JPH11102413A - Pop-up correction method for optical character recognition output and device thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To attain comparison for a text, together with its optical character recognition interpretation by recognizing the characters from a document image to decide areas corresponding to a word of a document text and displaying the part of the document image, after securing the relation between the decided area and the corresponding word of the document text based on a correlation table. SOLUTION: The image of a certain document is generated by a scanner device 126, and an optical character recognizer 128 recognizes the characters included in the document image for generating a document text. A processor 112 decides an area for the document image, corresponding to a word of the document text, secures a relation between the decided area of the document image and the word corresponding to the document text, based on a correlation table and shows the document image part on the document text via a display device 120. Then it is preferably to display the area of the document image corresponding to the word of the document text and/or the corresponding word of the document text, so as to show each recognition likelihood parameter.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to optical character recognition technology, and more particularly, to a method and apparatus for displaying an optical character recognition output and correcting an error thereof.

[0002]

BACKGROUND OF THE INVENTION Obtaining text and graphics from paper documents is a significant problem for many industries. For example, a publisher may print hundreds or thousands of scholarly articles throughout the year. Often a publisher starts with a paper document that must be entered into the publisher's computing device. One conventional approach is to employ a keyboard enthusiast to read a paper document and type the document into a computer system. However, entering a document is time consuming and costly.

[0003] Optical character recognition (hereinafter referred to as OCR)
Is a technology that guarantees benefits to the publishing and other industries. The reason is that the input processing speed of the OCR device is much higher than the input speed of the keyboard input person. Thus, publisher employees often begin work with scanned documents. The document has been converted by the OCR device into a computer readable text format, for example, ASCII. However, even the high recognition rates possible with modern OCR devices (often above 95%) are not sufficient for industries such as the publishing industry that require high accuracy. Thus, publishers often employ proofreaders to manually modify OCR output.

[0004]

However, manually calibrating the OCR output is very time consuming and difficult for humans to perform. The proofreader must compare the original paper document with the printed or screen display of the OCR output and compare them word by word. Even if the recognition rate is high, when a person calibrates the OCR output, one person tends to miss and miss an error.

Another conventional option is to spell check the resulting computer readable text. However, it does not recognize all misspelled words. In addition, the entered words may be very distorted, so the proofreader must refer back to the paper text throughout the entire spell check. Once the proofreader looks at the text on the paper, he determines the correct word, and then taps the correct word into the OCR output text. This technique is time consuming and has proven to be somewhat error prone, so the proofreader does not need to refer to the original document used to generate the OCR interpretation,
It would be useful for a proofreader to be able to compare text representing a document image along with the OCR interpretation of that text.

[0006] Viewing a document image along with its OCR interpretation of the text requires that publishers use ASC rather than paper form.
It is particularly useful when trying to reprint and sell OCR output in the form of II text. Publisher OC in electronic form
When obtaining the OCR output for the purpose of reselling the R output, not only does the OCR output contain the correct word, but also the form of the OCR output changes when the OCR output is later displayed on a computer monitor. There is an additional relationship that the form remains the same. The ability of the proofreader to compare the OCR output and the document image side-by-side during the editing phase greatly facilitates this purpose.

It is an object of the present invention to enable a user to compare text consisting of a document image along with an OCR interpretation of the text.

Another object of the present invention is to eliminate the need for the user to reference the original document used to generate the OCR interpretation, and to allow the user to combine the text represented by the document image with the OCR interpretation of that text. Is to be able to compare.

It is yet another object of the present invention to provide a method for a user to convert text represented in a document image to an OCR interpretation of the text in order to correct errors made during the conversion of the original text to text in the OCR output. Is to be able to compare.

[0010]

SUMMARY OF THE INVENTION There is a need to facilitate human calibration of OCR output. To meet this need, the characters of the document image obtained from the original paper document are used to generate the document text (eg, O
(Via CR). A region of the document image corresponding to the region of the document text is determined, and a recognition accuracy parameter is determined for each region. The user can select the word from the document text by positioning the cursor over the word. When the user clicks (presses) one button of the mouse or presses a similar function key, the portion of the document image corresponding to the selected word appears as a pop-up window. If the user clicks another mouse button or presses another similar function key, a pop-up menu for the corresponding OCR output is displayed.

In particular, a computer-implemented method for displaying a document image portion on document text includes the steps of generating a document image of a document, recognizing characters from the document image to generate the document text. ,
Determining a region of the document image corresponding to the word of the document text, associating the region of the document image with the corresponding word of the document text using a correlation table, and displaying a document image portion on the document text Are combined. The selected text of the document text is then corrected.

[0012] These and other aspects and advantages of the present invention will become apparent with reference to the following description, drawings, and claims.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a method and apparatus for pop-up correction of an optical character recognition output according to the present invention will be described in detail with reference to the drawings. Signs are attached.

[1. FIG. 1 is a block diagram of a computer system 100 on which an example of the present invention can be implemented. Computer system 100 includes a bus 110 or other communication means for communicating information, and a processor 112 for processing information.
Are connected to the bus 110. The computer system 100 further includes a random access memory (RAM) or other dynamic storage device 114 (shown as main memory),
4 is connected to a bus 110 for storing information and instructions to be executed by the processor 112. Main memory 114 may also be used to store temporary variables and other intermediate information while processor 112 is executing instructions. Computer system 100 also includes one or both of read-only memory (ROM) and other static storage devices 116, which are connected to bus 110 and store static information and instructions for processor 112.
The data storage device 118 is, for example, a drive for a magnetic disk or an optical disk and a disk corresponding thereto,
It may be connected to bus 110 for storing information and instructions.

An input / output device can be connected to the computer system 100 via a bus 110. For example, the computer system 100 uses a display device 120, such as a cathode ray tube (CRT), to display information to a computer user. Further, the computer system 100 uses a keyboard 122 and cursor control means 124 such as a mouse. In addition, computer system 100 may use a scanner 126 to convert paper documents into a computer readable format. Furthermore, the computer system 100 can use the OCR device 128 to recognize characters in a document image generated by the scanner 126 or in a document image stored in the main memory 114 or the data storage device 118. Alternatively, the functions of OCR device 128 may be implemented in software by executing instructions stored in main memory 114 with processor 112. In yet another example, the scanner 126 and the OCR device 128 can be combined into a single device designed to scan a paper document and recognize the characters there.

The present invention relates to using the computer system 100 to view the original text and the output text on the same display device 120. According to one embodiment, this task is performed by computer system 100 in response to processor 112 executing a sequence of instructions stored in main memory 114. Such instructions may be stored in main memory 11 from another computer-readable medium, such as data storage device 118.
4 may be read. By executing a series of instructions stored in the main memory 114, the processor 112 performs the processing steps described later. In another example, hardwired circuitry may be used in place of, or in conjunction with, software instructions to implement the present invention. Accordingly, the present invention is not limited to any particular combination of hardware circuits and software.

[2. Synthetic Document Architecture] A synthetic document has many representations of a document, and treats the many representations as a logical whole. The composite document 200 shown in FIG. 2 is stored in a memory such as the main memory 114 or the data storage device 118 of the computer system 100, for example.

The composite document 200 includes a document image 210, which is a bitmap representation of a document of the document (eg, a TIFF file generated from the scanner 126). For example, a copy of the United States Constitution may be read and scanned by scanner 126 to produce an image of the United States Constitution in the form of a document image 210.

A bitmap representation is a row of pixels,
Monochrome (eg black and white) or multicolor (eg red, blue,
Green, etc.). The position of the rectangular area of the document image 210 can be specified, for example, by combining the upper left corner and the lower right corner of the rectangle. In the example of reading and scanning the United States Constitution, the first letter of the word “form” in the preceding sentence (ie, “f”) is a rectangle with coordinates (16,110) in the upper left and (31,119) in the lower right. May be arranged in the document image 210 in the inside. And the last character of the same word (ie, "m") is (16,140)
, And the lower right corner can be arranged in the rectangular document image 210 having the coordinates (31, 149).

The composite document 200 is composed of the document text 22
0 and a correlation table 230, which may be generated by the method shown in the flowchart of FIG. The document text 220 is encoded ASCII, E
It is made up of a series of 8-bit or 16-bit bytes that encode characters in BCDIC or Unicode. Thus, characters in the document text 220 may be located within the document text 220 by offset. In the above example, as represented in the offset column of the correlation table 230, the first letter of the word "form" in the preamble may be placed in the document text 220 at offset 57, and the last letter of the same word May be placed in the document text 220 at offset 60.

Referring to FIG. 3, step S25 will be described.
0, the characters in the document image 210 are
8 or equivalent, and is saved to generate document text 220 in step S252. The OCR device 128 is designed to output the coordinates of the recognized character in the document image 210 in step S250. Thus, characters recognized at known offsets in the document text 220 may be associated with regions of the document image 210. In the above example of the preamble image, the document text 220
The first letter of the word "form" (located at offset 57) has coordinates (16, 110) and (31,
119) is associated with the document image 210 area. Similarly, “fo” of the document text 220
rm "(located at offset 60) are the coordinates (16,140) and (31,14).
9) is associated with the document image 210 area defined by the above.

In step S254, the document text 220
Is specified by, for example, interpreting characters between spaces as words. In step S254, the area of the document image 210 corresponding to each character of these words is merged into a larger document image 210 area corresponding to each word of the document text 220. In one embodiment, the area of the document image 210 is specified as a rectangle having the upper left and lower right coordinates of the area corresponding to the individual words of the document text 220. For example, the area of the document image 210 corresponding to the word "form" (offset 57-60) of the document text 220 has the coordinates (16, 110) and (31, 149).
Alternatively, especially for documents having characters of various sizes, the coordinate list for each character of the document text 220 and the corresponding document image 210 area may be stored individually.

In addition, some embodiments of the OCR device 128 may recognize, as is well known in the art, the possibility of words or phrases in the document text 220 containing incorrect OCR interpretations. It is designed to output accuracy parameters. For example, in a font, the character "m" in the document image 210 may be recognized by the OCR device 128 as a concatenation of characters "rn" (e.g., because the OCR device may interpret the word as "modern", OCR device 128 may output a low likelihood parameter for the word "modem"). As a result, words containing the letter "m" appear to be assigned a relatively lower probability than words consisting entirely of features. In the example of the above preamble, the word “form” may be assigned a recognition accuracy parameter of 55% because the word “m” is included in the word.

In step S256, the document text 220
Are stored in the correlation table 230 so that regions of the document image 210 can be associated with words of the document text 220. In particular, the correlation table 230 stores a coordinate pair 232 specifying an area in the document image 210, an offset pair 234 specifying a word in the document text 220, and a recognition accuracy parameter 236 for the word. In the above example, the word “form” in the document text 220 is (16, 11
0) and (31,149) coordinate pairs 232, 57 and 60 offset pairs 234, and a 55% recognition accuracy parameter 236.

Using the correlation table 230, each offset in the document text 220 is
, And vice versa. For example, given a character in document text 220 at offset 58, to determine that the character falls into a rectangular area in document image 210 having coordinates (16,110) and (31,149), The offset field of the correlation table 230 can be consulted. Then, the regions in the document image 210 at those coordinates (the word “form” in the example above) can be derived from the document image 210 and displayed. On the other hand, given the coordinates (23,127) of the document image 210, to determine that the coordinates of the given document image 210 are found within words of the document text 220 having an offset of 57-60. Next, the coordinate field of the correlation table 230 can be examined. The word in the offset region of the document text 220 (the word "form" in the example above) can then be identified. Thus, the composite document architecture described herein provides one way to relate the placement of words in the document text 220 to corresponding areas of the document image 210.

[3. Pointing out words that have the possibility of misrecognition)
In the image displays 300 and 350 of FIGS. 4 and 5, the words in the document text 220 that are most likely to be misrecognized are identified by various methods (for example, brightening, changing colors and fonts, and underlining). , Flash, etc.). These words can be determined by comparing the recognition accuracy parameters 236 of all recognized words to a specified threshold. For example, the recognition accuracy parameter 23
Words in which 6 is 60% or less are displayed in red, and can draw the user's attention to potentially incorrect words in the text. For example, the original word "form" results in a 55% recognition accuracy parameter 236, which is displayed in red. In another example, words with a low recognition accuracy parameter 236 are identified from the document text 220 by changing the background color of the word (eg, the word “form” in the document text 220 is highlighted in a clearly visible color). obtain).

In another example, the recognition accuracy parameter 236 further associated with each word of the paper text may include the document text 2
An individual display color for each of the twenty words is determined and compared to a plurality of thresholds to form a "heat map" of the recognized word. The heat map is a chart showing values of parameters (for example, frequency, temperature, or recognition accuracy) at various points in the spectrum using a plurality of colors. The resulting heat map helps guide the user to the most problematic portions of document text 220 for OCR output. In this example, the words of the document text 220 displayed to the user are represented in various colors.

Referring to FIG. 6, a heat map is generated for the document text 220 by a loop controlled in step S410. Step S410
Is the document text 22 to be displayed on one or both of the image display 300 and the image display 350.
Loop over all 0 words. Step S420
Then, the correlation table 230 is consulted to find a recognition accuracy parameter 236 corresponding to the displayed word of the document text 220. Then this parameter 236
Is subsequently compared to a plurality of thresholds, such as 60%, 80% and 90%.

Steps S422-S434 show a process of generating a heat map display when the threshold values are set to 60%, 80% and 90%, for example. First, the lowest threshold of 60% is used as a threshold for comparison. If the recognition accuracy parameter 236 is lower than the threshold, the color of the word is set to red (step S424). In the above example, the word “form” is the recognition accuracy parameter 2
Because 36 is 55%, it is highlighted in red. In the examples shown in FIGS. 4 and 5, another word set to red is "general
"And" Constitution ".

Next, in step S426, the next lower threshold value of 80% is used as a comparison threshold value. If the recognition certainty parameter 236 is lower than the threshold, the color of the word in the document text 220 is set to green (step S428). In the above example, the word "Union" may have a recognition accuracy parameter 236 of 75%, in which case it is displayed in green. In the examples shown in FIGS. 4 and 5, the other words set to green are “insure” and “se
cure ".

In step S430, the last threshold value of 9
0% is used as a threshold for comparison. If the recognition accuracy parameter 236 is lower than the threshold, the word color of the document text 220 is set to blue (step S43).
2). In the examples shown in FIGS. 4 and 5, the words set in blue may be “Tranquility” and “establish”. On the other hand, if the recognition accuracy parameter 236 is higher than all thresholds, the color of the word in the document text 220 is set to black, which can be used as a default color (step S434). Once the color is set, the document text 22
The word 0 is displayed in that color (step S436).

It will be appreciated that the numbers and colors for the thresholds may vary from embodiment to embodiment without departing from the spirit of the invention. For example, one threshold, 2
One, three, or even ten. As another example, the color selection may change (e.g., red, orange,
yellow). Actually, display attributes other than the display color such as blinking and underlining may be employed. It is also understood that the thresholds and display colors or other display attributes may be entered in one table and continually examined in one loop without strictly organizing the branches as shown in the flowchart of FIG. Can be done.

[4. Show Document Image Window] To reduce the time associated with browsing the original paper document,
The scanned image portion of the original paper document (ie, document image 210) is displayed on the OCR interpretation of the text. In the example of reading and scanning the United States Constitution, the scanned image portion of the preamble is
As shown in FIG. 5, it may be displayed in the image display 350 of the window on the OCR output.

In the image display 350, the document text 220 is displayed on a monitor as shown in FIG.
The user then selects a word from document text 220 by positioning cursor 360 over any word in document text 220. When the user clicks a button with the mouse or presses a similar function key, a portion of document image 210 corresponding to the area surrounding the selected word appears as pop-up window 390. This allows the user to immediately view the document image 210 when needed.

Document image 210 and document text 220
The coordinated motion with the document image 210 is determined by using the correlation table 230 to determine the location of each word in the document text 220.
This is achieved by relating to the corresponding region from Based on the information provided by the cursor control means 124, the position of the cursor 360 at any instant on the document text 220 is determined, and the mapping of the document text 220 from the coordinate system of the image display 350 is performed by mapping techniques well known in the art. It can be converted to an offset system. Then, using the correlation table 230, the offset of each word appearing in the document text 220 can be associated with coordinates for a corresponding region of the document image 210. The portion of the document image 210 that includes the corresponding region is then extracted so that the image portion can be displayed in a pop-up window 390. The user can then view the displayed portion of the document image 210 to make sure that the corresponding word in the document text 220 matches the document image 210.

In another example, a pop-up window 39
The portion of the document image 210 displayed in 0 is heat mapped in the same manner used to heat map the document text 220. Document text 22
In order to associate the display state of the word 0 with the corresponding area of the document image 210, the same recognition accuracy parameter 236 used to clearly indicate the display state of a particular word in the document text 220 is used. It is used to clearly indicate the display state of the area to be displayed. For example, by assigning a 55% recognition accuracy parameter 236 to the word "form" in the correlation table 230, the word is displayed in red in both the document text 220 and the corresponding area of the document image 210. Becomes In another example, both document text 220 and document image 210 are not displayed in a heat map, and only a portion of document image 210 is displayed in a window on document text 220.

[5. Error Correction of OCR Output] FIG. 6 is a flowchart showing an OCR output according to an embodiment of the present invention.
A process for correcting an output error is also shown. To make corrections, the cursor 310 may be, for example, a mouse, trackball, joystick, and pop-up window 3.
The mouse button used to display 90 or a mouse button other than the function key or a mouse button or a cursor control means such as a function key is disposed on any part of the document text 220 by using the cursor control means 124.

At step S440, the processor 112
Receives input from the cursor control means 124 regarding the position of the cursor 310 on the image display 300. This input may be automatically generated by the cursor control means 124 whenever the cursor 310 is placed on the image display 300 or only when the user operates a button. In the latter case, when the user operates the button, the cursor control means 124 sets the cursor 3
Send 10 current locations as input.

The position of the cursor 310 associated with the input received in step S440 is determined by using a mapping technique well known in the art.
00 from the coordinate system of 00 to the offset system of the document text 220. In the example shown in FIG. 4, the position of the cursor 310 in the image display 300 may correspond to the offset 59 of the document text 220.

In step S442, a description specifying offset pair 234 including the offset obtained from the input received in step S440 is obtained, and the correlation table 2
30 is examined. In the above example, offset 59 is included in offset pair 57-60. This offset pair is
Used to extract strings placed in document text 220 at offsets within offset pair 234.

At step S444, the offset 57-6
A possible replacement for the string at 0 is generated. A wide variety of techniques are known in the art for generating possible replacements, but do not require any particular technique to practice the invention. For example, word-level responses may be considered (eg, spell-checked) to generate potential replacements. As yet another example, phrase-level information (eg, a Markov model of continuous words in a database)
Can be used. Furthermore, these various techniques can be combined and weighted. In the above example, step S
444 is a set of the following possible replacements for the selected text "domestic": "dominat
"e", "demeanor" and "demotion" may be generated.

At step S 446, possible replacement words for the selected text are displayed in pop-up menu 330 near cursor 310. These replacement words are displayed in a pop-up menu 330 in an order according to the potential replacement of the selected text.
(I.e., if the selected text is deemed to be incorrect, the replacement word at the top of the list in the pop-up menu 330 is likely to be used as the most replacement word) ). In one embodiment, a delete option is also provided in the pop-up menu 330 near the cursor 310 to allow the user to delete a portion of the document text 220 without pause.

According to another example, if the cursor 310 is over a word in the document text 220, a pop-up menu 330 for the selected text is automatically displayed. Thus, the user can move the cursor 310 over the text display column of the document text 220 and move the selected text to the pop-up menu 33.
It can be compared quickly with possible replacement words in 0.

When the pop-up menu 330 is displayed, the user looks at the pop-up window 390 containing the portion of the document image 210 and looks at the document text 220
You may decide that the selected text is incorrect. In this case, the user will see the possible replacement words in the pop-up menu 330 to select the correct replacement word. Once the correct replacement word is found, the user can select the correct replacement word (eg, highlight the appropriate word and click or release the button on cursor control means 124). In the above example, the correct replacement for the word "domestic" is
It may be "demeanor" displayed between "dominate" and "demotion" in 0.

At this point, as in step S448, processor 112 receives the input for the intended correction and replaces the words in document text 220 with the user-selected correction as in step S450. However, if there is no correct replacement word in pop-up menu 330, the user may enter the correct replacement word in a conventional manner (eg, via keyboard 122). By generating the potential replacement words and displaying them in the pop-up menu 330, the time spent on corrections to the OCR output is reduced.

Once the user makes corrections to the document text 220 or in some way
, The correlation table 230 must be updated to reflect that this action has taken place.
In addition, the correct word recognition accuracy parameter 236 of the document text 220 is automatically reset to 100% and the selected text in the document text 220 returns to the default color (eg, black).

Although the present invention has been described and illustrated in considerable detail with reference to certain preferred embodiments, other variations are possible. After reading the above description, it will be apparent to one skilled in the art that modifications in the above description or figures may be made in form or detail without departing from the spirit or scope of the invention.

[0048]

As described above, in the method and apparatus for pop-up correction of optical character recognition output according to the present invention, a user compares a text consisting of a document image with an OCR interpretation of the text. It has an effect that can be done. Also, the user does not need to refer to the original document used to generate the OCR interpretation,
This has the effect that the user can compare the text represented by the document image with the OCR interpretation of the text. In addition, to correct mistakes made during the conversion of the original text to the text of the OCR output, the user may replace the text represented by the document image with the OCR of the text.
It has an effect that can be compared with CR interpretation.

[Brief description of the drawings]

FIG. 1 is a high-level block diagram illustrating a computer system on which the present invention can be implemented.

FIG. 2 is a block diagram illustrating a composite document architecture.

FIG. 3 is a flowchart illustrating a process of generating a composite document.

FIG. 4 is a diagram showing an example of a screen display according to an embodiment of the present invention.

FIG. 5 is a diagram showing an example of a screen display according to another embodiment of the present invention.

FIG. 6 is a flowchart illustrating error detection and correction processing in an OCR output according to an embodiment of the present invention.

[Explanation of symbols]

 112 processor 120 display device 122 keyboard 124 cursor control means 126 scanner device 128 optical character recognition device 210 document image 220 document text 230 correlation table 236 recognition accuracy parameter 330 pop-up menu

Claims (18)

[Claims] 1. A method for displaying text, comprising: generating a document image of a document; recognizing characters from the document image to generate a document text; Determining a region of the document image; associating the region of the document image with a corresponding word of the document text using a correlation table; displaying a portion of the document image on the document text; A method for pop-up correction of an optical character recognition output, comprising: 2. The method for pop-up correction of optical character recognition output according to claim 1, wherein the regions of the document image are displayed to indicate respective recognition accuracy parameters. 3. The corresponding word of the document text is:
The method for pop-up correction of an optical character recognition output according to claim 1, wherein each recognition accuracy parameter is displayed to indicate the recognition accuracy parameter. 4. The optical character of claim 1, wherein both the area of the document image and the corresponding word of the document text are displayed to indicate respective recognition accuracy parameters. A method for pop-up correction of recognition output. 5. An input for selecting a location in the text of the document, determining a selected text corresponding to the location of the text of the text, and input for correcting the selected text. 2. The method of claim 1, further comprising: receiving the selected text, and updating the correlation table with information including up-to-date information to reflect corrections made to the selected text. For pop-up correction of the optical character recognition output of a computer. 6. The optical character recognition output pop-up of claim 5, wherein receiving an input to correct the selected text comprises deleting the selected text. Method for correction. 7. The method of claim 1, further comprising: receiving one or more replacement words for the selected text; determining one or more replacement words for the selected text; and receiving one or more replacement words for the selected text. Displaying the replacement word, receiving input indicating a replacement word for the selected text, and replacing the selected text with the replacement word. 5. The method for pop-up correction of the optical character recognition output according to 5. 8. The method for pop-up correction of optical character recognition output according to claim 7, wherein receiving an input indicating a replacement word comprises receiving a keyboard input of the replacement word. . 9. The method for pop-up correction of optical character recognition output according to claim 7, wherein the one or more replacement words are displayed in a pop-up menu. 10. An apparatus for displaying text, comprising: a scanner device for generating a document image of a document; an optical character recognition device for recognizing characters in the document image for generating document text; A processor for determining a region of the document image corresponding to the word of the document text, and associating the region of the document image with a corresponding word of the document text using a correlation table; A display device for displaying a portion of a document image, and a device for correcting a pop-up of an optical character recognition output. 11. The apparatus according to claim 10, wherein the display device displays the area of the document image so as to indicate respective recognition accuracy parameters. Equipment. 12. The optical character recognition output pop-up correction of claim 10, wherein the display device displays the corresponding words of the document text to indicate respective recognition accuracy parameters. Equipment for. 13. The display device of claim 10, wherein the display device displays both the area of the document image and the corresponding word of the document text to indicate respective recognition accuracy parameters. For pop-up correction of the optical character recognition output of a computer. 14. The system further comprising: cursor control means for receiving an input for selecting a location in the document text; wherein the processor determines a selected text corresponding to the location of the document text; 11. The method of claim 10, further comprising receiving an input for correcting the selected text, and updating the correlation table to include the latest information to reflect the correction made to the selected text. For pop-up correction of the optical character recognition output of a computer. 15. The optical character recognition output of claim 14, wherein the processor receives an input to correct the selected text by deleting the selected text. Device for pop-up correction. 16. The processor, wherein the processor determines one or more replacement words for the selected text, and controls a display device to display the one or more replacement words for the selected text. Receiving input indicating a replacement word for the selected text, and receiving input for correcting the selected text by replacing the selected text with the replacement word. 15. An apparatus for pop-up correction of an optical character recognition output according to claim 14. 17. The apparatus according to claim 16, further comprising a keyboard for inputting the replacement word for the selected text. 18. The optical character recognition output of claim 16, wherein the display device is controlled to display the one or more replacement words in a pop-up menu. Equipment for.