to the next reading means 82, and this reading means 82
Then, the unit memory 82 is accessed and compressed information "1" and "0" of each unit from the upper to the lower part of the raster are read out in association with the raster address and the unit address. Next, the read compressed information "1" and "0" are supplied to the reading means 84, and from this reading means 84, the actual pattern memory 7
8, and if the compressed information is "1", the actual pattern of the corresponding unit is read out and sent to the output means 86, from which the information of this actual pattern is output. On the other hand, if the compressed information is "0", this signal is sent to the output means 86 to output the same unit information pattern as the information pattern of the unit immediately before the unit in question. By operating the COM device using these outputs obtained in this way, form information such as ruled lines and customer items, account items, and other variable information are combined and recorded. I can do it.
In the above-described embodiment, the raster information of the form is input using the keyboard, but it is also possible to record it on a form slide, read the form slide with a light beam such as a laser, and convert each raster into a digital code. You can also use it as

(Effects of the Invention) As is clear from the above description, according to the form information compression method of the present invention, compression in the raster order direction is first performed using the raster information pattern, and then compression in the unit order direction is performed using the unit information pattern. Since it is performed using patterns, compression can be performed more easily and accurately than with conventional methods, and the compression ratio can be increased.

Furthermore, the COM device can be configured to compress the form information and temporarily store it in memory, and then read the form information from the memory when actually recording the information, which simplifies the configuration of the device. Manufacturing costs are also low.

Since a memory that stores such compressed information is used, information decompression processing when writing form information is easier than in the conventional case.
Furthermore, the time required for information expansion can be significantly shortened compared to the conventional case; for example, information for one pixel can be written within 30 ns.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing an operation flowchart for explaining the form information compression method of the present invention, FIGS. 2, 3A to 3C, and 4 are diagrams for explaining the method of the present invention, FIG. 5 is a diagram showing an example of an apparatus for implementing the method of the present invention, FIG. 6 is a diagram for explaining the invention, and FIG. 7 is a diagram showing the decompression of information compressed by the method of the present invention. FIG. 50...Screen, 52...Photoelectric conversion device, 54...
...A/D converter, 56... Raster address assigning means, 58... Line memory, 60... First coincidence detection means, 62... Raster information memory, 64...
Address reading means, 66...Raster address memory, 68...Address changing means, 70...Unit address assigning means, 72...Unit information memory, 74...Second coincidence detection means, 76...Unit memory, 78 ...actual pattern memory, 8
0, 82, 84...reading means, 86...output means.

Claims (1)

[Scope of Claims] 1. In a form information compression method in which form information is stored in a memory for each raster, a raster address is given to each raster in the raster scanning order, and a raster information pattern is created between the current raster and higher-order rasters. If they match, the raster address of the current raster is changed to the raster address of the matching higher raster, and if they do not match, the raster address of the current raster is stored in the raster address memory, and the raster addresses match. If not, divide the raster information pattern of the current raster into multiple units consisting of multiple bits, give each unit a unit address, compare the unit information patterns of this unit and the previous unit, and if they match, it is determined that they are a match. A form information compression method characterized by storing a code in a unit memory, and in a case of a mismatch, storing a mismatch code in a unit memory, and further storing a unit information pattern of the unit as a real pattern in a real pattern memory in a case of a mismatch. 2. The form information compression method according to claim 1, wherein the raster information pattern of the first raster in the raster scanning order is always stored in the raster information memory. 3. The form information compression method according to claim 1, wherein a mismatch code is always given to the first unit of each raster stored in the raster information memory. 4. Claim 1, wherein the raster address memory, unit memory, and actual pattern memory are memories used for decompression processing.
Form information compression method described in section.