JPS6022876A - Picture coding system - Google Patents

Abstract

PURPOSE:To realize the picture coding system with good compressing rate and high efficiency by dividing a read picture into plural blocks and using a different coding system depending on the content of picture. CONSTITUTION:An original picture signal from a reader 10 is applied to a storage section 20 having a storage capacity of plural number of lines, where the signal is divided into plural blocks of picture information and outputted sequentially at each block. This output is applied in parallel to an MH (modified Haffmann) coding section 50 and a comparator section 30, in which the output is compared with a character from a character generator 40. When the block information is coincident with any character from the character generator 40, a character coding confirming signal C0 is outputted to a 2-byte coding section 60 to obtain 2-byte JIS Kanji (Chinese characters) codes and also the MH coding section 50 is made inactive. When this comparison is dissident, the MH coding section 50 is activated and the block picture information from the storage section 20 is coded by the MH system.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image encoding method used in facsimile machines and the like, and particularly to a method for efficiently encoding a read image in correspondence with the image content.

[Prior art]

Conventionally, the encoding method used is the Modified Huffman method (hereinafter referred to as MH method) used for facsimile communication.
Alternatively, redundancy suppression encoding methods such as the modified read method (MH method), the JISi code method used in ordinary word processors and computers, the Ebisudic method (EBCDIC method), and the ASCII method (ASCI method) are used.
There were code systems such as I system), JIS 3 unit code system, and BCD standard exchange code system. As is well known, the redundancy reduction encoding method recognizes the image to be read as simply a combination of white or black pixels without distinguishing between characters or figures, and adjusts the redundancy of the white or black pixel combination as appropriate. Moreover, the above code system is a system in which various characters such as kanji, alphanumeric characters, special characters, kana characters, etc. are written in a number of pitles of about 2 by 1 -, and for character representation, IJ: This is an optimal encoding method with good compressibility.

However, the encoding method provided in each conventional push device is limited to either the redundancy suppression encoding method or the encoding method described above, and a different optimal encoding method is selected depending on the content of the image read. There was nothing in use.

[Object of the Invention] This invention has been made in view of the above-mentioned actual situation, and by encoding using different optimal encoding methods depending on the content of the read image, it is possible to achieve high efficiency with a good compression rate. The purpose is to provide an image encoding method for

], Structure of the Invention This invention divides an image to be read into a plurality of blocks by grid-like division so that each block corresponds to one character, and the image of each divided block corresponds to a preset character. (Kanji, alphanumeric characters, special characters, katakana, 0,
If it is determined to be the character information, the JIS kanji code system, BCD standard exchange code system, ASCII code method, JIS
3. Encoding the image for each block using a coding method such as a unit code method or Ebisudic method, and if it is determined that the image information is other than the character information, redundancy such as a modified Huffman method is performed. The image is encoded for each block using a suppression encoding method.

〔Example〕

Hereinafter, the present invention will be described in detail based on a first embodiment shown in the accompanying drawings.

FIG. 1 shows an embodiment of the present invention, in which the present invention is applied to a facsimile machine.

In FIG. 1, a reading device 10 includes, for example, an image sensor and a sensor driver that drives and controls the image sensor, and reads an image of a transmitted document. The original image signal read by the reading device 10 is input to the storage section 20 . Here, the storage unit 20 is different from a line buffer used in a normal facsimile machine, and its storage capacity corresponds to that of a plurality of line buffers. That is, in this case, the image reading by the reading device 10 is normal line scanning, but the storage unit 20 temporarily stores the line-scanned image information for a plurality of lines, and stores the temporarily stored image information for the plurality of lines. Based on the readout control of the control unit 80, the image information is further divided into a plurality of blocks of image information in the main scanning direction, and each block is sequentially output. For example, as shown in FIGS. 2 and 3, if the original to be sent is
When a pixel consists of one pixel in the sub-scanning direction and is further divided into a plurality of block image information with one pixel in the main scanning direction and n block image information in the sub-scanning direction as a basic unit, the memory The section 20 is output from the reading device 10 (kx
n) pieces of image information are temporarily stored, and the temporarily stored image information is stored as (mxn) according to the control of the control unit 80.
The block image information is read out in units of block image information made up of 41 pixels.

The output of the storage unit 20 is sent to the comparison unit 30 and the MII encoding unit 50.
is output in parallel. The comparison section 30 compares the block image information outputted from the storage section 20 and the character information outputted from the character generator 40.

Since the character generator 40 is assumed to be a facsimile machine used in Japan, the character generator 40 is JIS
Each character corresponds to the first level kanji, and the name character is the same as the block image information m
Suppose that it has a dot matrix configuration of n. The comparison unit 30 performs the above comparison, and if the block image information matches any character output from the character generator 40, it compares the character encoding confirmation signal C8 with the logical value "1" with the block image information. 2-byte code converter 6
0, and also outputs an image information encoding confirmation signal C with a logical value of 10'' to the MH encoding unit 50. As a result, the 2-byte code conversion unit 60 becomes active, and converts the block image information into a 2-byte J code corresponding to the information.
The MH encoding unit 50 is inactive and does not operate. On the other hand, when the comparing unit 30 compares, if the above comparison does not match, the comparing unit 30 converts the constant signal co into a 2-byte code encoding unit 60ζ At the same time, the image information encoding of 1-1 j is determined by the logical value 1. Bow C1 M H
It is output to the encoding section 50. As a result, the 2-byte code converter 60 becomes inactive and does not perform 1rij7 operation, but
The MH code conversion unit 50 becomes active and writes j, ='y unit 2.
Block 1i output from 0! Ii image information is encoded using the MH method. 6. Comparison; 11X30 comparison! The IJ has the flexibility of adjusting the number of times as appropriate, and does not guarantee complete agreement in these comparisons.

Next, 2 Paitoko! Since the parsing unit 60 used this 2-byte JIS kanji code system, the top hit is always 1iIiii J'J! (lRight1()1) However, even if other Kanji code conversion methods are used, it is preferable to always set the most significant bit to the logical value "0".Because in this embodiment, MH encoding Part 50
During encoding conversion, the top focus is always set as a logical value.
1" to distinguish it from the code conversion result in the 2-byte code converter 60. Due to the difference in these most significant bits, for example, a facsimile machine that receives the information can easily identify which method should be used for decoding. When the 2-byte coder 60 completes the encoding process for one block, it generates a character encoding completion signal e indicating this fact. is sent to the control unit 80.

Next, the MH encoding section 50 encodes the block image information inputted from the storage section 20 using the normal modified huffing method. The most significant bit of the encoded string is always set to a logical value of 111. Of course, in the normal MH encoding system, the most significant bit may change to 10-1 or [1-1 depending on the white or black run length 1, but as mentioned above, if the most significant bit is If you add 1 bit, it will be:l! 11 will disappear. Note that a 24x24 or 32x32 bit matrix is usually required to identify kanji. Therefore, for example, assuming that the block image information is composed of 32 x 32 bits, the total number of bits per block is 1024 bits, which is less than the number of bits of one line of A4 size, 1728 bits, and ± When encoding Hc block image information, the encoding process can be performed without making any changes to the normal MW encoding method. When the MII encoding unit 5o completes the encoding process for the I block I, it inputs an image information encoding completion signal e1 indicating this to the control unit 80.

In this way, the control unit 8o receives the character information encoding completion signal eo from either the 2-byte coder 6o or the MH encoder 5o each time the encoding of one block is completed.
Alternatively, image information encoding completion No. 111 e is input, and the signal e. , el, the control unit 80 sends an encoding end signal g instructing the storage unit 2o to transfer the image signal of the next block.
At the same time, a 2-pitch code section 6 is sent to the switching section 7o.
A selection signal S is sent out indicating which of the coded information coded by o or the coded information coded by the MH encoder 5° is to be selected. As a result, the storage unit 20 transfers the block image information of the next block to the comparison unit 3° and the MH encoding unit 5.
The switching section 7o selects one of the above encoded information based on the selection signal S, and transmits the selected encoded information to the communication line 1oo via the transmission section 9o.

Then, the control unit 8o performs sub-scanning for 9470 minutes (nX in FIG.
When it is recognized that the encoding process for one area of k) has been completed, the image information for the next step read by the reading device f&10 is input into the storage unit 2o, and the same process is performed using the block image information as a unit. Let's do it.

Note that the manner in which signals are exchanged, particularly in the manner in which control signals are exchanged, shown in the embodiment of FIG. 1 is merely an example, and it goes without saying that the apparatus of this embodiment may be operated in other manners. Further, in the above embodiment, the reading device 10 is caused to perform line scanning,
Although the block image information is obtained by dividing the image read in the storage unit 20 into a plurality of block image information and reading them out, the reading device 10 is replaced by a photoelectric conversion element corresponding to each pixel of the single block. Consists of a matrix,
The block image information may be obtained by the reading device 10. In this way, the storage capacity of the storage unit 20 can be reduced.

In addition, when realizing facsimile communication using a facsimile transmitter as shown in the above embodiment, the facsimile receiver is also equipped with two different decoding units corresponding to the encoding unit on the transmitting side, and By appropriately switching and operating according to the content of block image information, facsimile communication using the encoding method of the present invention can be easily realized.

Furthermore, in this embodiment, since the application to facsimile machines in Japan was considered, the 2-byte code converter 60
We adopted a 2-byte commercial code system such as the JIS letter code system. However, when exporting a facsimile device to a foreign country, it is advisable to adopt an ASCII code system consisting of 7 to 9 bits, a JIS 8 unit code system, an Ebisu disk system, or a BCD standard exchange system for the encoding section. good. Of course, when encoding character information, an original encoding method that is completely different from these methods may be adopted.

Now, the encoding method of the present invention is particularly useful when transmitting by facsimile a document recorded on a word processor, typewriter, etc., or when a word processor and a facsimile device are directly connected and an image edited on the screen of a word processor is transmitted as is by facsimile. to demonstrate its effectiveness. This is because, in the above case, the format or font of the original image is unified, and the encoding unit of the present invention operates effectively.

In addition, the signals used in word processors include carriage return information that indicates the next recording line and the return of the carriage. For example, if a completely white line occurs in a scanned image, this carriage return information can be used to improve Compression rate encoding can be realized.

Furthermore, the present invention is not applied only to facsimile machines. For example, if the present invention is used in an input/output control device for a large-capacity storage device, information can be stored with a high compression ratio, and the storage area can be used effectively.

[Effects of the Invention] As explained above, according to the image encoding method according to the present invention, character information indicating characters and information other than the character information are encoded using different optimal encoding methods. Since this is done, it is possible to realize highly efficient encoding with good compression ratio even for a document image containing a mixture of characters and figures.

Therefore, the scope of the present invention is wide-ranging, including facsimile machines and data storage input/output control devices.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention, and FIGS. 2 and 3 are explanatory diagrams for explaining the operation of FIG. 1. DESCRIPTION OF SYMBOLS 10... Reading device, 20... Storage part, 30... Comparison part, 40... Character 5 generator, 50...
・MH encoding unit, 60...2-byte code encoding unit, 70
... switching section, 80... control section, 90... transmission section,
100...Communication line.

Claims (1)

[Claims] (1) The image to be read is divided into a plurality of blocks, each of which corresponds to one character, and the image of each divided block is character information indicating a preset character or information other than the character information. determine whether
An image encoding method, characterized in that the image for each block is sequentially encoded using different Ml and a second encoding method based on the determination. (2) The first encoding method for the character information is JI
The encoding method is S-kanji code system, BCD standard exchange code, ASCII code, JIS 8 unit code, or Ebisudik, etc., and the second encoding method for information other than the above-mentioned character information is redundancy suppression such as Modified Huffman method. Image encoding method according to claim 1.