JPH06103411A - Document reader - Google Patents

Abstract

PURPOSE:To read the contents of a document, and to exhibit a corrected picture to a user or to store it by detecting an angle of rotation, and correcting the input picture to a correct direction even in the case that the document is input ted while being rotated by an arbitrary angle to the set direction of a scanner. CONSTITUTION:This reader is provided with a picture input means 105 for inputting a document picture, a character line extracting means 110 for extracting the character line of the inputted document picture, an inclination extracting means 125 for extracting the inclination of the document, a character line coordinate rotating means 165 for rotating the extracted character line by the angles obtained by adding 0 deg., 90 deg., 180 deg., 270 deg. to the inclination of the document, and a document rotation angle judging means 150 for recognizing respectively four rotated character strings and judging the angle of rotation of the most correct one among them as the inclination of the document, and the inclination of the document is corrected.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document reading device, an electronic file device, a facsimile device, a device for converting a paper document input to a copying machine and a computer into electronic data, and more particularly to a predetermined scanner. The user is not aware of the (scanning line) reading direction (moving direction), and the paper document is read on the scanner from 0 degrees to 3 degrees.
A document reading device and an electronic device with improved usability such as reading the contents described in a document or correcting an input image in the correct direction regardless of the angle set up to 60 degrees The present invention relates to a file device, a facsimile device, a copying machine, and a computer.

[0002]

2. Description of the Related Art Originally, there has been known a document reading device which can read the contents described in a document only when the reading direction of the scanner and the setting direction of the document on the scanner match. However, in recent years, with the realization of document reading devices, when a document is input with a tilt, and at the same time, the reading direction of the scanner and the setting direction of the document on the scanner do not match, the setting state is insufficient. There is a situation in which the input document must be read.

As a conventional example for solving the above problems, Japanese Patent Laid-Open Nos. 62-14277 and 2-105266 are available.
There is a gazette.

In the apparatus disclosed in Japanese Patent Laid-Open No. 62-14277, contour extraction is performed from an image, the inclination of a straight line portion such as a ruled line is extracted from the extracted contour, and the inclination of this straight line portion is horizontal (or By correcting the inclination so that it becomes (vertical), the inclination of the image can be corrected.

In the apparatus disclosed in Japanese Patent Application Laid-Open No. 2-105266, the counting results counted by the black pixel counting means are compared, and the document image input based on the comparison result is rotated, whereby the vertical relationship of the document images is automatically detected. Fix it.

[0006]

However, in the above-described conventional apparatus, the vertical relationship is the same, and the vertical relationship is the same (the inclination is 0 to 90 degrees), or the input document has no inclination and the vertical relationship is different. The contents described in the document can be read only in the open state (inclination is 180 degrees).

That is, in the above-mentioned conventional apparatus, when the input is rotated by an arbitrary angle (0 to 360 degrees) with respect to the setting direction of the scanner, the contents cannot be read or the input is made in the correct direction. It does not have a function that considers the usability of the user such as modifying the image and not presenting the image to the user.

As shown in FIG. 3, the direction 2310 in which the scanning line of the A3 scanner 2300 moves and the area 2320 for setting a document are predetermined. Therefore, in the conventional apparatus, the processing direction of the system cannot be processed unless the scanning line of the scanner moves.

Therefore, as the allowable range of processing, as shown in FIG. 4, the document 2400 has the same vertical relationship and is slightly tilted, or, as shown in FIG. 5, the document 2500 has no tilt and the vertical relationship. Was the flipped one. As shown in FIG. 6, when the document 2600 has a tilt in addition to the rotation of 90 degrees, 180 degrees, or 270 degrees, the conventional system cannot read the content described in the document and corrects the input image. It did not have a function to present to the user.

Here, the rotation angle is defined as the difference between the reading direction of the scanner and the direction indicated by the vertical direction of the document. For example, the arrow 2605 in FIG. 5 indicates the vertical direction of the document, and the rotation angle is an angle considering the concept of the vertical relationship of the document.

Further, in the above apparatus, the user is required to detect when the user mistakenly inputs the back side of the document or when the portion to be read of the document is out of the reading area of the scanner or the method of processing such a document. It does not have a function considering the usability of.

Further, in the above apparatus, additional information such as an author, a place of acquisition, a date and time of acquisition, a memo, etc. relating to a document not described in the document can be input in association with the information regarding the input document, and the relationship between documents can be established. Since there is no function to have it, it is not possible to register or search information that is not described in the document, and it is not possible to search information related to the desired document from information of other related documents, etc. Does not have a function that considers the usability of the user.

Further, in the above apparatus, a file capacity checking function at the time of data registration, a reading function for a large amount of image data stored and managed, an image rotation correction function, or external character processing at the time of character recognition The user-friendliness of functions and the like are not considered.

As described above, the conventional device has a problem that it is inconvenient for the user.

Therefore, a first object of the present invention is to set a document at an arbitrary angle (from 0 degree to 360 degrees) with respect to the setting direction of the scanner.
Even if the image is rotated and input, the contents can be read, or a document reading device or an electronic file that considers the usability of the user such as the function of correcting the input image in the correct direction and presenting it to the user. To provide an apparatus, a facsimile machine, a copying machine, or a computer.

A second object of the present invention is to detect when the user mistakenly inputs the back side of the document or when the portion to be read of the document is out of the reading area of the scanner, and to process such a document. A method and a document reading device, an electronic file device, a facsimile, a copying machine, or a computer, which considers the user's usability such as automatically reading the contents of the protruding area without the user having to set the document again. .

The third object of the present invention is to register and retrieve additional information related to a document which is not described in the document, such as author, source, date and time of acquisition, memo, etc. Another object of the present invention is to provide a document reading device, an electronic file device, a facsimile device, a copying machine, or a computer having a function in consideration of user's usability, such as searching information related to a desired document from information of another document.

Further, a fourth object of the present invention is to check a file capacity at the time of data registration, a reading function for a large amount of image data stored and managed, an image rotation correction function, or character recognition. Another object of the present invention is to provide a document reading device, an electronic file device, a facsimile device, a copying machine, or a computer in consideration of user's usability such as an external character processing function.

[0019]

In order to achieve the above first object, means for inputting a document image, means for extracting character lines of the input document image, and means for extracting inclination of a document. And the above extracted character line to the inclination of the document as 0
A unit for rotating an angle of 90 °, 180 °, 270 ° and four rotated character lines are respectively recognized, and the most correct rotation angle among them is used as the inclination of the document to correct the image.

In order to achieve the second object, the means for extracting the character line of the document image described above is provided with means for determining that the document is input on the back side when the character line is not extracted.

Further, the character line to be read depends on whether or not two or more of the four vertices of the character line obtained by the means for extracting the character line at an arbitrary position are present on the four sides of the scanner reading area. A means for determining whether or not it is outside the scanner reading area is provided.

Further, when it is determined that the document is protruding in the A4 scanner reading area by the above-mentioned means, a new document image is collected by the A3 scanner to automatically display the contents described in the document, including the contents of the protruding area. And means for reading without leaking.

Further, the apparatus is provided with means for registering the document number of the document image determined to be incomplete in the reject file, and means for displaying an error message on the window or prompting a warning by voice call. Is also good.

In order to achieve the third object, a means for registering a document number in the input document image and additional information which is not written in the document such as date and time of acquisition, purpose and destination of the document are input. Means for inputting additional information, electronically recording the input additional information in association with the document number or the document information obtained by processing the input document, and means for searching the additional information or the document information. .

Further, there may be provided means for electronically recording related information between documents and means for searching for related information between documents and searching for document information or additional information about a desired document.

In order to achieve the fourth object, the present invention provides
There is a means to display the free space to output the processing result of the input document image to a file, a means to display a warning on the window or a voice prompt when the free space is low, or an operator via the network. And a document reading device, an electronic file device, a facsimile device, a copying machine, or a computer.

Further, a means for storing a document image in which a plurality of documents are scanner-input into the data storage device, a means for adding a document number at the time of storage, a stored document image is sequentially loaded to perform image rotation correction or reading processing. And means for performing.

Further, in the reading process of the apparatus for inputting a document as a digital image, if the character code to be recognized does not exist on the system side, a means for registering the character image as an external character in the dictionary or a registered symbol as a reading result. Means for allocating and displaying or outputting to a file may be provided.

[0029]

With the above structure, a character line at an arbitrary position in the document image is extracted, an inclination of the document image with respect to the scanner setting direction is extracted, and an appropriate angle for obtaining the rotation angle of the document image with respect to the scanner setting direction is obtained. Select a character line, and select four angles of 0 degrees for the tilt that detected the partial image of the appropriate character line,
Rotate by 90 degrees, 180 degrees, and 270 degrees, respectively, and cut out characters in the partial image from each of the four types of rotated character line partial images, recognize them, and evaluate the rotation angle of the input document image with respect to the scanner setting direction. It is possible to evaluate the rotation angle of the input document image with respect to the scanner setting direction by using the means or the layout information of the input document in this means, and rotate and correct the input image by the obtained rotation angle of the document. Further, a document input at an arbitrary rotation angle can be displayed or stored on the display so that the user can easily see it, and the usability of the user's data operation can be greatly improved. In addition, the rotation correction character line image is processed by re-extracting the character line from the rotation correction image and reading the described content or by rotating the character line partial image by the obtained rotation angle and using the layout information. It is possible to encode the character image described in the document input at an arbitrary rotation angle, which was not possible in the past, by sequentially reading the described contents by determining the order to perform, and automatic registration using an auto feeder etc. Since there is no need to re-input data when the user manually inputs data, it is possible to improve the efficiency of input work.

Further, the input image can be corrected and rotated by the obtained rotation angle and the corrected images can be sequentially stored, and the user can register the document image without being aware of the scanner input direction. Further, since the document input is performed by using the cover of the scanner, the input work of the spread document such as a magazine is difficult because the carver of the scanner interferes with every other page. However, since the rotation angle of the document is determined and the image is corrected / stored, the user can easily input the document by reversing the document without paying attention to the setting direction of the document. Can be realized.

Depending on the presence or absence of character lines extracted from the document image, it is possible to automatically determine whether the input document is normally input on the front side or is input by mistake on the back side. It is possible to give instructions, and the efficiency of input work can be realized. In addition, because the document number of the input error is registered in the reject file, it is possible to check the input status during manual data input or automatic registration using the auto feeder, etc., and it is possible to greatly reduce human confirmation work by visual inspection etc. .

Further, 2 of the four vertices of the character line obtained by the means for the system to extract the character line at an arbitrary position.
It is possible to judge whether or not the character line to be read is outside the scanner reading area according to whether or not the vertices and above are present on the four sides of the scanner reading area, and thus it is possible to instruct the user of an error. The efficiency of input work can be realized. In addition, because the document number of the input error is registered in the reject file, it is possible to check the input status during manual data input or automatic registration using the auto feeder, etc., and it is possible to greatly reduce human confirmation work by visual inspection etc. .

When it is found by the above means that the input document is out of the A4 scanner reading area, a document image is newly picked up by the A3 scanner and the reading process is performed on the document image so that the user can visually check it. There is no need to check the protrusion, and the system can automatically read the contents of the document without leaking it even without resetting the document and performing the reading process again. it can.

When the system determines that the input is inadequate, the user can be informed of the error by displaying an error message on the window or prompting a warning by calling out by voice. The efficiency of can be realized.

Further, the document number is registered in the input document image, and the user inputs additional information for inputting additional information not written in the document such as the author, date and time of acquisition, and memo of the user. Then, since the input additional information is electronically recorded in association with the document number or the document information obtained by processing the input document, the user specifies the additional information or the document information,
The corresponding additional information or document information can be retrieved more efficiently than the retrieval means, and the information can be retrieved easily.

Further, in order to electronically record the relationship information between documents, the user can easily search the document relationship information by the searching means, search for a document from another document, and obtain document information about the document. The additional information can be taken out, and the desired information can be obtained from the user's vague memory by using the information about other documents.

The free space for outputting the processing result of the input document image to a file is displayed, and when the free space becomes small, a warning is displayed on the window or voice is urged, or an operator is requested via a network. It is possible to avoid the redo of the input work and the adverse effect on the system by urging the other device having a warning to warn.

Further, by storing a document image in which a plurality of documents are scanner-input into the data storage device, adding a document number at the time of storage, and sequentially loading the stored document images to perform reading processing or image rotation correction. A character image can be encoded while managing information about the document for a large number of input document images, and the correction work of the user can be reduced.

Further, in the reading process of the device for inputting a document as a digital image, if the character code to be recognized does not exist on the system side, the character image is registered in the dictionary as an external character, and a certain symbol is registered as the external character reading result. By allocating and displaying or outputting to a file, it is possible to deal with input of unrecognizable characters that do not exist in the system. In addition, the user can easily process the reading result without outputting the meaningless recognition result, and the reading accuracy is improved.

[0040]

The present invention will be described in detail below with reference to the embodiments shown in the drawings. The present invention is not limited to this.

FIG. 2 is a block diagram of a document reading apparatus according to an embodiment of the present invention.

This document reading device includes a CPU 2110, a main memory 2120, an image memory 2130, and a CRT 21.
40, a keyboard 2150, a mouse 2160, a scanner control unit 2170, a scanner 2180, a data storage unit 2190, a magnetic disk 2200, an optical disk 2210, a magneto-optical disk 2220, a printout device 2230, and a speaker 2240. And the bus 225
It consists of 0 and.

FIG. 1 is a block diagram of an embodiment of the document reading apparatus of the present invention.

The image input means 105 is the scanner 210.
8, a scanner control unit 2170, a CPU 2110, and an image memory 2130. The document 101 is read to obtain a document image, which is temporarily stored and stored in the CRT 2
Display at 55.

The character line extracting means 110 is the CPU 211.
A character line in the document is extracted from the document image which is composed of 0 and is stored in the image memory 2130. This character line extracting method is disclosed in, for example, Japanese Patent Laid-Open No. 62-165284.

The front / back determination means 115 is the CPU 2110.
And the result of the character line extraction means 110 is used to determine whether or not the input document 101 is mistakenly input on the back side. When it is determined that the input is made on the back side, a signal is sent to the reject warning means 245, and this means 245 causes the CRT 2 via the input / output control means 250.
The display on the screen 55 or the speaker 280 is used to prompt the user for a back side input warning. Also, the reject registration means 2
05, the document number added by the document number registration means 195 is registered in the reject file 210.

The protrusion determination means 120 is the CPU 21.
It is constituted by 10 and using the result of the character line extraction means 110, it is determined whether or not the input document image is outside the reading area of the scanner 2108. When it is determined that the reading area has been pushed out, a signal is sent to the reject warning means 245, and this means 245 prompts the user to give a backside input warning using the display on the CRT 255 or the speaker 280 via the input / output control means 250. .
Further, the reject registration means 205 registers the document number added by the document number registration means 195 in the reject file 210.

The inclination extracting means 125 is the CPU 2110.
The inclination of the input document is extracted from the document image stored in the image memory 2130. This inclination extraction method is disclosed in, for example, Japanese Patent Application Laid-Open No. 62-14277.

The optimum character line selection means 130 is the CPU 2
Character recognition based rotation angle evaluation means 1
In order to realize high-accuracy and high-speed processing as the character line for evaluation in 45, a plurality of optimum character lines are selected from the character lines obtained by the character line extracting means 110.

The character line image rotating means 140 is the CPU 2
An image of a plurality of character lines, which is composed of 110 and is selected by the optimum character line selection unit 130, has four types of rotation angles, that is, the inclination obtained by the inclination extraction unit 125 has 0 degree, 90 degrees, or 180 degrees. It rotates by the rotation angle of 270 degrees.

The character recognition base rotation angle evaluation means 145 comprises the CPU 2110, and the optimum character line selection means 1
A plurality of character lines selected by 30 are rotated by the character line image rotating means 140 at four types of rotation angles, and character recognition is performed by performing character segmentation and character recognition on each rotated character line image. Four kinds of rotation angles are evaluated using the similarity of the result.

A method of utilizing the layout information of the document in addition to the result of the character recognition-based rotation angle evaluation means 145 to obtain the determination of the rotation angle of the document with high accuracy will be described.

The character line coordinate rotating means 165 is the CPU 2
110, and the coordinates of the character line obtained by the character line extraction means 110 are added to four types of rotation angles, that is, the inclination obtained by the inclination extraction means 125, 0 degree, 90 degrees, 180 degrees or 270 degrees. Rotate by the rotation angle considered.

The layout information extraction means 170 uses the CP
The character line coordinate rotating means 16 is composed of U2110.
Layout information is extracted with respect to the character line coordinates obtained by rotating at four types of rotation angles obtained in step 5. This layout information extraction method is disclosed in, for example, Japanese Patent Application Laid-Open No. 1-130293.

The layout base rotation angle evaluation means 175 is composed of the CPU 2110, and the layout knowledge 18
The evaluation is performed by using 0 and the layout information obtained by rotating the layout information extracting means 170 at four kinds of rotation angles.

The document rotation angle determination means 150 is the CPU 2
Character recognition based rotation angle evaluation means 1
45, or the rotation angle of the input document is determined on the basis of the evaluation results of the four types of rotation angles obtained by this means and the layout-based rotation angle evaluation means 175. When the determination result of the rotation angle is ambiguous by this determination means, a signal is sent to the reject warning means 245, and this means 245 uses the display on the CRT 255 or the speaker 280 via the input / output control means 250. Prompt the user with a warning that the rotation angle cannot be determined. Further, the reject registration means 205 registers the document number added by the document number registration means 195 in the reject file 210.

The image rotation means 155 is the CPU 2110.
The document image stored in the image memory 2130 is rotated by the rotation angle obtained by the document rotation angle determination means 150.

The reading means 160 is composed of the CPU 2110, and converts the character image in the corrected image into a character code for the corrected document image rotated by the rotation angle by the image rotating means 155.

The reading result correction means 240 is the CPU 21.
10, the reading result and the correction result are displayed on the CRT 255 for the contents processed by the reading unit 160 via the input / output control unit 250, and the keyboard 260 is used.
Alternatively, the mouse 265 is used to correct the reading result.

The document number registration means 195 is the CPU 21.
The input document 101 is provided with a document number, and the document number is registered in the document number file 200.

The input image registration means 185 is the CPU 21.
The input document image which is composed of 10 and is stored in the image memory 2130 together with the document number assigned by the document number registration means 195 is registered in the input image file 190.

The corrected image registration means 215 is the CPU 21.
The modified document image composed of 10 and the modified document image modified by the image rotation means 155 together with the document number.
Register to 0.

The reading result registration means 225 is the CPU 21.
The result read by the reading unit 160 is registered in the read result file 230 together with the document number.

The additional information registration means 275 is the CPU 21.
It is composed of 10 units, and the information input from the keyboard 260 or the mouse 265, not the information input from the scanner 2180, is added or associated and managed.
Then, information input from other than the scanner 2180 such as the keyboard 260 or the mouse 265 is registered in the additional information file 270.

The file control means 235 is the CPU 21.
It is composed of 10 and registers or manages a plurality of files as described above, or associates information between these files. Then, the management is performed by using the relationship between the information in the same document and the information in the different documents between a plurality of files.

The above file group is stored in the magnetic disk 2200 or the optical disk 221 via the data storage unit 2190.
0 or stored on the magneto-optical disk 2230.

Next, a rough processing flow of this system will be described with reference to FIG.

-Correct according to the claim.

First, the image input 2700 converts paper document data into electronic image data. Then, the document number is added to the document image converted in the document number registration 2705. Then, the document image is registered in the document image registration 2710. Then, in the character line extraction 2715, the character line existing in the document image is extracted. And front / back determination 2735
At, it is determined whether or not the input document is mistakenly input on the back side according to the presence or absence of character lines. Then, a rejection determination 2740 determines whether or not the input document should be rejected. If rejected, the reject warning 2745 and the document number are reject registered 2750. Then, it is determined whether or not the input document is out of the reading area of the scanner in the out-judgment determination 2755. Then, a rejection determination 2760 determines whether or not the input document should be rejected. If rejected, the reject warning 2765 and the document number are reject registered 2770. Then, the tilt detection 277
At 2, the inclination which is the difference between the horizontal line in the scanner reading area and the horizontal line of the input document is detected. Then, in the optimum character line selection 2774, a plurality of optimum character lines to be evaluated are selected in the character recognition base rotation angle evaluation 2778 for obtaining the rotation angle of the input document. Then, the character line image rotation 277
0 is added to the inclination extracted from the plurality of character line images selected in 6.
It rotates by four types of rotation angles including degrees, 90 degrees, 180 degrees, and 270 degrees. Then, in the character recognition base rotation angle evaluation 2778, characters are cut out from the character line image rotated at each rotation angle and recognized, and four types of rotation angles are evaluated according to the similarity at that time. Then, in layout analysis evaluation execution 2780, character recognition based rotation angle evaluation 277 is performed.
8 is branched depending on whether or not the evaluation of the rotation angle using the layout information is added. If the evaluation of the rotation angle using the layout information is also taken into consideration, the inclination of extracting the character line coordinates in the character line coordinate rotation 2782 is 0 °, 90 °, 180 °.
Rotate by 4 kinds of rotation angles, which are 270 degrees and 270 degrees.
Then, layout information extraction 2784 extracts layout information from the character line coordinates at four types of rotation angles. Then, the layout base rotation angle evaluation 8655 evaluates the rotation angle using the layout information and the layout knowledge at each rotation angle. Then, the document rotation angle determination 2779
The rotation angle of the input document is determined based on the character recognition-based rotation angle evaluation result previously obtained or in the layout-based rotation angle evaluation result. Then, whether the image rotation is necessary is determined by the image rotation determination 2790, and when the image rotation is required, the document image is rotated using the rotation angle determined by the image rotation 2792. Then, the rotated corrected image is registered (2794). Then, the character image in the corrected image rotated by reading 2796 is converted into a character code. Then, the read result is registered (2798), and the read result is corrected (2799) according to the instruction of the user.

Next, a method of determining whether the input document is normally input on the front side or mistakenly input on the back side according to the presence / absence of character lines extracted from the input image will be described.

As shown in FIG. 9, first, a character line is extracted (900). Then, it is determined whether a character line exists (910). If the back side is input, the input document is blank and the character line is not extracted, so there is no character line. That is, if the character line exists, the next process is executed (920), and if the character line does not exist, it is determined that the input is made on the back side and the reject process is executed (93).
0). In the reject processing, the user is prompted to reject input or the document number is registered as a reject document in the reject document number file.

Next, a method of determining whether the character line to be read is outside the scanner reading area will be described.

As shown in FIG. 10, first, a protrusion flag indicating whether or not the scanner reading region is protruded is set (1000). Then, the next process is repeated several times for the character lines (1005). C, which indicates the number of overhanging character lines
OUNT is set to 0 (1010). Then, the next process is repeated four times, which is the number of vertices of the current character line (1015). At 1020, it is determined whether or not the vertex of interest exists on the scanner reading area boundary line (*, 0). If 1020 is satisfied, COUNT is incremented (1025). Here, * indicates an arbitrary numerical value. Then, at 1030, it is determined whether or not the vertex of interest exists on the scanner reading area boundary line (0, *). If 1030 is satisfied, COUNT is incremented (1035). Then, at 1040, it is determined whether or not the vertex of interest exists on the scanner reading area boundary line (Xe, *). If 1040 is satisfied, COUNT is incremented (1045). Here, Xe is the maximum X coordinate of the scanner reading area. Then, at 1050, it is determined whether or not the vertex of interest exists on the scanner reading area boundary line (*, Ye). If 10
If 50 is satisfied, COUNT is incremented (1
055). Here, Ye is the maximum Y coordinate of the scanner reading area. Then, at 1060, it is determined whether COUNT is 2 or more. Here, the fact that COUNT is 2 or more means that the character line is in the non-reading area of the scanner and the input document is protruding. If COUNT is 2 or more, the number of the character line is registered (1065). Then, the protrusion flag is set to 1 (1070). After the above processing is completed, it is determined whether or not the protrusion flag is set to 1 (10
75). If the protrusion flag is set to 1, reject processing is executed (1080).

Next, the screen display when the protruding character line is detected will be described.

As shown in FIG. 11, character lines are displayed on the window 1110 displayed on the screen 1100. At this time, in order to make it easy to recognize that the character line is protruding, the display color of the object 1120 in which the character line does not protrude and the display 1130 in which the character line protrudes are displayed in different colors.

Next, a method for the user to adjust the inclination of the input document will be described.

As shown in FIG. 12, a character line 1220, a tilt indicator 1240, its initial position 1230, and a pointing device 1250 for operating the tilt indicator are displayed on a window 1210 displayed on a screen 1200. In this case, the input image has a tilt and is a document rotated by 180 degrees.
Then, the user can directly superimpose the tilt indicator displayed in the center on the displayed character line, and can easily and accurately adjust the tilt. Then, when the adjustment is completed, the end button 1 on the operation panel 1260
The process is terminated by pressing 270 or the cancel button 1280.

Next, a method of selecting a character line by the optimum character line selection means 130 will be described.

As shown in FIG. 8, first, COUNT, which is a counter of selected character lines, is initialized (80).
0). Then, the following processing is repeated until the counter COUNT reaches the number of selected character lines (810). When the aspect ratio or the aspect ratio of the current character line is equal to or greater than a certain value (820), the character line is regarded as one of the optimum character lines (8
30). Then, the counter COUNT is incremented (840). When the number of selected characters is reached (850), the optimum character line selection means is terminated (86).
0). Here, when the aspect ratio of the current character line is equal to or greater than a certain value, it indicates a horizontally written character line, and when the aspect ratio is greater than a certain value, it indicates a vertically written character line. In this way, as a basis for judging with a certain constant value, one is that when judging the rotation angle based on the similarity of character recognition, it is necessary to judge with high accuracy, and it seems that there are multiple characters in one character line. This is because The reason is that if there are few characters in one character line, Japanese words such as "mouth" and Chinese numerals "1" are 0 degrees, 9
This is because there is no big difference in the shapes even if they are rotated by 0 degrees, 180 degrees, and 270 degrees, and it is difficult to determine the rotation angle based on the similarity of character recognition. For example, since the chapter number or the like is written as "1.1", if this is determined as a character line, the determination as to whether the rotation angle is 0 degrees or 180 degrees becomes ambiguous.

Next, the character line image rotation means 140 and the character recognition base rotation angle evaluation means 145 will be described.

First, as shown in FIG. 13, an image 1 of a plurality of character lines selected by the optimum character line selection means 130.
305 and the inclination 1300 of the input document image are input to four types of rotation processing units (1315-1325), and images of a plurality of character lines are (-a) degrees, (90-a) in each rotation processing unit.
A character line image rotated by 180 degrees, (180-a) degrees, and (270-a) degrees is obtained. Then, in 1330, character segmentation (1335) is performed on the character line image at each rotation angle.
˜1350) and character recognition (1355-1370). Then, the character similarity evaluation processing unit (1375 to 13)
In 90), the evaluation value for the character line image at each rotation angle is obtained using the similarity of character recognition. Here, as the evaluation function of the character similarity evaluation processing unit, the average value or the median value of the similarity of all characters is used. Then, the best rotation angle determination unit 1397 determines the document rotation angle 1399 from the evaluation value obtained at each rotation angle, and outputs reject information 1399 when the evaluation value is ambiguous.

Next, the character recognition based rotation angle evaluation means 14
Example 5 will be described.

As shown in FIG. 14, when determining the rotation angle of an input document, if the evaluation based on the similarity is ambiguous, the target document is treated as rejected.
The character similarity evaluation values at the obtained four types of rotation angles are input (1400). Then, the best evaluation value (a) is calculated (1410), and the evaluation value (b) of the next point is calculated (142).
0). Then, it is determined whether the absolute value of ab is larger than a certain value K (1430). If the condition is satisfied, the rotation angle of the document is determined (1440), and if not, reject processing is performed (1450).

Next, the character line coordinate rotation means 165, the layout information extraction means 170 and the layout base rotation angle evaluation means 175 will be described.

First, as shown in FIG. 15, the inclination 1500 of the input document image and the coordinates 1505 of the plurality of character lines selected by the optimum character line selection means 130 are set in four types of rotation processing units (1510 to 1525). Input the coordinates of multiple character lines in each rotation processing unit (-a) degrees, (90-a)
The character line coordinates rotated by degrees, (180-a) degrees, and (270-a) degrees are obtained. Then, the layout analysis processing unit (1530 to 1545) executes layout analysis from the character line coordinates at each rotation angle. Then, the layout collation unit (1550 to 1565) uses the layout knowledge (1570 to 1585) to evaluate the rotation angle (15).
90-1597).

Next, a method of evaluating the rotation angle using the layout information when the vertical writing / horizontal writing information of the input document is known in advance will be described.

First, as shown in FIG. 16, the difference from the case of FIG. 15 is that the layout information signal (1699) is input to the layout collating section (1650 to 1665) because the layout information is known in advance at the time of layout collation. The layout knowledge (1670 to 1685) to be collated is limited.

Next, a method of collating layout information will be described.

A document is composed of one or more sections (blocks), and the vertical writing and horizontal writing information of the attributes that the blocks constituting the document necessarily have in FIG. 17 are 0 degrees, 90 degrees, 180 degrees, and 27 degrees.
The image when rotated 0 degrees is shown. As you can see from this figure, each of the horizontal and vertical writing rotated has a feature in the indentation coordinates at which the characters begin and the coordinates at the end of the sentence that ends midway. The rotation angle can be calculated. Figure 17 in the layout knowledge
The rotation angle of the input document is identified by storing the information of (8) of (1) of (1) and checking which pattern of (1) to (8) of FIG. Will be possible.

Next, the character recognition based rotation angle evaluation means 14
A method of determining the document rotation angle that combines the result of No. 5 and the result of the layout-based rotation angle evaluation unit 175 will be described.

As shown in FIG. 18, a character recognition base rotation angle evaluation value (a) is obtained (1800). Then, the layout base rotation angle evaluation value (b) is obtained (1810).
Then, in each of the document rotation determinations (1820), when the values a and b are equal, the rotation angle is determined (1830) and the reading process is performed (1840). If the values are not equal, reject processing is executed (1850).

Next, a processing method for converting a character image in a document image into a character code by using the obtained rotation angle of the input document will be described.

As shown in FIG. 19, character line coordinates 190
5. Input the document rotation angle 1910 and the input image 1915 into the image rotation unit 1920. Then, the image rotation unit 1920 rotates and corrects the input document by the document rotation angle. Then, the character line extraction unit newly extracts a character line from the corrected image. Then, the character cutting unit 1935 cuts out the character,
Character recognition unit 1940 performs character recognition and character code 1
Converted to 955 and output.

Next, a processing method for converting a character image in a document image into a character code by using the obtained rotation angle and layout information of the input document will be described.

As shown in FIG. 20, the document rotation angle 2110.
And the character line partial image 2105 are input to the partial image rotation unit 2125. Then, the image rotation unit corrects the rotation of the input document by the document rotation angle. Further, the character line number 2114 and the layout information 2115 are input to the reading order determination unit 2120. The reading order determination unit 2120 obtains the order of character lines to be converted into character codes. Then, the rotated character line image and the reading order information are input to the character cutout unit 2130. Then, according to the order obtained by the reading order determining unit 2120, the character cutting unit 21
A character is cut out at 30, a character recognition unit 2135 executes character recognition, and a character code 1955 is output.

Next, a method of automatically reading the page number of the input document and registering it as additional information will be described.

As shown in FIG. 21, the rotation processing unit 2210.
The rotation angle 2200 and the input image 2205 are input to and the document image is rotated and corrected. In order to recognize the page number, the page number image extraction unit 2215 uses the knowledge stored in the page number layout information file 2220 to extract the page number partial image from the corrected image. Extract. Next, the page number is cut out from the extracted page number partial image by the character cutting unit 2225 one by one,
The character recognition unit 2230 converts the image data into a character code. Finally, the recognized page number is treated as additional information by the additional information registration unit 2235, and this is added to the additional information file 224.
Register to 0.

Next, the registered information will be described.
The registered data includes a document number, input image, reject number, corrected image, reading result, and additional information. As shown in FIG. 22, the additional information further includes a document name 23
05, author name 1: 2310, author name 2: 2315, magazine name 2320, source 2325, date and time of acquisition 2330, etc. are registered, and related information is linked within the same document. Then, as shown in FIG. 23, the file control means 2
440 extracts the relationship between the documents and registers it in the document relationship file 2445.

FIG. 24 shows an information file relating to the magazine name. The content of the item and the document number of the item are registered in this file. Files of other registered items have the same structure. The file control means for generating a document relation assumes that document numbers having the same item contents are related to each other when referring to files of each registered item, and lists related document numbers in order from the document number. As a result, a document relationship file as shown in FIG. 25 is created.

As described above, in the system described above, not only the information in the same document is searched, but also the related document is searched from the registered item, and even if the information about the document desired by the user is vague, the documents are not matched. Relationships can be used to obtain information about a desired document from other documents as well.

Next, a method for solving the protrusion of the document from the reading area which may occur at the time of reading with the A4 scanner will be described with reference to FIG.

Many electronic filing apparatuses are capable of reading A3 as a scanner, and a document is not input in A4 but is input in the reading area of A3 (280
0) and thereafter, the contents in the input document image are read by the same method as that shown in FIG. 7 (2805-2870).
Here, what is different from FIG. 7 is that the protrusion determination is unnecessary and the reject process accompanying it is unnecessary. Further, the processing can be easily realized because only the processing area shown in FIG. 7 needs to be expanded.

Next, in the method described above, since the A3 reading area is processed every time, the processing time is longer than that in the case of supporting A4. Therefore, first, an image is input in the A4 reading area (2900), a protrusion process is performed (2905), and a protrusion determination is performed based on the result (2900).
910). If it does, the image is re-input in the A3 reading area (2915) and the processing shown in FIG.
-2870) Here, the process 29 is executed (292).
0). If the document image does not extend beyond the reading area, processing 29 is performed on the image in the A4 reading area (2925). By the method described above, when the user wants to read the document out of the reading area, the contents can be automatically read without adjusting the input document and re-inputting each time.

Next, a method for determining whether or not the character line to be read is outside the scanner reading area will be described with reference to FIG.
Will be explained.

FIG. 10 shows a method of determining whether or not the input document is out of the reading area based on the coordinates of the four corners of the character line. Here, a method for determining whether or not the input document extends beyond the reading area by a simpler method will be shown. FIG. 29 shows that the input image (3000) has a character line (300
5) is present and it extends beyond the scanner reading area. Here, a rectangle having a width of N bits with respect to the top side, the bottom side, the left side, and the right side (for example, 3010,
3015), the peripheral distribution or the cumulative number of black dots is integrated. When the input document extends beyond the reading area, it always touches any of the four sides, so the cumulative number of black dots in any one of the four rectangles exceeds a certain value V. In the case of FIG. 29, the cumulative number of black dots in the rectangle 3015 exceeds the value V. In this way, by checking the values of the cumulative number of black dots in the four-sided rectangle, it is possible to easily determine whether or not the input document exceeds the reading area.

Next, the result of character recognition is used in determining the direction angle of the input document to be read. At this time, it is necessary to rotate the character image or the character line image at an arbitrary angle for character recognition. is there. As a method of this rotation, a method of performing rotation for each bit is conceivable, but this processing amount is not small, so a character image is extracted from the character line image,
For a method that uses a character skeleton or contour as a character recognition method, the character skeleton or contour is described in, for example, FIG.
It is expressed by the 8-direction chain code shown in 1 and can be easily realized by greatly reducing the processing amount only by shifting the chain code number by the angle required for rotation. If the number of directions of the chain cord is increased to say 16 directions or 32 directions depending on the sentence of the rotation angle or Ino, it is possible to easily rotate at a fine angle. For example, 8 for the skeletal data of the character "Ta"
This is shown in FIG. 29 in which this is expressed by the chain code of the direction.
In this way, the skeleton data can be represented by a chain in 8 directions, and rotation can be easily performed in 45 degree units.

[0107]

According to the document reading device, the electronic file device, the facsimile, the copying machine, or the computer of the present invention, the document is input by being rotated at an arbitrary angle (0 to 360 degrees) with respect to the setting direction of the scanner. Even if the input document is rotated, the rotation angle of the input document can be detected, the input image can be corrected in the correct direction, and the content can be read. Further, the corrected image can be presented or stored to the user.

Further, it is possible to judge whether the input document is mistakenly input on the back side by the user and prompt the user for a message when inputting on the back side. In addition, it is possible to determine whether or not the portion of the input document to be read extends beyond the reading area of the scanner, and if the portion is input beyond the reading area, a message can be prompted to the user.

Further, additional information such as a source, a date and time of acquisition, a memo, etc. relating to a document which is not described in the document is recorded in association with the input image and the processing result thereof, and the document information is designated.
You can search and retrieve information related to the document. Further, it is possible to record the related information of other related documents, search for the related information of the documents, search for the document from a certain document, and search for and retrieve the information related to the desired document.

Even if the input document extends beyond the A4 scanner reading area, the extension area is determined, and the document image is automatically input again in the A3 scanner reading area to read the content described in the document without leaking. You can

Furthermore, by checking the file capacity at the time of data registration, it is possible to judge whether or not the processing result is stored and prompt the user for a message. Also, a large amount of documents can be input, the rotation angle of the input image can be corrected, the corrected images can be stored, and the images can be managed at the same time.

It is another object of the present invention to provide a document reading device, an electronic file device, a facsimile device, a copying machine, or a computer in consideration of user's usability.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of a document reading device of the present invention.

FIG. 2 is a block diagram of an embodiment of a document reading device of the present invention.

FIG. 3 is a diagram showing a scanning line direction of a scanner and a processing direction of a system.

FIG. 4 is a diagram showing a state in which a document is tilted on a scanner and set.

FIG. 5 is a diagram showing a state in which a document is set upside down with respect to the processing direction of the system.

FIG. 6 is a diagram showing a state in which a document is set with a processing direction of the system and an arbitrary rotation angle.

FIG. 7 is a diagram showing a rough process of the document reading apparatus of the present invention.

FIG. 8 is a diagram showing a flow for selecting an optimum character line.

FIG. 9 is a diagram showing a flow of determining whether or not an input is made on the back side by mistake.

FIG. 10 is a diagram showing a flow of determining whether or not the reading area of the scanner is protruded.

FIG. 11 is a diagram showing a display method of a portion outside the reading area of the scanner.

FIG. 12 is a diagram showing a method for a user to set the inclination of an input document.

FIG. 13 is a diagram showing evaluation of a rotation angle of a document using character recognition.

FIG. 14 is a diagram showing a method of determining a rotation angle from four types of candidate document rotation angles.

FIG. 15 is a diagram showing evaluation of a rotation angle of a document using layout information.

FIG. 16 is a diagram showing evaluation of a rotation angle of a document using preset layout information.

FIG. 17: Vertical writing / horizontal writing blocks 0, 90, 18
It is the figure rotated 0,270 degree.

FIG. 18 is a diagram showing a method of determining a rotation angle of a document based on a rotation angle obtained by using character recognition and a rotation angle obtained by using layout information.

FIG. 19 is a diagram showing reading of the described contents by correcting the input document by the rotation angle.

FIG. 20 is a diagram showing a method of reading the described content by correcting only the rotation angle of the input document using layout information.

FIG. 21 is a diagram showing an embodiment in which a document image is rotated and corrected, a page number is recognized, and the page number is registered as additional information.

FIG. 22 is a diagram showing an example of contents registered as additional information.

FIG. 23 is a diagram showing a method of extracting and registering information of inter-document documents from files of a plurality of items.

FIG. 24 is a view showing an example of a magazine name information file.

FIG. 25 is an example of a document fellowship information file.

FIG. 26 is a diagram showing a processing flow of an embodiment of the document reading apparatus of the present invention.

FIG. 27 is a diagram showing a reading flow in which an A3 scanner is used and the A4 scanner does not extend beyond the reading area.

FIG. 28 is a diagram showing a processing flow for realizing reading by detecting an overflow from the reading area of the A4 scanner and further inputting an image in the reading area of the A3 scanner.

FIG. 29 is a diagram showing a flow for determining whether or not the reading area of the scanner is protruded.

FIG. 30 is a diagram showing a chain code used for an expression in which the character recognition target is rotated at high speed by the chain code without rotating the character image.

FIG. 31 is a diagram showing a skeleton expressed by a chain code.

[Explanation of symbols]

 101 document, 105 image input means, 110 character line extraction means, 125 inclination extraction means, 150 document rotation angle determination means, 165 character line coordinate rotation means.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Yoshihiro Shima 1-280, Higashi Koigokubo, Kokubunji, Tokyo Inside the Central Research Laboratory, Hitachi, Ltd.

Claims (21)

[Claims] 1. An input unit for inputting an image of a document, a character line extracting unit for extracting a character line of the input image, an inclination extracting unit for extracting an inclination of the document, and the extracted character line. To the inclination of the document 0 degrees, 90 degrees, 180 degrees, 270
A rotation means for rotating an angle added with a degree and a determination means for recognizing the rotated four types of character lines and determining the inclination of the document from the recognition are provided, and the inclination of the document is corrected to read. A document reading device characterized by performing. 2. The document reading apparatus according to claim 1, wherein the input means inputs all the document images by using an A3 input scanner. 3. The character line extracting means comprises means for adopting an aspect ratio of a character line or an aspect ratio of a certain value or more so that a plurality of characters are present in the character line. The document reading device according to claim 1, wherein 4. The determining means recognizes a character cut out from each of the character lines rotated by the four types of angles, and calculates an average value, a median value, and a filter of similarity degrees of a plurality of characters obtained by the recognition. The document reading device according to claim 1, wherein the document inclination is determined based on at least one of the values obtained by the multiplication. 5. The document reading device according to claim 1, wherein the determining means obtains layout information from the character line rotated by the rotating means and determines the inclination of the document based on the layout information. apparatus. 6. The document reading device according to claim 5, wherein the layout information is four types of information in which vertical writing / horizontal writing and vertical / horizontal writing information are combined. 7. A document reading apparatus according to claim 1, wherein first means for adding a document number to said document and additional information for inputting additional information not written in a document related to said document are inputted. A second means, a third means for electronically storing the input additional information in association with the document number or the document information obtained by processing the input document, and a fourth means for searching the additional information or the document information. A document reading device comprising: 8. The document reading apparatus according to claim 7, wherein the first means recognizes the page number image of the document, recognizes the extracted page number, and recognizes the extracted page number. A document reading device, comprising: means for registering a page number as additional information. 9. A document reading apparatus according to claim 7, wherein means for electronically recording relationship information between documents and means for retrieving the relationship information between the documents to display document information and additional information relating to a desired document. A document reading device comprising a searching unit. 10. The document reading device according to claim 1,
A document reading device comprising means for determining whether the input document is normally input on the front side or is input on the back side by mistake according to the presence / absence of character lines extracted from the input digital image. 11. The document reading apparatus according to claim 1,
It is determined whether the character line to be read is outside the scanner reading area according to whether two or more of the four vertices of the character row extracted by the character row extracting means are present on the four sides of the scanner reading area. A document reading device comprising a means. 12. The document reading device according to claim 11, wherein when it is determined that the character line to be read of the document image is outside the area, an error occurs in a window displaying the document image or another window. A document reading device, comprising: means for enlarging and displaying a portion or a means for displaying a character line in which an error has occurred by changing the color of the character line from that of another character line. 13. The document reading apparatus according to claim 11, wherein, when it is determined that the character line to be read of the document image is outside the area, whether the user continues to read the input image is set by a mode. A document reading device comprising means capable of performing. 14. The document reading device according to claim 11, wherein when it is determined that the character line to read the document image is outside the area, a means for registering the document number of the document image in a reject file or an error. A document reading device comprising: means for displaying a message on a window or urging a warning by voice call. 15. A device for inputting a document as a digital image, a means for displaying a free space for outputting the processing result of the digital image to a file, and a warning window when the free space becomes small. A document reading apparatus comprising means for prompting by means of a display or voice to a user and means for prompting a warning to another device in which an operator is present via a network. 16. A device for inputting a document as a digital image, wherein the cumulative number of black dots (N is an integer) of a rectangle corresponding to the upper side N bits of the inputted image data and the cumulative number of black dots of a rectangular region corresponding to the right side N bits and the left side. A means for determining the cumulative number of black dots in the rectangle for N bits and the cumulative number of black dots in the rectangle for the bottom N bits, and whether the values of the four types of cumulative black dots obtained above exceed the value V (V is an integer). A means for determining whether or not, and a document reading device, characterized in that when the above determination result exceeds a value V (V is an integer), it is determined that the written content to be read is outside the scanner reading area. 17. A character reading device for expressing a skeleton or outline of a character by a chain code in N directions (N is an integer), means for cutting out a character from an appropriate selected character line image, and the cut out character image. Means for extracting a skeleton or contour of a character from the expression and expressing it as a chain code in the N direction, and four kinds of angles 0, 90, 180 for the detected inclination.
A document reading apparatus comprising means for rotating the chain code by an angle corresponding to an angle of +270 degrees and means for recognizing the rotated chain code. 18. An electronic file device comprising: input means for inputting an image of a document; storage means for storing the input image; and output means for outputting the stored image. A character line extracting means for extracting a character line of an image, a tilt extracting means for extracting a document inclination, and the above-mentioned extracted character lines for a document inclination of 0 °, 90 °, 1
The inclination of the document is corrected by including a rotation unit that rotates an angle of 80 degrees and 270 degrees and a determination unit that recognizes the four types of rotated character lines and determines the inclination of the document from the recognition. An electronic filing device characterized by performing reading. 19. An electronic file device comprising input means for inputting an image of a document, storage means for storing the input image, and output means for outputting the stored image. A character line extracting means for extracting a character line of an image, a tilt extracting means for extracting a document inclination, and the above-mentioned extracted character lines for a document inclination of 0 °, 90 °, 1
The inclination of the document is corrected by including a rotation unit that rotates an angle of 80 degrees and 270 degrees and a determination unit that recognizes the four types of rotated character lines and determines the inclination of the document from the recognition. An electronic filing device characterized by performing reading. 20. A facsimile device comprising an input means for inputting an image of a document and a transmitting means for transmitting the input image, a character line extracting means for extracting a character line of the input image, An inclination extracting unit that extracts the inclination of the document, and the extracted character line to the inclination of the document is 0 degrees,
Rotating means for rotating an angle obtained by adding 90 degrees, 180 degrees, and 270 degrees, and determination means for recognizing the rotated four types of character lines and determining the inclination of the document from the recognition are provided. A facsimile machine characterized by correcting inclination and reading. 21. A character line extracting means for extracting a character line of the input image in a copying machine having an input means for inputting an image of a document and an output means for outputting the accumulated image, Inclination extraction means for extracting the inclination of the document, and the above extracted character lines to the inclination of the document are 0 °, 90 °, 1
The inclination of the document is corrected by including a rotation unit that rotates an angle of 80 degrees and 270 degrees and a determination unit that recognizes the four types of rotated character lines and determines the inclination of the document from the recognition. A copying machine characterized by scanning and reading.