JPS58146171A - Character recognizing device - Google Patents

Abstract

PURPOSE:To eliminate the need for hard copy even at a failure such as a line failure, or at normal picture information reception and recognizing processing, by providing a storage means storing received picture information temporarily at each information unit and possible for readout/write. CONSTITUTION:A character recognizing device is provided with a communication control section 12, a reception buffer control section 18, a reception buffer 22, an expansion section 24, a changeover switch section 32, a recognizing section 38, a hard copy output section 44, a system control section 46, and a system bus 58. The section 22 stores the received picture information temporarily at picture information unit. The picture information stored in one unit is transferred in the section 38. When the operator instructs the output of hard copy, after the recognition of one unit of picture information is finished at the section 38, the stored one unit of picture information is transferred to a hard copy output section 44.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a character recognition device, and more particularly, to a character recognition device that receives image information including characters from a communication line in the form of a facsimile signal, and recognizes the characters included in the image information. be.

For example, when inputting image information including characters transmitted by facsimile into a computerized information processing system via an optical character reader (OCR), there is little room for manual errors to intervene in the input process. , work efficiency is also improved. As an alternative character recognition system, a character recognition device that organically combines a facsimile device and an OCR has been proposed. In this system, image information in the form of an OCR form received by facsimile is input as an electrical signal to a character recognition section, and character recognition processing is performed. Receiving image information from a communication line is possible in accordance with, for example, the International Telegraph and Telephone Consultative Committee (CCITT) Recommendation T, 30.
The transmission control procedure is performed according to the following. The received image information is usually, for example, MH (Modified Huffma).
n) Since it is a serial image signal compressed by the encoding method, it is converted into a two-electro signal representing black/white level and output as a hard copy from the recording unit, and is also simultaneously transferred to the character recognition unit. Ru. The character recognition unit is a processing/system that includes memory, a floppy disk (FD), a CRT (cathode ray tube) display, a keyboard, etc., and detects the timing mark of the form included in the transferred electronic signal and determines the inclination of the form. Character recognition processing is performed on this corrected two-way signal, and all recognition results are processed for each form regardless of the presence or absence of unrecognizable characters. The image information is stored on the floppy disk, and the process moves on to receiving the image information corresponding to the next form. Although it is common for unrecognized characters, or recognition errors, to be dealt with, that is, correction work is done all at once after the end of communication, but each time an unrecognized character is detected in the character recognition unit, a ``character'' pattern is displayed on the CRT display. Some systems are configured to prompt the operator for immediate correction via keyboard input.

Therefore, in such a system, the operator makes instant corrections based on a soft copy of the image information on the CRT display, and there is little need to refer to the hard copy output for this correction. , a hard copy is always output from the recording unit. therefore,
Hard copies of image information containing unrecognized characters before correction will pile up. Further, since hard copies are always output in this way, even if facsimile communication is interrupted due to an external factor such as a communication line failure, unnecessary hard copies are output. Furthermore, since the character recognition unit executes character recognition processing based on the two electric signals transferred at the same time as the facsimile signal is received, in the event of such communication interruption, the character recognition processing will be canceled and the It is necessary to perform an interruption process such as discarding the recognition results stored on the floppy disk.

Therefore, the present invention eliminates the need to output unnecessary note copies of image information even during normal image information reception recognition processing, or even in abnormal situations such as line failures, and eliminates the need for unnecessary copying of image information such as interruption processing. The purpose of the present invention is to provide a character recognition device that does not require processing.

This object is achieved according to the invention by a character recognition device as follows. In other words, this device includes a receiving means for receiving image information including characters from a communication line, a storage means that can read and write the received image information for each unit of the image information, and a storage means for temporarily storing the received image information in units of the image information, and a storage means for storing the characters included in the image information. A recognition means for recognizing, an output means for outputting the image information as a hard copy, and a storage means connected between the recognition means and the output means, and the image information stored in the storage means is connected to the recognition means and the output means for outputting the image information as a hard copy. The control means controls the transfer means when the storage means accumulates one unit of image information, and causes the storage means to accumulate subsequent units of image information. Transfers one unit of image information stored in the storage means to the recognition means, regardless of whether it is inside the storage means or not.
When the operator instructs to output a note copy, when the recognition means finishes recognizing one unit of image information, the transfer means is controlled and the one unit of image information stored in the storage means is outputted. Transfer to.

Next, embodiments of a character recognition device according to the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram showing an embodiment of a character recognition device according to the present invention. In the figure, the present device has a communication control unit 12 connected to a communication line 10, which has network control functions such as incoming call detection and line maintenance, and a demodulation function of signals received from the line 10. The communication line 10 may be a public line, for example, and an image signal is transmitted from a remote image terminal such as a facsimile machine or an image file at a center. This image signal is, for example, a serial signal based on the rask scanning method in which the screen is scanned by main scanning and sub-scanning, and is a serial signal using an appropriate band compression method, such as M
A facsimile signal based on the H (Modified Huffman) encoding method may be used. The communication control unit 12 is connected to the line 1
If there is an incoming call from
The reception of facsimile signals from the line 10 is controlled in accordance with Recommendation T.30. Note that the communication control section 12 may have a function of transmitting a facsimile signal from the present device to the line 10.

The communication control section 12 is connected to a reception buffer control section 18 by a data line 14 and a control line 16, and the latter is connected to a reception buffer section 22 by a data line 20. In the figure, among the data lines through which image information such as facsimile signals is transferred, those through which image information in compressed and encoded format is transferred, such as data lines 14 and 20, are indicated by hatched arrows. Receive buffer control section 1
8 corrects the inclination and lateral shift of the form based on the timing mark of the form included in the compressed image information transferred from the communication control unit 12 through the data line 14, and sends the corrected image information to the receiving buffer unit 22. It also manages the image/information storage area in the reception buffer section 22. The control section 18 also controls the reading of image information stored in the reception buffer section 22 in response to a request from the decompression section 24, as will be described later.

The reception buffer unit 22 stores compressed image information in multiple pages (
This is a large-capacity memory such as a random access memory (RAM) or a magnetic disk device, which can store data for a large number of documents. The maximum storage capacity can be designed according to the communication traffic applied to this system (for example, the number of pages of image information), but if a memory with a storage capacity of 512 KB (kilobytes) is used, main scanning according to the CCITT G standard Direction linear density: 8 dots/van, sub-scanning direction linear density: 77 dots/τ, average compression ratio by MH encoding method: 1
/7 A4 size (210X297 tone) image information (approximately 72K
B) can be accumulated for about 7 pages. Under the control of the reception buffer control section 18, the reception buffer section 22 can read accumulated image information to the reception buffer section 24 while writing image information from the communication control section 12.

A decompression unit 24 is connected to the reception buffer control unit 18 by a data line 26 in a compressed format and a control line 28, and this decompression unit 24 converts image information in a predetermined compression format such as the MH encoding system received through the data line 26 into black/white. It has the function of converting into two electric signals indicating.

A changeover switch unit 32 is connected to the expansion unit 24 via an uncompressed data line 3 and a control line 32, and a recognition unit 38 is connected to the changeover switch unit 32 via a similar data line 34 and a control line 36. , still connected to a hard copy output unit 44 via similar data lines 4o and control lines 42, respectively.

The changeover switch section 32 sends the image information received from the decompression section 24 via the data line 30 to the data a 134 and transfers it to the recognition section 38, or sends it to the data line 4o and sends it to the hard copy output section. This is a switching circuit that switches whether to transfer the data to 44 or not. This switching operation is performed under the control of the system control section 46 as described later.

Peripheral devices such as a CRT (cathode ray tube) display 48, a keyboard 50, a printer 52, and a floppy disk (FD) 54 are connected to the recognition unit 38, and the recognition unit 38 receives uncompressed image information received from the data line 34. This is a processing device that recognizes the included characters and stores the recognition results on the floppy disk 54. When an error occurs in character recognition (detection of an unrecognizable character) in the image information, the ``character'' pattern is visually displayed on the display 48 to request instructions from the operator.The operator operates the keyboard 50 to recognize the character. You can perform correction operations on invalid characters.

The hard copy output unit 44 is a plotter that visually outputs the uncompressed image information received from the data line 40 as a note copy 56.

Output control of the output unit 44 is performed by a system control unit 46 as described later.

These communication control section 12, reception buffer control section 18, decompression section 24, selector switch section 32, recognition section 38, and hard copy output section 44 are connected to a system control section 46 via a system bus 58. The system control unit 46 is a processing unit that monitors and centrally controls the operation of the entire device, and as described later, instructs each unit to operate according to the operation status of each unit via the system bus 58, and monitors the operation status of each unit. Let's do it.

FIG. 2 is a system flow diagram showing the operation of this apparatus, and the operation of this apparatus will be explained with reference to this. In FIG. 2, time is plotted in the horizontal direction, and the operating state of each part is plotted in the vertical direction, as shown at the left end of the figure. In this figure, elements that are the same as those shown in FIG. 1 are designated by the same reference numerals.

When a call is received from the communication line 10 to the apparatus as shown by the triangle in the upper left corner of FIG. 2, the communication control section 12 starts a reception operation according to a predetermined transmission control procedure, for example, T2O. When image information compressed using a predetermined band compression encoding format, for example, the MH encoding method, is received from the line 10, the communication control unit 12 demodulates it and sends it to the data line 14. The image information on the data line 14 is written into an empty area of the reception buffer section 22 via the data line 2o under the control of the reception buffer control section 18.

When image information for one page (form) is accumulated in the reception buffer section 22 in this manner, the communication control section 12 notifies the system control section 46 of the status through the system bus 58. System control unit 46 instructs decompression unit 24 and recognition unit 38 to start operation via system bus 58,
The data line 30 from the extension section 24 is connected to the changeover switch section 32.
to the data line 34 to the recognition unit 38. As a result, a transfer path for uncompressed image information from the decompression section 24 to the recognition section 38 is set.

Therefore, the system control section 46 instructs the reception buffer control section 18 to read the image information from the area in which the storage of image information for one page in the reception buffer section 22 has been completed.

In response to this, the reception buffer control section 18 reads out the compressed image information stored in the storage area of the reception buffer section 22 and transfers it to the decompression section 24, and the decompression section 24 uses this MH encoding method. The image information is decoded and converted into two electric signals representing normal black and white levels, and data #j! The recognition unit 38 via the 30° changeover switch unit 32 and the data line 34
Send to. Note that during such a read operation from the reception buffer section 22, if the communication control section 12 receives image information corresponding to the next page and subsequent pages from the line 10, the reception buffer control section 18 reads the image information from the reception buffer section in the same way as described above. This image information is written into the next free storage area at 22.

By the way, the recognition unit 38 performs recognition processing for characters included in the image information received from the data line 34. This recognition processing is sequentially performed on the image information transferred from the decompression section 24 to the recognition section 38, and the results are sequentially stored on the floppy disk 54. In this way, the reading of image information for one page from the reception buffer section 22 and the recognition process in the recognition section 38 are performed in parallel.

When an unrecognizable character is detected in the image information during recognition processing by the recognition unit 38, the image pattern of the portion that appears to be a text character table is displayed on the display 48 to request instructions from the operator. If the operator can read the unrecognized character from the contigs of the image information, he operates the keyboard 50 to correct the character. If the information cannot be read, an abnormality processing key such as a "form ejection" key or a "skip" key provided on the keyboard 50 can be operated as necessary to instruct the recognition unit 38 to interrupt the process.

In this case, the recognition unit 38 notifies the system control unit 46 that processing interruption is requested, and the system control unit 46 responds to this by informing the reception buffer control unit 18 and decompression unit 24.
instructs to stop the operation. As a result, the image information of the page containing characters that the operator cannot read, which is stored in that area of the reception buffer section 22, is discarded according to the operator's instructions, but only the trench is output as a hard copy. You can make corrections using the keyboard while looking at the screen.

When an instruction to output a hard copy is given by the operator's manual operation from the keyboard 50, the system control unit 46 receives a notification from the recognition unit 38 that the recognition process for the image information of that page has been completed, and then controls the changeover switch unit 324. data line 3
゜ is connected to the data line 40. As a result, the extension part 2
4 is connected to a hard copy output section 44. Next, the system control unit 46 instructs the reception buffer control unit 18 to sequentially read out again the image information of the page for which the recognition process has been completed, and instructs the hard copy output unit 44 to perform a hard copy creation operation. As a result, the compressed format image information of the area where the recognition process has been completed is read out again from the reception buffer section 22 to the decompression section 24, where it is converted into an uncompressed format signal, and the latter signal is sent to the data line 30. , are transferred to the hard copy output section 44 through the changeover switch section 32 and data line 40. The hard copy output unit 44 outputs this image information as a hard copy 56. Note that this hard copy output operation can also be executed by an operator's instruction from the keyboard 50 without waiting for the recognition processing in the recognition unit 38 to be completed.

The storage area of the reception buffer section 22 that had accumulated the image information of the page for which recognition processing and/or hard copy output has been completed is released as an empty area under the control of the system control section 46 and the reception buffer control section 18. and is used for writing image information on another page.

In this way, this device performs character recognition processing in the recognition unit 38 and/or outputs notes from the note copy output unit 44 in parallel with receiving image information from the communication line 10, and performs operation. A hard copy of the image information after recognition processing can be obtained as necessary according to instructions from a person.

Even during such operations, image information can still be received from the communication control unit 12 to the reception buffer unit 22 without interrupting the processing.
The storage operation is performed sequentially for a series of pages.

Simultaneous parallel processing of receiving and storing image information, character recognition, and hard copy output requires, for example, the following time. If an A4 size image is received in the G III standard mentioned above at a main scanning speed of 20 milliseconds per horizontal scanning line and a sub-scanning length of 297 lines with a linear density of 787 dots/song in the sub-scanning direction, it will take 46 seconds. However, in the meantime, in one example, character recognition is performed on the A4 size image of the previous page in about 10 seconds, and then the main scanning speed is 5 milliseconds per horizontal scanning line, and the line density in the sub-scanning direction is 7.7. Dot/pus sub-scan length 297
It takes about 12 seconds to output a hard copy of the key, which takes about 22 seconds in total. Therefore, in this example, while one page of A4 size image information is being received, character recognition and hard copy output of the previous page can be performed with sufficient time.

By configuring the character recognition device according to the present invention in this way, the hard copy output of image information can be obtained only when requested by the operator, so that image information including unrecognizable "characters" is not wasted as a hard copy. In addition, unrecognized characters can be corrected immediately during facsimile reception.If a large number of unrecognized characters occur due to abnormal operations such as communication line failures, recognition processing may be delayed. Even if the facsimile signal reception is interrupted, or even if the facsimile signal reception itself is interrupted, useless note copies will not be output.With this device, after receiving the image information for one prefecture, the image information is Therefore, especially in the case of an interruption in facsimile communication, extra processing other than discarding the relevant image information stored in the reception buffer, which was required in the conventional technology, is not required. There are advantages to not doing so.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the character recognition device according to the present invention, and FIG. 2 is a 7-stem flow diagram for explaining the operation of the device shown in FIG. Explanation of symbols of main parts 12...Communication control unit 18 - Reception buffer control unit 22 - Reception buffer unit 32 - Changeover switch unit 38 - Recognition unit 44 - Hard copy output unit 46 -・／Stem control unit patent applicant Ricoh Co., Ltd.

Claims (1)

[Scope of Claims] Receiving means for receiving image information including characters from a communication line; storage means capable of reading and writing for temporarily accumulating the received image information in units of the image information; recognition means for recognizing characters; output means for outputting image information as a hard copy; connected between the storage means, the recognition means and the output means, and the image information stored in the storage means is transmitted to the recognition means and the output means; It includes a transfer means for transferring to one of the output means, and a control means, and the control means controls the transfer means when the accumulation means accumulates one unit of image information, so that the accumulation means thereafter transferring one unit of image information stored in the storage means to the recognition means, regardless of whether or not the unit of image information is being stored;
When the operator instructs hard copy output,
A character recognition device characterized in that when the recognition means finishes recognizing the one unit of image information, the transfer means is controlled to transfer one unit of image information stored in the storage means to the output means. .