JPS61190677A - Shift controlling system for handwritten character recognition - Google Patents

Abstract

PURPOSE:To make shifting, displaying and correction of a specified character category by simple operation even when designation of character category is omitted by storing other character categories successively after designated character categories. CONSTITUTION:A result of recognition storing section 15 builds up the result of recognition to a format shown in the figure (a). In this format, a header of 1 byte that indicates the category of data, and data consisting of n1 Chinese characters, n2 Japanese syllabaries and n3 English symbols allotted 2 bytes for each shift character. Code of each character category is written in blank space of 2 bits after code '110000' of the header in the figure (b) at the time of inputting shift character of character category. The figure (c) shows the number of shift characters of each character category written after code '10' of the header of handwritten character input. By this constitution, corresponding shift character can be arrived at by shift even when designation of character category is not made.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention is a handwritten character recognition device that inputs handwritten characters by specifying the character type to be changed and performs real-time recognition,
The present invention relates to a modification of a shift control method for handwritten character recognition that shifts and displays a predetermined character type with a simple operation when the character type is not specified.

[Conventional technology]

Conventionally, in order to recognize handwritten characters, handwritten characters are written in a character frame 2 in a character input area 1 using a tablet 10 illustrated in FIG. 4 using an electronic pen. Further, each key of the function input section 3 is provided to control the handwriting input. Among these, kanji,
Kana and English symbols 4 are character types, and by specifying the next character type immediately before the change, it is useful to increase the mm ratio.

Hereinafter, an execution key is provided along with keys for modification such as cancellation, insertion, and deletion.

In this case, the recognition unit performs real-time recognition by comparing the character with the candidate character based on the order of strokes of the electronic pen (9 strokes, 1 position).

That is, the feature of this recognition is that it is performed in real time by specifying the character type and recognizing strokes.

[Problem that the invention seeks to solve]

In the conventional recognition described above, it often happens that the user forgets to specify a change in character type when inputting handwritten characters.

In this case, since the candidate characters stored in the storage section cannot be limited in advance by character type, correct recognition results cannot be obtained. Therefore, it was necessary to move the cursor back, specify a change to the character type, and enter the characters again. This does not place a heavy burden on the operator, and when recording the handwritten results in a journal or the like, there is a problem in that the corrections are overwritten, making it difficult to read.

An object of the present invention is to provide a handwriting system that, when inputting handwritten characters by specifying a character type to be changed, can shift and display the predetermined character type with a simple operation even if the character type is not specified. The object of the present invention is to provide a shift control method for character recognition.

[Means for Solving the Problems] In order to achieve the above object, the handwritten characters &! The shift control method of the system includes a shift state specifying means for specifying the character type to be changed, an input means for inputting handwritten characters, a recognition unit for recognizing input characters based on the character type and handwritten input from the means, and a recognition unit for recognizing input characters based on the character type and handwritten input from the means. A handwritten character recognition device comprising a storage unit that stores candidate characters in each shift state, and a display unit that sequentially displays recognized characters, the storage unit storing a specified character type as a recognition candidate character in each shift state. This system is characterized in that other character types are also stored, and the recognition candidate characters are displayed after shifting to a predetermined character type based on the designation of the immediately preceding character type.

[For production]

Conventionally, shifting was performed only for the specified character type, but with the present invention, even if you forget to specify the character type, you can move on to other character types by proceeding with the shift, so the original character can be recognized and displayed. Then a tortoise is formed. Furthermore, even if a character type (incorrect) is discovered after inputting multiple characters, the user only has to return the cursor and correct the characters within the range of the incorrect character type, allowing quick recognition with simple operations.

〔Example〕

FIG. 1 is an explanatory diagram of the configuration of an embodiment of the present invention.

In the figure, the tablet 10 specifies the character type to be changed, writes handwritten characters in the character input area 1, and sends the characters to the input data storage unit 12 via the input control unit 11, as described in FIG. The designation of the character type and the data of the stroke order, number of strokes, and position are stored in . The character type is specified by the shift control section 16. It is set in advance in the recognition result storage section 15 via. On the other hand, in the present invention, each character is stored in the recognition dictionary section 14 so as to include a shift character of the foreign character type in addition to a shift character of the original character type. For example, for “ko”,
In addition to the original character type ``ko'' in hiragana shift, there are ``ni'' in kanji shift, ``17'' in katakana shift, and symbols. A shift "=" is stored, and the shift character of the character type from the input data storage unit 13 is recognized by comparing it with the candidate character in the recognition dictionary 14. This recognition result is stored in the recognition result storage unit 15Y? -Remember;
It is displayed on the display section 18 via the display control section 17. The g-texture result storage unit 15 assembles the recognition results into the format shown in FIG. 5(α). This format includes a 1-byte header indicating the type of data at the beginning, and 2 bytes for each shift character.
Data consisting of byte allocation, one kanji character, two kana characters, and an English symbol cho is provided. The same figure (6) shows the header code ("110000") when inputting the shift character type.
The code for each character type is written in the 2-bit blank space after the .

(c) t1 handwritten character input header code (“1
o”) followed by the number of characters shifted for each character type*n2*n
8 is shown.

With this configuration, it is possible to reach the corresponding shift character (by shift), not only when a character type is specified, but also even when a character type is not specified.

FIG. 2(G) is an explanatory diagram of the operation based on a data example of the present invention.

The figure shows an arrangement of candidate characters for each character type, kanji, kana, and alphabetic symbol shift, when this data is recognized. The lower row of candidate characters shows code examples.

The operating procedure will be explained below based on this data example.

(1) Data processing procedure for normal input The correct input operation for the data example is as follows.

ko indicates input of changed character type.

r Loro Mountain S e, t Teniro Tabikuni t J Also, the stored data in this case is as follows.

t mouth O mountain 85 3B33 2455ko
9E 4673 3B52 2455 254B
2573 2161e 95 4757 2
428 2349 95 3235 244
6 25570 mouth O trip 81 4E39 eye blade 84 2472 The procedure will be explained below with reference to FIG. 1 for the data example of FIG. 2.

(2) When the character type 4 kanji shift t Bentouch is touched on the tablet, the input control unit 11 determines that it is a shift instruction and sends the information to the shift control unit 16. The shift control unit 16 stores the kanji shift information “C” in the MW & result storage unit 15.
Send O' and change the internal shift to Kanji.

■ When the Tapuretsu MOK handwritten character "mountain 1" is input, the stroke data is stored in the input data storage section 12.The recognition section 15 uses this stroke information and the recognition data stored in the storage dictionary section 14 to , performs recognition corresponding to character type 4: Kanji shift, Kana shift, and English symbol shift.

First, recognition is performed with a kanji shift, and 3B33" is recognized as ``1 mountain" and sent to the recognition result storage section 15. Next, recognition is performed with a kana shift, and 1 fu is recognized as ``2455''.
is sent to the recognition result storage section 15. When all recognition is completed, a recognition completion message is sent to the recognition result storage section 15.

■The recognition result storage unit 15 assembles the data from Figure 3 (ω) based on the data from the recognition unit 13, stores 185'' since there is one candidate character each for Kanji and Hirakana, and then recognizes Data “3833” and 2455” are stored.

At the same time, it is determined that the current shift state is a kanji character, and the code 13B33'' corresponding to the kanji character is selected from the recognition data and sent to the display control section 17.

(2) The display control section 17 displays a mountain on the rule display section 18 based on this data.

■Similarly, store the data shown in the example. The display will be as follows.

(11) Next, the character type shift Kugata is missing (forgotten)
In this case, consider the following input operation as an example.

The stored data in this case is 7 of the stored data in a normal case.
The shift information @01'' of the byte is omitted.

The display at the end of input will be as follows.

■==匡=] In this case, ``1 over'' is erroneously displayed as 2 otsu 1 for 卺 according to the recognition result of the kanji shift.

Previously, the cursor was moved back to the 20 position and then pressed again.

I had to re-enter "over the top."

The processing in the case of the present invention will be explained below.

■Move the cursor back to the “2” position.

■Touch the kana shift section on the tablet.

The kana shift information is provided by the input control section 11. Shift control section 1
6 to the recognition result storage section 15.

■When the recognition result storage unit 15 receives the kana shift information,
The code 2455'' corresponding to the kana shift is selected from the recognition data for the character ``2'' at the current cursor position, and sent to the display control unit to advance the cursor by one character.

(2) In the same way, change the display until shift information appears in the stored data.

〔Effect of the invention〕

As explained above, conventional shifts were performed only for the specified character type, but with the present invention, even if you forget to specify the character type, the shift shifts from the current character type to another character type, so you can quickly shift the original character. can be reached, recognized, and displayed. Even if you find an incorrect character type after inputting multiple characters, you can return it to the missing character type as before and simply shift the corresponding character type without having to correct everything, making the operation easier. quick in
! ! knowledge becomes possible.

[Brief explanation of the drawing]

) Fig. 1 is a configuration explanatory diagram of an embodiment of the present invention, Fig. 2 is an explanatory diagram of the procedure of the present invention using data examples, Fig. 3 (α) to (c)
is an explanatory diagram of the data format used in the present invention, and FIG. 4 is an explanatory diagram of a handwritten character input tablet.
1 is a character input area, 2 is a character frame, 3 is a function input section, 4 is a character type shift input, 10 is a tablet, 11
12 is an input control unit, 12 is an input data storage unit, 13 is a recognition unit, 14 is a recognition dictionary unit, 15 is a recognition result storage unit, 16 is a shift control unit, 17 is a display control unit, and 18 is a display unit. .

Claims (1)

[Claims] a shift state specifying means for specifying a character type to be changed; an input means for inputting handwritten characters; a recognition unit for recognizing input characters based on the character type and handwritten input from the means;
A handwritten character recognition device comprising a storage unit that stores candidate characters in each shift state during recognition, and a display unit that sequentially displays the recognized characters, wherein the storage unit stores character types designated as recognition candidate characters in each shift state. This shift control method for handwritten character recognition is characterized in that other character types are also stored following the previous character type, and a recognition candidate character is displayed by shifting to a predetermined character type following the shift specified by the immediately preceding character type.