JPH07182461A - Handwritten character recognition device - Google Patents

Abstract

PURPOSE:To learn the graphological data which are inputted in a handwriting input mode by registering the correspondence between the character selected out of the displayed candidate characters and the graphological data on the selected character when the learning of the graphological data is instructed. CONSTITUTION:A character recognition part 12 collates the graphological data inputted from a character input part 11 with the dictionary data stored in a dictionary part 18 for recognition. Thus the part 12 recognizes the inputted characters and successively extracts the candidates in response to the result of recognition and in the order of higher possibility out of those characters included in the part 18 in consideration of the coincidence and the using frequency of those characters. A candidate character display part 13 displays the extracted candidate characters. When the learning is instructed by an input operator to the input graphological data, a learning deciding part 15 informs a character registering part 17 of a fact that the graphological data should by learnt. The part 17 secures the direct correspondence between the graphological data and the character desired by the input operator and selected by a character selecting part 14 to generate the new data and registers the data in the part 18.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a handwritten character input device, and more particularly to a handwritten character recognition device having a learning function.

[0002]

2. Description of the Related Art In the conventional handwritten character recognition, when learning a handwritten character, a method of designating a character to be learned and then inputting handwritten character data for the character is adopted. . As a method of specifying this character, enter the character code from the keyboard,
There is a method of inputting correct characters by handwriting and designating characters to be registered (“JP-A-1-103787”).

A learning method and device in this conventional online handwritten character recognition device will be described below with reference to the drawings. FIG. 8 is a block diagram of an example of a conventional handwritten character recognition device. In the figure, 81 is a character input unit, 82 is a character recognition unit, 83 is a recognition dictionary unit, 84 is a recognition result output unit, 85 is a learning character designating unit, and 86 is a learning character. It is a character data input part,
87 is a character registration unit.

In FIG. 9, a character is designated for learning,
FIG. 9 is an explanatory diagram showing a state of being registered in the dictionary unit in combination with handwriting data corresponding to the character. In the figure, 91 is a character input for designating a learning character, 92 is a character code converted from the input character, and 93 is handwriting data of a character to be learned.
First, handwritten character recognition in a conventional online handwritten character recognition device will be described.

The character input section 81 allows the input person to input characters by handwriting. For input, a configuration of an electronic pen (stylus pen, hereinafter also referred to as “input pen” or simply “pen”) and a tablet (electronic input handwriting detection panel, digitizer) is generally used. This is done by shifting the Z (depth) direction inside the flat electrical insulator and then X
An alternating current is passed through metal wires that are densely provided at equal intervals in the Y direction and the Y direction, while the electronic pen is magnetized to determine whether the electronic pen is in contact with a flat plate, the contact coordinate position, and the like. It is like detecting as an AC disturbance in a thin line. However, since these are well-known techniques described in, for example, Shigeru Watanabe et al., Edited by Ohmsha, "Microcomputer Handbook (1985) II, Chapter 8, 3 Pen touch input", etc. principle,
Detailed description of the configuration and the like is omitted.

The handwritten character data input by the character input unit 81 is in the form of handwriting data such as a sequence of coordinate points and the character recognition unit 8
Passed to 2. The character recognition unit 82 refers to the received handwriting data with the dictionary unit 83 to obtain a candidate character, and further obtains a character code for the candidate character as the recognition result. The obtained character code is passed to the recognition result output unit 84, and finally the character corresponding to the code is presented to the input person. Regarding the technology for obtaining character codes from handwriting data, basically, electronic collation of input strokes and registered strokes, and dictionary collation of confirmed input strokes and letters are the basics. Becomes However, regarding these, "Tomimoto et al .: Online handwritten kanji and hiragana recognition, IEICE technical report,"
PRL 82-37, p. 49 (1982) ”and other various documents, and therefore detailed description of its principle, configuration, etc. is omitted. Furthermore, the technique of selecting or confirming the character intended for input by the inputter from the candidate characters is basically the same technique as that of a keyboard-type Japanese word processor. For this reason,
With respect to this, too, detailed description of the principle and configuration will be omitted.

Next, learning of handwritten characters in the conventional online handwritten character recognition device will be described. The learning character designating section 85 allows the input person to designate a character code of a character to be learned. As a method of designating a character code, there are a method of inputting a character with a code from a keyboard, a method of carefully inputting a character by handwriting in a regular print, and obtaining a character code from the recognition result. For example, when the character “mountain” is designated as the learning character as shown at 91 in FIG. 9, the character code 3B33 may be directly input from the keyboard, or the learning character 3B33 shown in 92 as a character code (here, JI of "mountain")
S code) may be obtained. Next, the learning character data input unit 86 causes the input person to input the handwritten character data corresponding to the designated character code using the pen and the tablet, and extracts the handwriting data to be learned from the data. And
The handwriting data obtained as a result is passed to the character registration unit 87.

The character registration unit 87, which has received the input learning character code and handwriting data, generates new character recognition data based on these data, and the recognition dictionary unit 8
Register it in 3. For example, the collapsing character of “mountain” as shown by 93 in FIG. 9 is not recognized as it is as “mountain”. Therefore, it is newly generated as handwriting data corresponding to the code 3B33 of “mountain” and recognized. Dictionary section 8
Let 3 register for learning.

[0009]

However, in the above-mentioned conventional handwritten character recognition device, learning registration of characters must be performed by a procedure different from normal character input, which is not only complicated but also complicated. In some cases, there may be a difference between the data newly input for learning and the data actually input in the state of handwriting in a hurry, and the learning effect may be impaired. Moreover, the method of designating the character code to be learned is also troublesome.

In view of the above problems, the present invention makes it possible to learn handwriting data input during the input of normal handwritten characters as it is, and it is unnecessary to input a new character code, so that handwritten characters can be easily learned. The purpose is to provide a recognition device.

[0011]

In order to achieve the above object, according to the invention of claim 1, a dictionary unit in which characters and handwriting data are registered in association with each other, and input handwriting data is used as a basis. The character recognition unit that performs character recognition, the candidate character display unit that displays the candidate characters obtained by the character recognition unit in a predetermined procedure, and the intention to input to the input person from the candidate characters displayed on the candidate character display unit. In a handwriting character recognition device having a character selection unit that makes it possible to select a selected character and a learning unit that exerts a learning function by registering input handwriting data in the dictionary unit in association with characters specified by the input person , When the input instruction is given to the input handwriting data by the input person, the input portion is intended to be input by the input person from the candidate characters displayed corresponding to the input handwriting data to the learning unit. Shi The dictionary unit as it associates with the selected character as the character is characterized by having a learning instruction unit that newly registered.

According to a second aspect of the present invention, the learning instruction section is provided at a position close to the candidate character displayed on the candidate character display section. According to the invention of claim 3, a pen instruction input unit is provided which enables the character selection unit and the learning instruction unit to select a character intended to be input with a pen for handwriting input and to instruct learning. It is characterized by being.

According to another aspect of the present invention, the pen instruction input unit recognizes a predetermined locus created by a pen for handwriting input with respect to the displayed candidate characters, thereby instructing selection of an intended object. It is also characterized by being able to give instructions for learning. According to the invention of claim 5, the character recognition unit has redundancy in recognition of candidate characters.

[0014]

With the above structure, in the invention of claim 1,
When the input instruction is given to the handwriting data by the learning instruction unit by the input person, the input portion is selected by the inputter from among the candidate characters displayed corresponding to the handwriting data input to the learning unit. The character selected as the character intended for input is directly associated with the character and is newly registered in the dictionary unit.

In the invention of claim 2, the learning instruction section is provided at a position close (including adjacent) to the candidate character displayed on the candidate character display section. In the invention of claim 3, the pen instruction input unit enables the character selection unit and the learning instruction unit to select and learn a character intended to be input with a pen for handwriting input.

In the invention of claim 4, the pen instruction input section recognizes a predetermined locus of the input pen when the displayed candidate character is selected, thereby giving an instruction for selecting an intended object and an instruction for learning. Is possible. In the invention of claim 5, the character recognition unit has redundancy in recognition of candidate characters, and displays characters with some possibility as candidates.

[0017]

EXAMPLES A handwritten character recognition apparatus according to the present invention will be described below based on examples. (First Embodiment) FIG. 1 is a block diagram of a first embodiment of an online handwritten character recognition apparatus according to the present invention. In the figure, 11 is a character input unit, 12 is a character recognition unit, 13 is a candidate character display unit, 14 is a character selection unit, 15 is a learning determination unit, and 16 is a recognition result output. 17 is a character registration unit, and 18 is a dictionary unit for recognition. In addition to the above, a floppy disk storage portion, various operation portions, and the like are provided, but since these are obvious, illustration and description thereof are omitted.

FIG. 2 is a diagram for explaining how the screen is displayed in this embodiment. In the figure, 21 is a display of a handwritten character group, 22 is an input frame for handwritten character input, 23 is a displayed candidate character group, and 24 is whether or not to learn. Is a button for instructing. Since it is directly related to the gist of the present invention,
Here, the principle and configuration of the button will be described. In the present embodiment, metal fine wires are embedded near the surface of the button at coarse intervals, and an alternating current is flown through the metal fine wires. On the other hand, a magnet is embedded in the tip of the pen. Therefore, when the handwriting input person touches the button with the pen, the magnetic force disturbs the alternating current of the thin metal wires in the button, and by detecting this, it is possible to input the touch button operation and eventually the predetermined instruction can be input. It is said that.

FIG. 3 shows a change in the screen state when the "confirm" button for selecting or specifying the character intended by the input person from the characters displayed as candidates from the state of FIG. 2 is pressed. It is a figure. In the figure, 31 indicates the display change of the "confirm" button, and 32 indicates the state of the input text. FIG. 4 is a diagram showing handwriting data and character codes to be learned. In the figure, 41 is handwriting data, and 42 is a character code of a character to be learned.

Next, the learning operation of this embodiment will be described with reference to FIGS. First, the input person inputs characters by handwriting from the character input unit 11. The character input unit 11
Upon receiving the input handwritten character data, information is extracted in the form of a coordinate point sequence, stroke data, segment data, and the extraction result is passed to the character recognition unit 12 as input handwriting data. The character recognition unit 12 recognizes the input character by collating the received handwriting data with the dictionary data in the recognition dictionary unit 18. Here, the recognition of a character means, for example, obtaining a character code corresponding to the character. The character recognizing unit 12 extracts a plurality of candidates in order from the one having the highest possibility in consideration of the matching degree and the frequency of use in the characters in the dictionary unit according to the recognition result, and passes them to the candidate character display unit 13. The candidate character display section 13 is
The received candidate characters are displayed as indicated by 23 in FIG. At the same time, the character selection unit 14 makes it possible to determine that one of the candidate characters is intended or to instruct learning. The character that is first set in the instructable state is the first candidate character that has the highest possibility. It should be noted that the characters that can be normally designated are displayed in reverse video so that they can be distinguished from other candidate characters. The input person designates, from the displayed candidate characters, the one he or she intended to input the instruction.
The character selection unit 14 changes the candidate character designated by the input person to the instructable state. In FIG. 2, the input person designates the fourth candidate “heaven” character, and the character selection unit 14 is capable of instructing confirmation and learning that this character is the character intended to be input. This is indicated by reverse display. Next, the input person is instructed whether or not to learn the candidate character.

In this embodiment, two buttons of "confirm" and "learning" are prepared as instructions by the input person, and the learning judging section 15 indicates "in case of not learning the candidate character in the instructable state". The presence or absence of learning is identified by lightly pressing the "confirm" button or the "learning" button when learning. If the input person presses the "confirm" button in the state shown at 31 in FIG. 3, the character "heaven" is simply confirmed as the character intended to be input, and the input text is changed as shown at 32. The word "heaven" is entered. At this time, it goes without saying that the learning of characters is not performed. Similarly, when the input person instructs the “learn” button, the learning determination unit 15
Informs the character registration unit 17 that the input handwriting data should be learned as the character "heaven". As shown in FIG. 4, the character registration unit 17 includes the handwriting data 41 extracted by the character input unit 11, the character "heaven" selected by the character selection unit 14, and the character code 42 (lexically) of the character. Here, new data is generated in association with JIS code 4537), and this is registered in the dictionary unit 18 for recognition. When a character similar to the handwriting registered here is input at the time of subsequent handwritten character input, this is registered in the recognition dictionary unit 18, so this is recognized as the immediate character "heaven". Will be done. At the same time, similarly to “confirmation”, characters may be input to the text being input. (Second Embodiment) Next, another embodiment of the handwritten character recognition device according to the present invention will be described with reference to the drawings. In this embodiment, a button for instructing whether to learn is selected,
The characters are displayed adjacent to the characters for which learning can be instructed. Specifically, the only pair of buttons for instructing the input person whether to learn or not are not provided at the dedicated fixed position,
When the candidate character display unit 13 displays the candidate character, the candidate character display unit 13 is provided so as to display the candidate character adjacent to the character that can be selected and learned by the current character selection unit 14. This state is shown in parts 51 and 52 of FIG. (However, in FIG. 5, two sets and two types of buttons are broadly shown for easy understanding of the following operation.) Hereinafter, the operation and action peculiar to this embodiment will be described with reference to FIGS. 1 and 5. Will be explained. For example, in 51 of FIG. 5, when the input person designates the character “MA” as the character to input the instruction, the “confirm” and “learn” buttons for instructing whether or not to learn are It is displayed adjacent to the bottom of the part where "Ma" is displayed. In 52, the input person selects the character “heaven”, and like in 51, two buttons are displayed adjacent to the lower part of the portion where “heaven” is displayed as a candidate. In this case, "heaven" is selected for the two buttons directly below the "ma",
Needless to say, it is dismissed when the instruction is given for learning. (Third Embodiment) In this embodiment, an instruction as to whether or not to learn is made not by pressing a button but by using a gesture, which is shown in FIG. Here, various gestures are defined in advance according to the shape of the trajectory written with a pen, and the device side must analyze the trajectory input by the pen for a predetermined instruction by the handwriting input person. The operation corresponding to it is performed. The basic principle of the gesture is to recognize a character from handwriting data and display the character on the display unit to recognize an operation command to perform this operation, which is not different from character recognition structurally. In addition, this is "handwriting input O
Birth of S, opening up a new market for personal computers: Nikkei Electronics No. 525, pp. 115-133 (199
This is a well-known technique described in the literature such as "1.4.15)", and the recognition of various trajectories and the correspondence between this and commands correspond to electromagnetic signal matching, which is not different from character recognition. . Therefore, detailed description thereof will be omitted. Now,
In the present embodiment, the downward convex gesture as shown by 61 in FIG. 6 is not learned for the character pointed by the convex portion (bent point portion) of the gesture, and the character is input by the input person. It is assumed that the upward convex gesture as shown by 62 indicates that it is to be determined that it was intended. For example, in 61 of FIG. 6, since the character “Ma” is convex downwards, the input person selects the character “Ma” from the characters displayed as candidates as the character intended to be input, It also indicates that the user has instructed not to study. When the input person inputs this gesture, the character selection unit 14 determines that the character “MA” is the character intended to be input by the input person, and at the same time, the learning determination unit 15 does not learn this character. judge. In addition, 62 in FIG.
Then, the input person makes a character "heaven" by a convex gesture.
Is designated as the character intended to be input, and it is indicated that the user has instructed to learn that the input handwriting data is the character "heaven". As a result, the character selection unit 14 determines that the character “heaven” is the character intended to be input and displays it at a predetermined position in the text being input, and the learning determination unit 15 determines the input handwriting data as the character. It is determined that learning is performed as "heaven." (Fourth Embodiment) In the present embodiment, the candidate character display unit 13 has redundancy, that is, not only the candidate character that the character recognition unit 12 obtains as a recognition result in a normal procedure but also a character to be recognized to some extent. If it is judged that it is added to the candidate character and displayed so that it is displayed,
This state is shown in FIG. It should be noted that this is performed by outputting all the characters included in the handwriting data, which have some features of strokes, etc., even if there are some differences.

The operation of this embodiment will be described below with reference to FIG. The character recognition unit 12 compares the input handwritten data with the data corresponding to the character in the recognition dictionary unit 18, and passes the recognition candidate character group to the candidate character display unit 13. At this time, the recognition candidate character group received from the character recognition unit 12 includes characters that are considered to be somewhat similar to the input character. And the candidate character display unit 1
3 displays a list of all these characters received from the character recognition unit 12. The display device according to the present embodiment is a portable device, and since the display unit is small, only 27 display devices can be displayed at the same time. Therefore, as shown at 71, a display movement button is provided so that the front candidate character group and the rear candidate character group can be displayed.

In the present embodiment, the input person can select an intended character from among the characters that are likely to be present. Therefore, the character recognizing unit 12 has entered a recognition candidate in a normal procedure because an extreme collapsing character or an abbreviation has been input. Characters that are not output as characters will also be presented. Therefore, it is possible to select and learn the character intended to be input from these wide recognition candidate characters. As a result, it becomes possible to even study curriculum and cursive. Further, it can be used for the use of a person with a handicapped arm, the lower grades of elementary school, or the study of characters that also play as an infant. It should be noted that characters other than the characters that are candidate characters in the normal procedure may be displayed in various orders such as the order with the smallest number of strokes, the Japanese syllabary order, and the character code order. It may be possible to easily find the character to be selected by setting the setting to.

Although the present invention has been described above based on the embodiments, the present invention is not limited to the above embodiments, and of course includes the following. (1) In the first embodiment and the like, the button operation with the pen is detected not by the magnetic disturbance of the alternating current but by the distortion or the change in the capacitance. (2) In the second embodiment, the "confirm" and "learn" buttons are a set, and the target candidate characters are moved and displayed above these buttons. (3) In the first embodiment and the second embodiment, the handwriting input of the next character also serves as the “confirm” instruction of the candidate character displayed as the input of the current instruction. (4) In the first embodiment and the second embodiment, "confirmation",
The "learn" button is provided in front of the displayed candidate characters. Note that this is a CRT that displays candidate characters just as it is used in automatic cash dispensers at banks.
It can be easily realized by providing a transparent input panel on the front surface of the etc. (5) In the first and second embodiments, the position of the button for instructing “confirmation” and “learning” is above the candidate character for convenience of manufacture and arrangement with other buttons. Alternatively, the positions are separated by 1 or 2 cm even at the lower side. (6) In the third embodiment, the gesture is a concept including not only handwriting data but also an operation called tapping that taps a character with the pen tip. So, for example,
If tapping is once, selection for inputting the instruction of the character displayed in the next rank, and if it is twice within 1 second, it is confirmed or selected that it was intended to input the character, and 3
If it is a time, it is called learning. (7) In the fourth embodiment, whether or not the function of the present embodiment is exhibited depends on a separate instruction from the handwriting input person. Alternatively, further, the recognition range is a step type (multistage type). (8) In each of the embodiments, it is possible to cancel an instruction made by mistake by using the pen, or by pressing a button on the top of the pen opposite to the input side. (9) When learning the handwriting data of a character that is not registered in the dictionary or a character that is not recognized in advance together with the character, enter the character intended for input by the character code. Alternatively, instead of providing redundancy as in the case of the fourth embodiment, a function is added in which characters intended to be input are separately and correctly written in a large size. Along with this, a function is added to enlarge the entry frame for one character. This allows
When a candidate character intended for learning is not displayed, the handwriting inputter does not have to refer to the character code one by one. Furthermore, it is possible to learn characters and the like that are not registered in JIS or the like. Then, in this case, it is still more convenient if the handwritten characters in this regular script are displayed in the text currently being created. (10) It goes without saying that the characters targeted by the present invention include not only kanji and kana but also alphabets, Arabic numerals, Hangul characters, Arabic characters, Russian characters, Greek characters, German characters and the like. Of course, it also includes gestures. In addition, letter means one (single)
It is needless to say that it includes not only the characters, but also a plurality of characters (in Japanese, since the distinction between the single and the double is not strict, in the present specification, it is described as “character” in principle). For this reason,
For example, a word entered in a cursive alphabet or an idiom composed of a plurality of kanji characters may be learned. Also in these cases, if the correct input of the kanji, etc. intended by the input person is added with the function that is performed by the side-by-side writing, the printed material, etc. as described above, a idiom such as "patented invention" is used. Needless to say, it becomes convenient to create various documents such as patent application specifications by learning with handwriting "/".

Further, in the case where a large number of types of characters can be input by handwriting, a function is added so that which type of character is to be input can be designated in advance in order to speed up the extraction of recognition candidate characters. This will be very effective when excluding Kanji. (11) For convenience of manufacturing, one component (requirement) of the present invention may be a plurality of components, or conversely, a plurality of components may be physically integrated components, or a combination thereof may be appropriately used. Furthermore, a part is configured as software.

[0026]

As described above, the present invention not only has improved learning ability and is convenient, but the invention of each claim has the following effects. According to the invention of claim 1, the input person can register the handwriting at any time as necessary while inputting the handwritten character. Therefore, it is possible to learn handwritten characters without performing a troublesome character registration operation.

According to the second aspect of the present invention, when instructing whether or not to learn after selecting a character, the amount of movement of the pen or the like is reduced and the instruction can be made more quickly. According to the invention of claim 3,
Instead of inputting the character you want to learn with the character code, you can just point with the pen tip from the displayed candidate characters,
In addition, since it is possible to instruct the displayed button with the pen tip whether or not to learn, the operation becomes easy.

According to the fourth aspect of the present invention, the character determination and the learning instruction can be performed in one operation by using the gesture. In the invention of claim 5, since an extreme collapsing character or abbreviation is input, even if the character recognition unit does not present the character to be input as the recognition candidate character, only that character can be selected from the list of recognition target characters. In addition, the handwriting data can be learned without specifying a character intended to be input by a character code or the like. Further, it is convenient for character education for children and use by persons with physical disabilities.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a first embodiment of a handwritten character recognition device of the present invention.

FIG. 2 is a diagram showing how a screen is displayed in the above embodiment.

FIG. 3 is an explanatory diagram showing a change in screen state when a confirm button is pressed from the state shown in FIG. 2 in the above embodiment.

FIG. 4 is a diagram showing the states of learned writing data and character codes in the above embodiment.

FIG. 5 is an explanatory diagram of a second embodiment of the handwritten character recognition device of the present invention.

FIG. 6 is an explanatory diagram of a third embodiment of the handwritten character recognition device of the present invention.

FIG. 7 is an explanatory diagram of a fourth embodiment of the handwritten character recognition device of the present invention.

FIG. 8 is a block diagram of an example of a conventional handwritten character recognition device.

FIG. 9 is a conceptual diagram showing a state in which learning characters are designated, combined with handwriting data, and registered in a recognition dictionary in the conventional example.

[Explanation of symbols]

11 character input unit 12 character recognition unit 13 candidate character display unit 14 character selection unit 15 learning determination unit 16 recognition result output unit 17 character registration unit 18 dictionary unit for recognition 21 input text (one of Part) 22 an input frame for inputting handwritten characters 23 a displayed candidate character group 24 a button for instructing the input person whether or not to learn

Claims (5)

[Claims] 1. A dictionary unit in which characters and handwriting data are registered in association with each other, a character recognition unit that recognizes characters based on input handwriting data, and a candidate character obtained by the character recognition unit is predetermined. Steps to display the candidate character display section, a character selection section that allows the input person to select the character intended for input from among the candidate characters displayed in the candidate character display section, and the input handwriting data In a handwritten character recognition device having a learning unit that performs a learning function by registering in a dictionary unit in association with a character specified by, when an input instruction is given to the input handwriting data by the input person, The input handwriting data is newly registered in the dictionary unit by directly associating the input handwriting data with the character selected as the character intended to be input by the input person from the candidate characters displayed corresponding to the learning unit. Handwritten character recognition apparatus characterized by having a learning instruction unit that. 2. The handwritten character recognition device according to claim 1, wherein the learning instruction unit is provided at a position close to the candidate character displayed on the candidate character display unit. 3. A pen instruction input unit that allows the character selection unit and the learning instruction unit to select a character intended to be input with a pen for handwriting input and to instruct learning. The handwritten character recognition device according to claim 1 or 2, which is characterized. 4. The pen instruction input unit is capable of instructing selection and learning of an intended object by recognizing a predetermined locus created by a pen for handwriting input with respect to displayed candidate characters. The handwritten character recognition device according to claim 2 or 3, wherein 5. The handwritten character recognition device according to claim 1, wherein the character recognition unit has redundancy in recognition of candidate characters.