JPH0411914B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a character recognition device, and particularly to a character correction method for correcting misread characters.

[Conventional technology and problems]

FIG. 1 is a diagram showing a comparison example of input characters and their recognition results. a indicates an input character, and b indicates a recognition result. As shown in Figure 1, when input character a is "Japanese" and recognition result b in the character recognition device is "Japanese", the recognition result of the second character is The word "tree" must be changed to "book." Conventionally, the character correction methods in such cases have been to display candidate characters for recognition of the input character ``book'' and allow the operator to select from among them; There are methods such as inputting a character code.
However, in the former method, the number of recognition candidate characters increases because each character is corrected, and the burden on the operator is heavy as the operator selects from among them and corrects the misread characters.
Again, the operator must operate the keyboard for each character to correct misread characters, which increases the burden on the operator. As described above, the conventional methods have the disadvantage that the operations are complicated and the burden on the operator becomes heavy.

[Purpose of the invention]

The present invention is intended to eliminate the above-mentioned drawbacks, and aims to facilitate the operation of correcting misread characters when the input character and the recognition result are different.

[Structure of the invention]

In order to achieve the above object, the character correction method of the present invention includes an observation unit that reads input characters from a medium on which the characters are written and photoelectrically converts them, and extracts characteristics of the input characters based on image information from the observation unit. a feature extraction section; a recognition dictionary section storing standard features of a plurality of character types; and a comparison between the features of the input character extracted by the feature extraction section and the standard features stored in the recognition dictionary section. an identification unit that selects a plurality of candidate characters for one input character using the identification unit; and a classification unit that sorts the plurality of candidate characters for one input character selected by the identification unit according to the rank of similarity. a word dictionary section storing a plurality of words, a display section, a character specifying means for specifying a misread word in the character string displayed on the display section, and a correction control. The correction control unit displays the candidate character string stored in the first rank of the candidate category buffer on the display unit, and corrects the misread word on the display unit by the character designation means. If specified, the candidate character of each rank corresponding to the misread word in the candidate category buffer is compared with the contents of the word dictionary section, the candidate word is reselected, and the reselected candidate word is The present invention is characterized in that it is configured to be displayed on the display section.

[Embodiments of the invention]

Hereinafter, the present invention will be explained with reference to the drawings.

FIG. 2 is a block diagram showing one embodiment of the present invention;
3 is a diagram showing an example of displaying input character recognition results, FIG. 4 is a diagram illustrating a character correction method using a word replacement display, and FIG. 5 is a diagram illustrating a character correction method using a word column display. FIG. 6 is a diagram illustrating an example of a word matching method. In FIG. 1, 1 is an observation unit, 2 is a feature extraction unit, 3 is an identification unit,
4 is a recognition dictionary, 5 is a candidate category/batshua, and 6 is a recognition dictionary.
7 indicates a correction control section, 7 indicates a word dictionary, and 8 indicates a display section.

In FIG. 2, an observation section 1 reads input characters from paper on which characters are written and photoelectrically converts them, and the output thereof is supplied to the input of a feature extraction section 2. The feature of the input character is extracted from the character/graphic information supplied from the section 1, and its output is supplied to the input of the identification section 3.
The identification unit 3 identifies candidate categories that are determined to be close to the input character by comparing the features of the input character supplied from the feature extraction unit 2 with standard character features stored in the recognition dictionary 4 in advance. The standard features are input from the recognition dictionary 4 and the output is fed to the input of the candidate category buffer 5.
The identification unit 3 selects a plurality of candidate characters for one input character. The candidate category buffer 5 stores the candidate categories supplied from the identification section 3, and its output is supplied to the input of the modification control section 6. The candidate category buffer 5 has an area of n stages, and the candidate character of the first rank is stored in the area of the first stage, and similarly the candidate character of the nth rank is stored in the area of the nth stage. . The size of the area in each row is such that candidate characters for one screen can be stored, for example. The modification control unit 6 controls the candidate category buffer 5.
The candidate character string stored in the first rank is displayed on the display unit 8. In addition, when a misread word is specified from the display section 8, a candidate word is selected by comparing the words corresponding to each rank of the candidate category buffer 5 with the contents of the word dictionary 7, and the candidate word is displayed on the display section 8.
to be displayed.

Next, a character correction method will be explained with reference to FIGS. 2 to 5. When the image information of the input character is read by the observation section 1, the image information is transferred to the feature extraction section 2. The feature extraction unit extracts the features of the input character from the image information, and the features are extracted by the identification unit 3.
will be forwarded to. The identification section 3 compares the features of the input character extracted by the feature extraction section 2 with standard features stored in the recognition dictionary 4, and selects a plurality of candidate characters for one input character. The plurality of candidate characters selected by the identification unit 3 are classified according to the rank of similarity, and the candidate character in the first rank is placed in the first stage area of the candidate category buffer 5, and similarly The ranked candidate characters are stored in the n-th stage area of the candidate category buffer 5. When each candidate character is stored in the candidate category buffer 5, the candidate character string stored in the first stage of the candidate category buffer 5 is displayed on the display unit 8 by the correction control unit 6. When a misread word is specified from the display unit 8, the correction control unit 6 selects a candidate word by comparing the words corresponding to each rank of the candidate category buffer 5 with the contents of the word dictionary 7, and selects the candidate word. The word is displayed on the display section 8.

FIG. 3 shows an example in which the recognition results of input characters are displayed on the display unit 8. In Figure 3, character strings that include misread characters include ``Bakusakuki'', ``Enji'', and ``Hibokujin''. Therefore, when the operator looks at the display unit 8 and indicates the misread word by using a light pen or a cursor, the correction control unit 6 Three-letter words are then sequentially read out from the word dictionary 7 and matched with the candidate categories in the candidate category buffer 5. After the words have been matched, the display section 8 is controlled so that several candidate words determined by the correction control section 6 to be close to the input character string are displayed. The display method is to replace the part where the original word is displayed with the maximum candidate word as shown in the oblique section in Figure 4, or to replace the part where the original word is displayed with the maximum candidate word as shown in the lower part of the display screen in Figure 5. There are methods such as displaying candidate words all at once in a separate column. In the latter method, misread characters are corrected by the operator specifying which word from among the displayed words. Once corrected, the corresponding first-ranked character in candidate category buffer 5 is replaced with the corrected character.

An example of a word matching method will be explained with reference to FIG. FIG. 6a shows that when the input character example "calculator" is input, the features extracted by the feature extraction unit 2 and the standard features stored in the recognition dictionary 4 are compared in the identification unit 3, and the input character is This is an example of candidate categories determined to be close. For the input character ``ke'', the input character ``calc'' is assigned in the order of ``ku'', ``ke'', and ``hu''.
For the input character ``Ki'', candidate categories are determined in the order of ``san'', ``shin'', and ``honour'', and for the input character ``ki'', the candidate categories are determined in the order of ``ki'', ``ori'', and ``machine'', and Weights 1, 2, and 3 are added according to the ranking. This first-ranked candidate category is selected for displaying the recognition results. FIG. 6b shows three words extracted by the correction control unit 6 from among the words registered in the word dictionary 7 for "Bakusakuki" which was designated as a misread character.
This is an example of a character word. FIG. 6c is a diagram showing an example of distance calculation between a three-letter word extracted from a word dictionary and a candidate category. When calculating the distance, for example, as shown in Figure 6a, the first-ranked candidate category is assigned the value "1," the second-ranked candidate category is assigned the number "2," and the third-ranked candidate category is assigned the number "2." Items that are not in the numerical value "3" candidate category are assigned a numerical value of "4" weight. This weight is the 6th
When adding up each of the three-letter words extracted from the word dictionary 7 in Figure 6, "calculator" is the smallest at 4, followed by "debate" at 9, and "counter" at 10, as shown in Figure 6c. Become. As a result, the word most similar to the input character string is determined to be "calculator." Candidate words are displayed in FIG. 4 or 5 in descending order of the calculation result, that is, in order of the ranking determined to be more similar.

In the above explanation, the candidate categories are accumulated in the candidate category buffer 5 up to the third rank, and a weight of numerical value "1" to "3" is added according to the rank, and characters other than the candidate category are numerically Although the explanation has been given by adding a weight of "4", there is no problem in changing these numerical values as appropriate.

〔Effect of the invention〕

As is clear from the above description, according to the present invention, misread characters and unrecognized characters can be corrected in character string (word) units, so the number of candidate characters for correction is reduced, and character correction operations This has the effect of making it easier.

[Brief explanation of drawings]

Fig. 1 is a diagram showing a comparison example of input characters and recognition results, Fig. 2 is a block diagram showing one embodiment of the present invention, and Fig. 3 is a diagram showing an example of displaying recognition results of input characters. 5 is a diagram illustrating a character correction method using a word replacement display, FIG. 5 is a diagram illustrating a character correction method using a word column display, and FIG. 6 is a diagram illustrating an example of a word matching method. DESCRIPTION OF SYMBOLS 1... Observation unit, 2... Feature extraction unit, 3... Identification unit, 4... Recognition dictionary, 5... Candidate category buffer, 6... Correction control unit, 7... Word dictionary, 8...
...Display section.

Claims (1)

[Scope of Claims] 1. A plurality of observation units that read input characters from a medium on which the characters are written and photoelectrically convert them, and a feature extraction unit that extracts features of input characters based on image information from the observation unit. The recognition dictionary section stores the standard features of the character types, and the features of the input character extracted by the feature extraction section are compared with the standard features stored in the recognition dictionary section to extract one input character. On the other hand, there is an identification section that selects a plurality of candidate characters, and a candidate category that sorts and stores the plurality of candidate characters for one input character selected by the identification section according to the rank of similarity. a buffer, a word dictionary section storing a plurality of words, a display section, a character specifying means for specifying a misread word in the character string displayed on the display section, and a correction control section; The control unit displays the candidate character string stored in the first rank of the candidate category buffer on the display unit, and when the misread word on the display unit is specified by the character designation means, Compare the candidate characters of each rank corresponding to the misread word in the candidate category buffer with the contents of the word dictionary section, reselect the candidate word, and display the reselected candidate word on the display section. A character correction method characterized by being configured as follows.