JPS60160482A - Recognizing device of character - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a character recognition device, and particularly relates to a recognition result display method suitable for correcting the occurrence of unreadable or misread characters.

[Background of the invention]

Conventional recognition result display methods have been broadly classified into two types, with the aim of making character correction easier.

FIG. 1 is a diagram showing the display screen when character recognition results are displayed at a brightness different from that of unreadable characters (Patent Publication 52-
No. 19415). In FIG. 1, 10 represents a display surface. The kanji and hiragana shown in 10 are characters that can be read correctly, and the question mark 11 represents unreadable characters. The unreadable characters are corrected character by character by the operator, but they have the following drawbacks.

In other words, since unreadable characters are displayed using code information (a question mark in this example), the operator can recognize their position, but cannot recognize what kind of character they are based on only the information on one screen. Therefore, in order to correct unreadable characters, the operator must read the original document.
A decrease in correction speed and the occurrence of incorrectly corrected characters are unavoidable.

FIG. 2 is a diagram showing the display screen of the character recognition device when an unreadable character button is transferred along with its position information (
Patent Publication No. 56-36463). 20 in Figure 2
represents the display diagram, 23 represents the recognition result character string, and 2
1 represents the transcribed unreadable character button, and 22 represents the position information of the unreadable character corresponding to 23. Although this method is recognized to have improved correction efficiency compared to the above-mentioned method, it has the following drawbacks. In other words, since the recognition result character string and the unreadable character ``bat'' exist separately, the operator is required to watch both character groups alternately.

However, since the positions of unreadable characters in the recognition result character string are random, a decrease in the correction speed and the occurrence of erroneously corrected characters are unavoidable.

[Purpose of the invention]

SUMMARY OF THE INVENTION An object of the present invention is to provide a recognition result display method that can quickly correct the occurrence of unreadable or misread characters in a character recognition device.

[Summary of the invention]

In order to achieve the above-mentioned object, the present invention subjects a character pattern to normalization processing, changes color or brightness, and embeds and displays the character pattern in the vicinity of the corresponding recognition result character. FIG. 3 is a diagram showing a display screen when displayed on a terminal using the present invention. In FIG. 3, 30 represents a display surface. Kanji characters 32 surrounded by squares represent character slam images, and other character strings 31 represent recognition result character strings. That is, by identifying the recognition result character string and character pattern by color or brightness, unreadable characters and misread characters can be corrected quickly and accurately.

[Embodiments of the invention]

An embodiment of the present invention will be described with reference to the drawings. FIG. 4 is a block diagram showing an embodiment of a character recognition device according to the present invention. In the figure, a photoelectric conversion unit 42 optically scans characters on a form 41 and converts them into electrical signals. Thereafter, the character pattern converted into an electrical signal is output to the recognition section 43 and the image processing section 44.

The recognition unit 43 performs necessary preprocessing such as nozzle removal on the character pattern, extracts features necessary for recognition, and calculates the degree of similarity or distance between the input slam feature and the standard slam feature. Authenticity or unreadability is determined based on the calculated similarity or distance. Thereafter, the result code and information are output to the nine character generator 46 via the selector 45. Note that the selector 45 selects the recognition unit 43 in the recognition mode, and selects the keyboard 51 in the character correction mode.

Character generator 46 generates a font image corresponding to the result code and outputs it to synthesis section 48 .

On the other hand, the image processing section 44 receives the position information of the unreadable characters from the control section 52, and as shown in FIG. Performs size normalization processing on character patterns. The normalized positional information of the unreadable characters is output to the color information adder 47.

The color information adder 47 adds 4
A color code different from 6 is added and output to the composition section 48. In this case, it is obvious that it is better to add another color code to the unreadable characters.

It is clear that the means for distinguishing from other character parts is not limited to color information, but may generally be brightness information or other means, and in this case, 47 is brightness information.

The composition section 48 sends the image information from the color information adder 47 and the character generation section 46 to the video memory 49 based on the address signal generated by the control section 52 .

The recognition result character string in which the corresponding original character pattern is placed nearby is converted into a video signal in the video memory 49, and the C
The voltage is applied to the RT display unit 50.

The operator pays attention to the screen displayed on the CRT display section 50 and uses the keyboard 51 to correct the unreadable characters and the misread characters. That is, first, the character of the illegible character pattern is recognized. After that, the keyboard 51
Use to insert characters that are illegible. Repeat this process until there are no unreadable characters. Next, the original character BATA and the corresponding recognition result character are compared to search for misread characters.

If a misread character is found, correct it using the same method as for misreading.

In the above correction means, the keyboard 51 preferably has a Japanese language editing function, such as a Japanese word processor, especially when the characters include kanji or the like.

Control signals necessary for the above operations are supplied from the control section 52 through signal lines (not shown). Note that a normal microprocessor or the like may be used as the control unit 52.

As described above, according to the present invention, the original character pattern can be displayed in the vicinity of the corresponding recognition result character by changing the color or brightness, so that the misread characters can be found and corrected without overlooking the unreadable characters. Can be done quickly. Therefore, the speed and reliability of character input work can be improved, and the effect is significant.

〔Effect of the invention〕

According to the present invention, misreading and unreadable characters can be quickly corrected, and the turnaround time of the device can be shortened.
This has the effect of reducing CR operating costs. Furthermore, since the apparent recognition performance of the device can be improved, there is an effect of improving the reliability of the device.

[Brief explanation of drawings]

1 and 2 are explanatory diagrams of a display terminal in a conventional character recognition device, FIG. 3 is an explanatory diagram of a display terminal according to an embodiment of the present invention, and FIG. 4 is a block diagram of an embodiment of the present invention. . 41... Form, 42... Photoelectric conversion unit, 43... Recognition unit, 44... Image processing unit, 45... Selector, 4
6...Character generator, 47...Color information adder, 48.
...Composition unit, 49...Video memory, 50...CR
T Display, 1st Village, No.

Claims (1)

[Scope of Claims] 1. In a character recognition device comprising a character pattern observation unit, a recognition unit, a result display unit, and a character correction unit for correcting erroneous and illegible characters, normalization of a slam image is performed. A character recognition device comprising an image processing means, a composition means for embedding the normalized image in the vicinity of a corresponding recognized character string, and a means for displaying the image obtained by the composition means. 2. The character recognition device according to claim 1, further comprising means for adding color or brightness information to the bang image. 3. The character recognition device according to item 2, further comprising means for adding color or brightness information to the unreadable baton image that is different from that of other baton images.