JPH0823410A - Facsmile equipment and character recognition system using the facsimile equipment - Google Patents

Abstract

PURPOSE:To prevent mis-recognition of characters written on a slip such as a received slip by eliminating a residual form of received image information. CONSTITUTION:Form pattern information of a slip is stored in advance in a form pattern storage circuit 25, and when image information of a slip is received through a telephone line network 16, a form pattern elimination circuit operates received image information and form pattern information to eliminate a document form from the received image information. The image information eliminated from the slip form is transferred to a character recognition system. Thus, even when a form pattern and characters are mixed in the image information of the slip sent from a remote location, only the information excluding the frame pattern is transferred to a host computer 13. Thus, even when the form pattern is left in the received image information, the system allows a character recognition device 14 to execute correct recognition processing.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a facsimile machine and a character recognition system using the facsimile machine.

[0002]

2. Description of the Related Art In recent years, a FAX-OC in which a facsimile device (FAX) and an optical character recognition device (OCR) are integrated.
Various R systems have been developed. This FAX-OCR
The facsimile machine used in the system has a function of receiving image information transmitted from a facsimile machine at a remote place via a telephone line network and transferring it to an OCR in the system.

An image received by the FAX-OCR system is usually a form described in a fixed format (hereinafter referred to as "form"). An example of this form is shown in FIG.
It shows in (A). As shown in the figure, a form pattern P consisting of ruled lines and an identification mark ID are printed on the form, and characters are arranged and entered in each frame defined by the form pattern P. The form pattern P on this form is usually printed in a color (drop color) that cannot be read by a facsimile device.

For this reason, the image information transmitted from the transmitting side facsimile apparatus becomes an image of only characters with missing form patterns, which is the received image as shown in FIG. 5B in the facsimile apparatus of the FAX-OCR system. As received. This received image is sequentially sent to the OCR.

In the OCR, after the received image is developed, the image corresponding to the recognition target character entry position is cut out based on the identification mark ID. In this image cutout, information called FC (field control) indicating the character entry position of the recognition target in XY coordinates is used, and character recognition is executed for each of these cutout images. An example of the character recognition result is shown in FIG. In FIG. 5C, the amount of money from the received image of FIG.
An example of reading information on an account and a date is shown.

As described above, in the FAX-OCR system, since it is possible to read the image information of the form transmitted from an arbitrary facsimile apparatus at a remote place, the totalizing work such as editing the result and compiling it as an output slip. Can be done efficiently.

However, since various models are used as facsimile machines on the transmission side in a remote place, depending on the scanner sensitivity at the time of image reading and the type of light source,
In many cases, the form pattern that should not be sent as an image is not completely removed and a part of it is sent together with the characters. This state is shown in FIG. Here, FIG. 6A is a form read by the facsimile transmission on the transmitting side, and FIG. 6B is a received image received from the facsimile apparatus on the transmitting side by the facsimile apparatus of the FAX-OCR system.

In this case, the received image of FIG.
If the character recognition is performed with, the OCR cannot distinguish the ruled line of the form pattern P from the character, and thus the erroneous recognition as shown by the underlined portion in FIG. 6C occurs.

[0009]

Conventionally, if a form pattern is included in the image information of a form transmitted from a remote place, the form pattern and the character cannot be distinguished, and erroneous recognition occurs. was there.

The present invention has been made in view of the above point, and even when a form pattern and characters are mixed in the image information of a form transmitted from a remote place, only the information excluding the form pattern is recognized by the character recognition device. It is an object of the present invention to provide a facsimile apparatus which can be transferred to a computer and can perform accurate character recognition even if a form pattern remains in image information, and a character recognition system using the facsimile apparatus.

[0011]

The facsimile apparatus of the present invention can receive image information of a form transmitted from a partner terminal via a line network and transfer the image information to a character recognition system. Device, a form form storage means for storing form pattern information indicating a predetermined form form, and image information of the received form and form pattern information of the form form storage means are calculated, and the received form is stored. And a form form deleting means for deleting the form form from the image information, and is configured so that only the image information not including the form form can be transferred to the character recognition system.

In this facsimile, the form pattern information of the form is stored in advance. When the image information of the form is received via the line network, the image information and the form pattern information are calculated and the received image is received. The form form is deleted from the information. The image information from which the form form has been deleted is transferred to the character recognition system. Therefore, even when the form pattern and characters are mixed in the image information of the form transmitted from a remote place, only the information excluding the form pattern can be transferred to the character recognition system. Therefore, even if the form pattern remains in the received image information, it is possible to cause the character recognition system to execute accurate recognition processing.

[0013]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a FAX-according to an embodiment of the present invention.
The configuration of the OCR system and the facsimile machine used in this system is shown. The FAX-OCR system 11 is a system for realizing online data entry, and not only receives a general document from a remote place as a FAX but also recognizes necessary information from the received image by various characters to input various data. It has a function of reading the contents of the slip and collecting the result as an output slip.

As shown, the FAX-OCR system 11 comprises a facsimile machine 12, a host computer 13, a character recognition device 14, and a printer device 15.

The host computer 13 controls the facsimile device 12, the character recognition device 14 and the printer device 15 which are the connection devices thereof, and performs data aggregation work such as slip arrangement, and is composed of, for example, a personal computer. There is.

The facsimile apparatus 12 is a telephone line network 16
, And transmits / receives image information to / from various facsimile apparatuses 17-1 to 17-n in a remote place via the telephone line network 16. Further, the facsimile machine 12 has a communication function with the host computer 13, and can transfer the image information transmitted from the facsimile machines 17-1 to 17-n to the host computer 13.

In this case, the sending side facsimile device 17-
For the image information such as the forms received from 1 to 17-n, in order to prevent the character recognition device 14 from erroneously recognizing, the processing for deleting the form pattern of the form from the image information is performed in the facsimile device 12. And then transferred to the host computer 13. As a result, only the identification mark information and the character information can be transferred to the host computer 13.

As shown in the figure, the facsimile apparatus 12 includes an image storage circuit 21 and a form pattern storage circuit 2.
5, a form pattern deleting circuit 26, and a host interface 27, which are interconnected by an internal bus 20.

The image storage circuit 21 is a device for receiving and storing the image information transmitted from each of the transmission side facsimile devices 17-1 to 17-n, and includes a communication control circuit 22, a storage buffer 23, and image information. Analysis circuit 2
4.

The communication control circuit 22 controls communication with the partner facsimile machine via the telephone line network 16. in this case,
The communication control with the partner facsimile is performed by, for example, the G3 standard facsimile communication control. Further, the communication control circuit 22 outputs the received image information such as a form from the partner facsimile to the accumulation buffer 23. When the received image information is in the standard mode (3.85 lines / mm resolution), it is finely processed. It is converted into the mode (resolution of 7.71 lines / mm) and output to the accumulation buffer 23.

The storage buffer 23 is a memory device for storing the image information received via the communication control circuit 22. The image information analysis circuit 24 develops an image of received image information such as a form accumulated in the accumulation buffer 23 on a memory in the image information analysis circuit 24, and analyzes the form pattern type and paper direction of the received form. To do.

The form pattern storage circuit 25 is a memory device capable of storing a plurality of form patterns, in which one or more predetermined form patterns are registered by the host computer 13.
In this case, the form pattern is fine (7.71 lines / mm
Resolution) mode is registered.

The form pattern deletion circuit 26 is an arithmetic circuit that deletes the form pattern from the received image information, reads the form pattern corresponding to the received image information from the form pattern storage circuit 25, and receives the image developed with the form pattern. The calculation with the image information is performed in pixel units.

The host interface circuit 27 is a device for exchanging various data and commands with the host computer 13. When the image information of the form is received, the form pattern deleting circuit 26 creates a form pattern. The deleted image information is transferred to the host computer 13.

Next, with reference to FIG. 2, the principle of the form pattern deleting process executed by the facsimile apparatus 12 will be described. FIG. 2A shows the form pattern information registered in the form pattern storage circuit 25, and this form pattern information includes a form pattern P1 consisting of ruled lines. Further, FIG. 2B shows the received image information in which the form pattern P1 is not partially omitted and remains.

The form pattern deletion circuit 26 is shown in FIG.
Inversion value A of each dot A in the form pattern information of (A)
The logical product (= A-B) of the dot value B of the received image information of FIG. 2B is calculated. As a result of this calculation, FIG.
As shown in (C), the form pattern P1 is deleted from the received image information in FIG. 2 (B).

When form pattern information corresponding to a plurality of types of forms is registered in the form pattern storage circuit 25, the received image information is based on the size (α, β) of the identification mark ID of the received image information. Of the form type is determined, and the form pattern information corresponding to the form type is used for the deletion calculation.

Next, with reference to the flowcharts of FIGS. 3 and 4, a specific processing procedure of the received image information by the facsimile apparatus 12 will be described. First, referring to FIG.
The processing procedure when there is one form pattern information registered in the form pattern storage circuit 25 will be described.

Here, the case where the facsimile machine 12 receives the image information from the transmitting side facsimile machine in the communication procedure according to the G3 standard will be described as an example. Note that FIG.
Symbol DIS, DCS, TCF, CFR, MPS, EO
P and MCF are signals defined by the G3 standard, and the description thereof is omitted here.

The facsimile apparatus 12 is set to the automatic reception mode, and when the transmission side facsimile apparatus issues a connection request, the communication control circuit 22 performs line connection (steps S11 and S12). After this, when the communication setting notification DCS is issued from the sending facsimile machine, the communication control circuit 22 causes the image mode (standard / fine) designated by the DCS to be registered in the form pattern storage circuit 25. It is determined whether or not it matches the image mode of the pattern information (steps S13 and S14).

As described above, the form pattern information is registered in the form pattern storage circuit 25 in the fine mode. Therefore, when the image mode designated by the DCS is the standard mode, the communication control circuit 22 immediately disconnects the line (step S15) or receives the image information from the transmitting side facsimile apparatus. , Storage buffer 2 while converting it to the resolution of the fine mode
3 (step S16). The reception process and the output process to the accumulation buffer 23 are sequentially executed until the output process of the information of the page last line image information is completed (steps S16 and S17).

On the other hand, if the image mode designated by the DCS is the fine mode, the communication control circuit 22 outputs the image information received from the transmitting side facsimile device as it is to the accumulation buffer 23 without converting the image mode. (Step S18). The reception process and the output process to the storage buffer 23 are sequentially executed until the output process of the information of the page last line image information is completed (steps S18 and S1).
9).

In this way, the image information transmitted from the transmitting facsimile is sequentially stored in the storage buffer 23 page by page (form unit). In this case, all the image information is compressed and encoded.

The image information analysis circuit 24 includes the accumulation buffer 2
The image information stored in 3 is sequentially developed in units of lines (step S20), and by detecting the presence or absence of the left and right identification marks, it is detected in units of forms whether or not the inserting direction of the forms in the sending facsimile device is correct. (Step S2
1). When the form format as described with reference to FIG. 2 is used, the case where the identification mark IDs are on the left and right sides is the normal insertion direction.

When an error in the form insertion direction, that is, upside-down inversion of the image information is detected, the image information analysis circuit 24
Converts the image information in the storage buffer 23 to the normal direction (step S22), and performs the processing of steps S20 and S21 again.

Next, the image information analysis circuit 24 detects the form type by checking the size of the identification mark ID (α and β in FIG. 2B) from the image information that has undergone image development, and the form pattern storage circuit 25. It is determined whether the form type is the same as the form pattern registered in (step S23). If the form types do not match, the image information analysis circuit 24 instructs the communication control circuit 22 to disconnect the line (step S24).

On the other hand, when the form types match, the form pattern deleting circuit 26 is operated. The form pattern deletion circuit 26 reads out the image information of the form for which a match of the form type is detected from the accumulation buffer 33, develops the image, and shifts from the misalignment of the left and right identification marks to the paper insertion angle (θ) in the sending facsimile machine. It is checked whether or not there is (skew) (step S25).

If there is a skew, the form pattern deletion circuit 26 uses the form pattern information read from the form pattern storage circuit 25 as the paper insertion angle (θ).
It is rotated only (step S26), and then a deletion process is performed (step S27). If there is no skew, the form pattern deletion circuit 26 performs the deletion process without rotating the form pattern information (step S27).

In the deleting process, as described above, the image information of the received form and the form pattern information are calculated, and if the form pattern remains in the received image information, it is deleted. The received image information from which the form pattern has been deleted is again compression-coded and stored in the storage buffer 23.

Thereafter, the host interface 27 reads the image information from the storage buffer 23 and sends it to the host computer 13 to complete the processing (step S28).

The host computer 13 receives the image information of the form sent from the facsimile device 12 from the character recognition device 1.
4 to perform character recognition. In the character recognition device 14, after the image information of the form is developed, the image corresponding to the recognition target character entry position is cut out based on the identification mark ID. In this image cutout, FC (field control) is used, and character recognition is executed for each of these cutout images. The character recognition result is transferred to the host computer 13 again.

When image information other than a form such as a general document is sent from the facsimile device 12, the host computer 13 prints it out to the printer device 15.

Next, referring to the flowchart of FIG.
The processing procedure when a plurality of types of form pattern information are registered in the form pattern storage circuit 25 will be described.
When a connection request is issued from the sending facsimile machine,
The communication control circuit 22 performs line connection (step S31,
S32). After that, when the communication setting notification DCS is issued from the transmission side facsimile device, the communication control circuit 22 recognizes the image mode (standard / fine) designated by the DCS (step S33).

After that, the communication control circuit 22 starts receiving the image information transmitted from the transmitting side facsimile device and outputting the received image information to the accumulation buffer 23 (step S34). This reception processing and accumulation buffer 23
The output processing to the is sequentially executed until the output processing of the image information of the last line of the page is completed (steps S34 and S35).

In this way, the image information transmitted from the transmitting facsimile is sequentially accumulated in the accumulation buffer 23 in page units (form units). In this case, all the image information is compressed and encoded.

The image information analysis circuit 24 includes the accumulation buffer 2
The image information stored in 3 is sequentially developed in units of lines (step S36), and the presence or absence of the left and right identification marks is detected to detect whether or not the inserting direction of the form in the facsimile apparatus on the transmitting side is correct in form units. (Step S3
7). When an error in the form insertion direction, that is, upside-down reversal of the image information is detected, the image information analysis circuit 24 converts the image information in the accumulation buffer 23 into the normal direction on a form-by-form basis (step S38). The processing of S36 and S37 is performed.

Next, the image information analysis circuit 24 detects the form type for each form by checking the size of the identification mark ID (α and β in FIG. 2B) from the image information that has been image-developed, and detects the form pattern. It is determined whether or not the form pattern of the corresponding form type is registered in the storage circuit 25 (step S39). When the form pattern of the corresponding form type is not registered, the image information analysis circuit 24 instructs the communication control circuit 22 to disconnect the line (step S40).

On the other hand, when the form pattern of the corresponding form type is registered, the image information analysis circuit 24
First, it is checked whether or not there are a plurality of form types in the received image information (step S41). If there are a plurality of form types, the form types are classified and re-output from the accumulation buffer 23 (step S42).

After this, the form pattern deletion circuit 26
Checks whether or not the image mode of the image information in the storage buffer 23 matches the image mode (fine mode) of the form pattern information registered in the form pattern storage circuit 25 (step S43). If they are different, the form pattern deletion circuit 26 reads the registered form pattern of the form pattern storage circuit 25 while converting it according to the image mode of the stored image information (step S44), and if they match, the resolution. The data is read as it is without conversion (step S45).

Next, the form pattern deletion circuit 26
Develops the received image information in the storage buffer 23, and checks whether there is a deviation (skew) in the sheet insertion angle (θ) in the transmitting side facsimile apparatus due to the deviation of the left and right identification marks (step S46).

If there is a skew, the form pattern deletion circuit 26 uses the form pattern information read from the form pattern storage circuit 25 as the paper insertion angle (θ).
It is rotated only (step S47), and then the deletion process is performed (step S48). If there is no skew, the form pattern deletion circuit 26 performs the deletion process without rotating the form pattern information (step S48).

In the deletion process, as described above, the image information of the received form and the form pattern information are calculated, and if the form pattern remains in the received image information, it is deleted. The received image information from which the form pattern has been deleted is again compression-coded and stored in the storage buffer 23.

Thereafter, the host interface 27 reads the image information from the storage buffer 23 and sends it to the host computer 13 to complete the processing (step S49).

As described above, in the facsimile apparatus 12 of this embodiment, the form pattern information of the form is stored in advance in the form pattern storage circuit 25, and when the image information of the form is received via the telephone line network 16. The form pattern deletion circuit calculates the received image information and the form pattern information to delete the form form from the received image information.

The image information from which the form form has been deleted is transferred to the character recognition system. Therefore, even when the form pattern and the characters are mixed in the image information of the form transmitted from a remote place, only the information excluding the form pattern can be transferred to the host computer 13. Therefore, even if the form pattern remains in the received image information, it is possible to cause the character recognition device 14 to execute an accurate recognition process.

Further, by deleting the remaining form of the received image information, the received image information amount can be reduced and the accumulated image information amount in the facsimile apparatus 12 can be increased. In addition, the host computer 13
It is also possible to shorten the communication time to. Furthermore, the processing speed (throughput) as a system from image reception to character recognition result output can be improved.

In this embodiment, the function for deleting the form is provided in the facsimile device 12. However, this function is, for example, the host computer 13 or the character recognition device 1.
It is also possible to provide it within 4.

Further, here, the case where the communication with the facsimile apparatus at the remote place is performed through the telephone line network has been described, but the FAX-OCR system 11 of this embodiment is as follows.
It can be similarly applied to an in-company system using a dedicated line such as a LAN.

[0059]

As described above, according to the present invention, even when the form pattern and the characters are mixed in the image information of the form transmitted from a remote place, only the information excluding the form pattern is transmitted to the character recognition device. It becomes possible to transfer, and it becomes possible to perform accurate character recognition even if the form pattern remains in the image information.

[Brief description of drawings]

FIG. 1 is a block diagram showing the overall configuration of a FAX-OCR system according to an embodiment of the present invention.

FIG. 2 is a diagram for explaining the principle of remaining form deletion processing executed by the facsimile apparatus provided in the system of the embodiment.

FIG. 3 is a flowchart showing an example of a procedure of an image receiving process of a facsimile device provided in the system of the embodiment.

FIG. 4 is a flowchart showing another example of the procedure of the image receiving process of the facsimile apparatus provided in the system of the embodiment.

FIG. 5 is a diagram for explaining the principle of character recognition processing of a conventional FAX-OCR system.

FIG. 6 is a diagram showing a situation in which erroneous recognition occurs when received image information includes a remaining form in a conventional FAX-OCR system.

[Explanation of symbols]

11 ... FAX-OCR system, 12 ... Facsimile device, 13 ... Host computer, 14 ... OCR device, 1
5 ... Printer device, 16 ... Telephone line network, 17-1 to 17
-N ... Sending side facsimile device, 21 ... Image storage circuit,
22 ... Communication control circuit, 23 ... Storage buffer, 24 ... Image information analysis circuit, 25 ... Form pattern storage circuit, 26
... Form pattern deleting circuit, 27 ... Host interface.

Claims (9)

[Claims] 1. A facsimile device capable of receiving image information of a form transmitted from a partner terminal via a line network and transferring the image information to a character recognition system, a form pattern showing a predetermined form form. A form form storage unit for storing information, and a form form deleting unit for calculating the image information of the received form and the form pattern information of the form form storage unit to delete the form form from the received image information of the form. And a facsimile machine configured to transfer only image information not including a form form to the character recognition system. 2. The facsimile apparatus has first and second image modes in which image information resolutions are different from each other,
The form pattern storage unit stores the form pattern information in the resolution of the first or second image mode, and the image mode of the image information transmitted from the partner terminal does not match the image mode of the form pattern information. At this time, the facsimile apparatus according to claim 1, further comprising image mode conversion means for converting image information transmitted from the partner terminal into a resolution corresponding to an image mode of the form pattern information. 3. The method further comprises means for detecting a form insertion direction in the partner terminal based on a reference mark of the received image information and converting the received image information in the same direction as the form pattern information. The facsimile apparatus according to claim 1, which is characterized in that. 4. The apparatus further comprises means for detecting the inclination of the form in the partner terminal based on the reference mark of the received image information and rotating the form pattern information according to the inclination. The facsimile apparatus according to Item 1. 5. A facsimile apparatus capable of receiving image information of a form transmitted from a partner terminal via a line network and transferring the image information to a character recognition system, wherein a plurality of form forms are displayed. Form form storage means for storing form pattern information of the form, form form detecting means for detecting the form of the form transmitted from the partner terminal by analyzing the received image information, and showing the detected form form The form pattern information is read from the form form storage means, the received image information and the form pattern information read from the form form storage means are operated, and the form form included in the received form image information is calculated. An image that does not include a form form that is equipped with a form form deletion method to be deleted. Facsimile apparatus characterized by being configured to forward only to the character recognition system distribution. 6. The apparatus further comprises means for detecting a form insertion direction in the partner terminal based on a reference mark of the received image information and converting the received image information in the same direction as the form pattern information. 6. The facsimile apparatus according to claim 5, wherein the form form detecting means detects the form of the form transmitted from the partner terminal according to the size of the reference mark. 7. The facsimile apparatus has first and second image modes in which image information resolutions are different from each other,
The form form storage means stores a plurality of form pattern information having different resolutions and types, and the form form deleting means stores the form forms according to the resolution of the image information transmitted from the partner terminal and the type of the form form. 6. The facsimile apparatus according to claim 5, wherein the corresponding form pattern information is read from the storage means. 8. The apparatus further comprises means for detecting the inclination of the form in the partner terminal based on the reference mark of the received image information and rotating the form pattern information according to the inclination. The facsimile apparatus according to Item 5. 9. A character recognition system for receiving image information of a form transmitted from a partner terminal via a line network and recognizing the characters of the form from the received image information, a form pattern showing a predetermined form form. A form form storage means for storing information, image information of the form received from the partner terminal and form pattern information of the form form storage means are calculated, and a form form included in the received image information of the form is calculated. A character recognition system comprising: a form form deleting unit for deleting; and a character recognizing unit for recognizing a character of the form from image information in which the form form is deleted by the form form deleting unit.