JPH0459669B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to an optical character reading device that reads characters and graphics recorded on a document.

[Technical background of the invention and its problems] In recent years, optical character reading devices that can read documents such as books, newspapers, and drawings containing a mixture of characters and graphics have been developed. Such a device requires the ability to identify characters and graphics, output character codes for characters after recognition processing, and read graphics as image data. However, with conventional devices, it is difficult to fully realize the above-mentioned functions, and a method that requires operator assistance for some processing has been adopted.

Specifically, there is a method in which, for example, format control information necessary for character detection and cutting processing is specified using a tablet or mark before reading starts. With this method, it is necessary to specify format control information for each sheet to be read.
The operator's work burden increases. Furthermore, there is a method in which the recognition result after character recognition processing is displayed on the screen of a CRT display device, and an operator compares the displayed result with the original document, and if there is a difference, performs correction processing. Furthermore, there is a method in which a scanned image obtained by scanning a sheet to be read (including photoelectric conversion) and a recognition result are simultaneously displayed, and an operator compares the two and makes corrections. In each of these methods, the final result after recognition processing is modified. For this reason, if the character detection/cutting process or the character/graphic identification process malfunctions, the number of misreadings or unreadable texts may increase significantly, and the time required for correction by the operator may increase.
Further, since the format of the original sheet to be read and the displayed reading result are usually different, it becomes difficult for the operator to collate the sheets, and errors are likely to occur.

Therefore, in the conventional optical character reading device, there is a problem in that the burden of support work on the operator is large and the efficiency of the reading process is reduced.

[Object of the Invention] An object of the present invention is to provide an optical character reading device that can reduce operator support work and realize highly efficient and reliable reading processing in a reading device for documents containing a mixture of characters and figures. It is about providing.

[Summary of the Invention] The present invention is provided with a detection/cutout means that executes detection/cutout processing of characters and figures from image data in an image memory and creates detection/cutout information at each detected/cutout position. Characters and figures corresponding to the cutoff information created by the detection cutout means and the image data output from the image memory are configured to be displayed by the display means. Further, the operator operates the confirmation/correction means based on the cutout information displayed on the display means, thereby confirming or correcting each detected cutout position of the characters and figures in the image memory. The character recognition means recognizes the character detected and cut out by the detection cutout means at the detected cutout position determined by the confirmation and correction means.

With such a configuration, an operator can efficiently perform reading processing on a document containing a mixture of characters and graphics simply by operating the confirmation and correction means.

[Embodiment of the Invention] An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of an optical character reading device according to an embodiment of the present invention. 1st
In the figure, a photoelectric conversion unit 10 scans a document to be read and converts characters and figures recorded on the document into image signals. An image buffer memory (hereinafter referred to as image buffer) 11 stores image data consisting of an image signal output from the photoelectric conversion section 10. The detection cutting section 12 identifies characters and figures from the image data in the image buffer 11, and performs detection cutting (cutting) processing of the characters and figures. The character recognition unit 13 executes recognition processing for the characters formed by the detection and extraction unit 12 and outputs a character code as a recognition result to the editing unit 14. The image compression section 16 compresses the image data of the figure detected by the detection cutout section 12 and outputs the compressed image data to the editing section 14 .
The editing section 14 edits the recognition results from the character recognition section 13 and the graphics from the image compression section 16 into a preset format. The output device 15 is a printer that prints the characters and figures edited by the editing section 14 or a magnetic disk device that stores them.

On the other hand, a first data buffer memory (hereinafter referred to as data buffer) 17 is a memory that stores cutout information indicating the cutout position of characters and figures created by the detection cutout section 12. The second data buffer 18 stores the recognition results output from the character recognition section 13. The display control section 19 synthesizes or selects the contents stored in the data buffers 17 and 18 and the image buffer 11 and displays them on the screen of a CRT display device (hereinafter referred to as CRT) 20.
The coordinate input device (mouse) 21 is a device for specifying and inputting position coordinates on the screen of the CRT 20 . The keyboard 22 is an input device for inputting various data. Further, the control section 23 is a circuit that controls the operation of the entire device.

In the optical character reading device having such a configuration, the operation of this embodiment will be explained with reference to the flowchart shown in FIG. First, step S1 in Figure 2
As shown in FIG.
When the image is scanned, for example, one page's worth of characters and graphics are converted into image signals. Image data consisting of an image signal is stored in an image buffer 11. Next, in step S2, the detection cutting section 12 starts the inspection processing. In the detection cutting section 12,
Characters and figures are identified from the image data in the image buffer 11. This identification process is normally executed by, for example, a method using a peripheral distribution histogram of image data or a method using a figure extraction method based on run length. Furthermore, the detection cutting section 12 detects the character line, character size, character pitch, etc. of the identified character image, and classifies each character. At this time, as shown in step S3, the cutting position information (hereinafter referred to as cutting frame information) created when each character is segmented is transferred from the detection cutting section 12 to the data buffer 17.
is stored in Similarly, cutting frame information created when cutting out a figure is also stored in the data buffer 17.

When the operation of the detection cutout section 12 is completed, the display control section 19 performs the operation of step S4. That is, the display control section 19 reads out image data from the image buffer 11, and reads out image data consisting of characters 30 and graphics 31 obtained by scanning a document as shown in FIG. 3a. Furthermore, the display control section 19
reads out the cutting frame information from the data buffer 17 and displays the characters 30 and figures 3 as shown in FIG. 3b.
1 and the corresponding cutout frames 32 and 33 are superimposed and displayed on the screen of the CRT 20. At this time, the display control unit 19 may, for example, display each cutout frame 32, 33 corresponding to a character and a figure in a different color, or display each character separately depending on whether the orientation of the character 30 is vertical or horizontal. Display in color.

In the next step S5, the operator uses the CRT20
Check the cutting result of the detection cutting section 12 using the cutting frame. That is, the operator determines the cutout position of characters and figures based on the position of the cutout frame, and if the cutout position is appropriate, inputs a confirmation instruction signal from the keyboard 22 to the control unit 23. If the determination result shows that the detection position is incorrect, the operator operates the mouse 21 to correct the detection position on the screen of the CRT 20. In this modification process, e.g.
A function selection image is displayed at the edge of the CRT20 screen, and the operator can move, delete, or
Select from functions such as adding, changing character direction, and excluding from reading targets. Thereafter, the operator specifies a change position of the inspection position using the mouse 21. The instruction to change the cutting position is transferred from the display control section 19 to the control section 23, and further transferred from the control section 23 to the detection cutting section 12. Thereby, the detection cutout section 12 modifies the cutout frame information stored in the data buffer 17.

When the confirmation and correction work by the operator is completed, the detection cutout section 12 performs a cutoff position for characters and figures from the image buffer 11 based on the newly determined cutoff information in the data buffer 17. The image data for each character detected by the detection cutout section 12 is output to the character recognition section 13 for recognition processing (step S6). Further, the image data corresponding to the figure cut out from the detection cutout section 12 is compressed by the image compression section 16 and then transferred to the editing section 14 . The character recognition unit 13 outputs a character code, which is a recognition result, to the editing unit 14. Here, the character recognition unit 13 stores the recognition result in the data buffer 18. The display control unit 19 reads the recognition results from the data buffer 18 and displays them on the CRT 20.
to be displayed on the screen. The operator selects the displayed recognition result (for example, it consists of multiple candidate characters)
For example, in case of reject, press mouse 2.
1 selects from a plurality of candidate characters. In case of misreading, the operator can input corrected characters from the keyboard to the displayed recognition result.
Perform correction work such as deleting characters. This results in
Confirmation and correction processing for the recognition result shown in step S7 is performed. The editing section 14 includes a character recognition section 13
The confirmation result outputted from the image processing unit 15 (confirmation result after correction processing in step S7) and the figure from the image compression unit 16 are edited into a preset format and then transferred to the output device 15.

Here, the image displayed on the screen of the CRT 20 by the display control unit 19 may be, for example, a portrait document as shown in FIG. 4a or a landscape document as shown in FIG. 4b. FIG. 4c is an example of a display in which a character image and a cutting frame are overlapped, and FIG. 4d is an example of a display of a circumscribing frame when the character is in italics. In addition to performing color display by distinguishing colors according to the circumscribed frames of characters and figures as described above, the display may be distinguished by changing the screen display over time. Further, in step S5, after the correction of the inspection position is completed, a function may be provided for specifying the output order of the character recognition results and figures to the outside using the mouse 21 or the keyboard 22. For example, step S5 above
As shown in FIG. 5, on the screen on which characters, figures, and their cutting frames are displayed on the CRT 20, use the mouse 21 to designate and display output order designation frames 50 to 52, and to display the output order. Specify the output order between order specification frames. in particular,
If you specify the output order in the order of 50 to 53,
In this order, the characters or figures each surrounded by the output order designation frame are output to the outside.

As described above, character recognition processing is performed after displaying the cutout processing results (cutout information) in the detection cutting section 12 and performing confirmation and correction processing, so that unrecognizability and misreading can be largely prevented. can. Therefore,
The operator's work to correct the recognition results can be reduced. In addition, the operator
Since the inspection results are confirmed and corrected while looking at the image and cutting frame displayed on the screen of 0, the confirmation and correction work by the operator is extremely easy.

[Effects of the Invention] As described in detail above, according to the present invention, in an optical character reading device that reads documents containing a mixture of characters and figures, the reading rate can be increased by the operator's simple confirmation and correction work based on inspection information. can be improved. Therefore, as a result, the operator's support work for the final confirmation result can be reduced, and highly efficient and reliable reading processing can be realized.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an optical character reading device according to an embodiment of the present invention, FIG. 2 is a flowchart for explaining the operation of the embodiment, and FIGS. 3 to 5 are respectively FIG. 7 is a diagram illustrating a display example of an image for explaining the operation of the embodiment. 10...Photoelectric conversion unit, 11...Image buffer, 12...Detection cutting unit, 13...Character recognition unit,
17, 18...Data buffer memory, 19...
Display control unit, 20...CRT display device,
21...Mouse.

Claims (1)

[Claims] 1. A scanning means that scans characters and figures recorded on a recording medium and converts them into image signals, an image memory that stores each image data of characters and figures obtained by this scanning means, and an image in this image memory. a detection cutting means for executing detection cutting processing of characters and figures from data and creating cutting information indicating each detected cutting position; a display means for superimposing and displaying characters and figures corresponding to the image data to be output; and confirmation for confirming or correcting each detected cutting position of the characters and figures in the image memory based on display information of the display means; What is claimed is: 1. An optical character reading device comprising a correction means and a character recognition means for recognizing a character detected and cut out at a detected cutout position determined by the checking and correction means.