JPS61153787A - Information processing device - Google Patents

Abstract

PURPOSE:To simplify an operation and to enhance a recognition rate by dividing an information input frame into plural areas and performing the recognition of input information by examining the area of the divided areas in which a handwriting inputs is inputted. CONSTITUTION:A character is inputted by hand from an user. As the result of a recognition, whether plural candidates appear or not is discriminated. When the plural candidates appear, whether the character in question can be judged to a single candidate or not. If there if a possibility by the single candidate, a judgement is done which area of a character entry frame 4 on a tablet 1 an input character uses. KATAKANA (kind of Japanese syllabary) is written in 1/2 of lower section of a character entry frame 4 on the tablet 1 and HIRAGANA (Japanese syllabary) and KANJI (Chinese character) are written almost all areas of the character entry frame 4. By clearly defining in the manner, different character types can be easily discriminated.

Description

Detailed Description of the Invention [Technical Field] The present invention relates to a man-machine interface for facilitating information transmission between devices and humans, and particularly relates to an information processing device that can recognize handwritten input information, etc. It is something.

E. Prior Art] Handwritten character recognition is effective as an input means for computers and word processors, especially as an input means for Japanese. However, Japanese texts contain a wide variety of characters such as kanji, hiragana, katakana, alphabets, numbers, and symbols, and there are many characters that are difficult to distinguish based on their appearance.For example, the kanji ``power''
and ``power'' in katakana. Also, the hiragana “tsu”
Even though they have the same shape like `` and ''.

It may be necessary to generate different character codes depending on the size of the shape.

Conventionally, the method of solving this type of difficulty has been to provide a mode such as kanji mode, katakana mode, etc.However, as you can see from the shift lock type keyboard, this type of mode The switching designation is not only complicated for the user, but also has the disadvantage that many switching errors occur.

[Objective] The present invention aims to eliminate the above-mentioned drawbacks of the prior art, and another object is to provide a man-machine interface that reduces the burden of human input operations to machines; An object of the present invention is to provide an information processing device capable of recognizing input information and easily distinguishing between a plurality of types of information.

[Example] An example of the present invention will be described below with reference to one drawing.

FIG. 1 is a block diagram of a handwritten character recognition device according to an embodiment of the present invention.

In the figure, 1 is a tablet for inputting handwritten characters, 10 is a tablet) The coordinates of the tip of the input pen 2 (see Figure 2) and the up/down information of the pen 2 are input from the tablet L, and characters are being input. 11 is the coordinates of the pen tip obtained from the input control circuit IO.

From the up/down information of the pen 2 and the character input end signal 21 obtained from the single character recognition start detection circuit 13, the number of strokes of the input character, the coordinates of the representative point, the classification of each stroke, the intersection information of each stroke, and the length of each stroke , a feature extraction circuit 12 that extracts features of input characters such as the maximum coordinate point and the minimum coordinate point, and temporarily stores the extracted various data in the N product storage circuit 15;
is an input character extraction circuit which extracts input characters based on the time when the pen is up, the coordinates of the pen tip, etc., and sends a single character input end signal 20 to the character recognition start detection circuit 13;
Reference numeral 13 outputs a matching start signal 22 to the matching circuit 16 based on the single character input end signal 20 sent from the input character extraction circuit 12 and the character recognition start designation signal 23 sent from the input control circuit 10, and starts the matching circuit 16. This is a character recognition start detection circuit. 14 is a matching circuit 1 that detects the position of a character to be recognized based on the position of the cursor displayed by the display circuit 18 or a recognition target character position designation signal sent from the input control circuit 10;
6. A recognition target character position detection circuit that tells which character is to be recognized among the features of several input characters stored in the time accumulation memory circuit 15; 15 is extracted by the feature extraction circuit 11; 16 is a character recognition start detection circuit 1;
3 receives the character recognition start signal 22 and is activated, the recognition target character position detection circuit 14 detects the recognition target character, reads out the characteristics of the corresponding input character from the temporary storage memory circuit 15, and reads the input character. and a dictionary 17, and outputs recognized characters to the display circuit and memory; 17 is a dictionary in which characteristic standard patterns of known characters are registered; 18 is a matching circuit 1;
19 a display circuit for displaying recognized characters recognized by 6;
is a memory that stores recognized characters recognized by the matching circuit 16.

Details of one embodiment of the tablet 1 are shown in FIG.

In FIG. 2, 2 is an input pen, 3 is a menu area used for recognition start designation, mode designation when there is mode switching, other commands, etc., and 4 is a character entry frame in which handwritten characters are written. Figure @3 is an enlarged view of the character entry frame 4. The dashed line in Figure 0 is a guideline indicating the region boundary when the character entry frame is divided into regions.

Hereinafter, the recognition process of the input handwritten characters of the device of this embodiment will be explained as follows.
This will be explained with reference to the flowchart shown in the figure.

First, in step 340, the user inputs characters by hand. When the input for one character is completed, the process advances to step S41, and in step S41, the feature extraction circuit 11
The characteristics of the input characters extracted by the above are stored in the temporary storage circuit 15.

As described above, this temporary storage circuit 15 contains information on the maximum coordinate point and the minimum coordinate point. This information will be used later when discrimination based on area information is required.

Next, in step 342, the matching circuit 16 matches the strokes of the input characters stored in the temporary storage circuit 15 with the strokes stored in the dictionary 17.

It is not necessary to limit the matching to strokes.

In the following step S43, as a result of matching, it is determined whether or not the input character strokes match the character strokes in the dictionary 17 and the input character has been recognized. If it is unrecognizable, step S
The process proceeds to step 44, where "?°'" is displayed on the display circuit 18, and the process returns to step 340, where characters and characters are entered again.

If the input character can be recognized in step S43, the process advances to step 545, and it is determined whether or not multiple candidates have emerged as a result of recognition. If a single candidate is selected from a plurality of candidates, the process advances to step 350 and joins the process described later. If multiple candidates appear, the dictionary is further checked to see if the character can be identified as a single candidate based on the area information. As a result, step 5
In step 46, it is determined whether there is a possibility of becoming a single candidate by using discrimination based on area information. If there is a possibility that there is a single candidate, the process advances to step S49, and if the input character is a tablet)
Determine which area of the character entry frame 4 above is being used.0For example, if all four areas are used as shown in Figure 3(a), the screen in Figure 3(b) 4 as shown
If you are using the divided lower half area, the third ffl
d This can be classified as 8, such as when only the lower left part of the area divided into four is used, as shown in (d).In other words, here, for example, the maximum y-coordinate is 1/1 of the height of the character frame. If we use the kanji character ``力 J'' when the value is north of 2, and the katakana character ``力'' when the value is less than l/2, we can get the following from Figure 3 (
a) is the kanji for “power,” and Figure 3 (b) is the katakana word for “power.”
It can be determined that the input is Similarly, by defining the case where the input character fits into the lower left area as a lowercase letter "tsu" and the other characters as a normal "tsu", Figure 3(c) shows the normal "tsu" and the lower left letter. Figure 3 (d) shows the lowercase letter “
It can be determined that In this way, one of the plural candidates is selected and the process proceeds to step S50. In this embodiment, the katakana character type is the l/ 2 (less than 1/2 of the height of the character frame), and the kanji type and hiragana character type are filled in almost the entire area of the character entry frame 4 as shown in fJS3 figure (a) or fJS3 figure (C), In the case of a lowercase character in the hiragana or katakana character type, it is defined to be written in the lower left corner of the character entry frame 4 as shown in FIG. 3(d).

By defining in this way, different character types can be easily distinguished.

In step 346, if there is no possibility that the candidates can be narrowed down to one even if the region discrimination information is taken into consideration, the process proceeds to step 547, and all of the plurality of recognition candidates are displayed on the display circuit 18. . Then, in the following step 348, the inputter (user) specifies and inputs one of the display candidates in the menu area 3 etc., as is often done in the past.
Select recognition characters.

In step 350, which is a confluence from step S45, step S48, and step S49, the character of the recognition result is displayed on the display circuit 18, and further, in step 351, the result is stored in the memory 19, and handwritten input character recognition processing for one character is performed. will end.

If a subsequent handwritten character input is required, the process returns to step 340 and recognition processing for the next handwritten character is executed.

In the following explanation, to distinguish between multiple character types and similar characters, the character entry frame is divided into regions without using mode specification, and input characters are discriminated based on the input conditions in these divided regions. By inputting only the upper left area of the area, for example, it is possible to define this as an input of a control command.

By determining (interpreting) in this manner, all kinds of information can be input using the character entry portion (character input portion) of the tablet 1.

[Effects] As explained above, according to the present invention, the information input frame in which handwritten input is to be performed is divided into a plurality of regions, and one handwritten input is input depending on which region of each divided region is input. It is possible to provide an information processing device that can recognize information, is easy to operate for the user, and has a high recognition rate.

Furthermore, by performing handwritten character recognition according to the present invention, character types can be easily identified, and a more efficient information processing apparatus can be provided. Furthermore, it has become possible to provide a man-machine interface that can reduce human input operations to the machine.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a handwritten character recognition device according to an embodiment of the present invention, and FIG. 2 shows details of a tablet for inputting handwritten characters of this embodiment. 3(a) to 3(d) are enlarged views of the character entry frame and diagrams showing an example of inputting handwritten characters into the character entry frame, and FIG. 4 is a flowchart of the character recognition process of this embodiment. In the figure, l...tablet, 2...input pen, 3...
・Menu area, 4...Character entry frame, lO...Input control circuit, 11...Feature extraction circuit, 12...Input character cutting circuit, 13...Character recognition start detection circuit, 1
4... Recognition target character position detection circuit, 15... Temporary storage storage circuit, 16... Matching circuit, 17... Dictionary, 18... Display circuit, 19... Memory.

Claims (2)

[Claims] (1) An input area having a plurality of information input frames, a reading means for reading information input into the input area, and a determining means for determining the input information read by the reading means, the determining means If there is a single discrimination candidate, the multiple information input frames mentioned above are ignored, and if there are multiple information input frames, it is further determined which area among the divided areas above the input information was input. An information processing device characterized by: (2) The handwritten information is handwritten character information, and the character type of the handwritten character information is recognized based on the input situation to the divided area of the information input frame.
The information processing device described in the section.