JPH01261793A - Character input device - Google Patents

Abstract

PURPOSE:To improve operability by removing characters unsuitable for recognized candidates from the recognized candidates and displaying the minimum recognized candidates required. CONSTITUTION:A position input device 1 is superposed on a display device 3 and constituted as a transparent tablet unitedly laminated on a device body 2. A display control part 2d executes processing for removing candidate characters by canceling conditions when plural recognized candidates for a handwritten character exist, and when plural recognized candidates are still left as the result of 1st canceling processing, displays the recognized result having the succeeding candidate differently from the display format of other recognized results. When the recognized result having the succeeding candidate is specified through the device 1, the minimum required other recognized candidates found out about the input character having the recognized results are displayed on the vicinity of the specified character having recognized candidates differently from the display format of the character. Consequently, the succeeding candidate can be simply and efficiently selected.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object 1] (Industrial Application Field) The present invention relates to a character input device that efficiently selects recognition candidates for characters input by hand via a position input device. .

(Prior Art) With the development of information processing technology, various character input devices have been developed for inputting character information used in document creation and various information processing.

Most of this type of character input device is usually equipped with a keyboard for inputting character information and a display device (for example, CRT display, screen) for displaying the input character information. Character information is displayed at a specified position and the character can be entered.

On the other hand, in recent years, instead of character input devices using keyboards, position input devices such as tablets have been used, and character information (writing stroke information indicating character patterns) input by hand via this position input device is being used. A character input device using an online handwritten character recognition method has been developed, which recognizes characters in real time and inputs characters.

This type of device uses the number of writing strokes of a character pattern input via a position input device and the characteristics of the writing strokes, such as information on a direction value sequence indicating a change in the direction of the line segments that make up the writing strokes. It recognizes written character patterns and finds the corresponding character code.

According to such a character input device, it becomes possible to input characters very easily, in the same way as writing characters on paper using a pen.

However, in reality, there are various problems such as character recognition accuracy.
Characters input by hand are not always recognized accurately.

For this reason, generally, multiple recognition candidates obtained during the character recognition process are saved, and if the characters displayed as a recognition result are different from the characters input by hand, the so-called next candidate selection is performed. The idea is to select the correct recognition result.

Specifically, similar to the next candidate selection seen in kana-kanji conversion in an n-honsaku word processor, "r selection" is instructed for the recognition result where the next candidate exists, and other recognition is performed by the instruction operation of "next candidate". The display changes to candidates one after another.

Then, when the desired recognition candidate is displayed, "select"
It is considered that the recognition result can be determined by

Alternatively, when "select" is instructed for a recognition result that exists as a next candidate, other recognition candidates required for the input character are displayed in the corner of the display screen, and from among them, the input purpose is selected. The idea is to select and specify the recognition results to be used.

However, since this operation for selecting the next candidate is based on information input operations using a keyboard, a considerable number of key input operations are required to select and specify the desired character, making it extremely inconvenient to use. It was something.

Recently, handwritten input has been introduced in online handwritten character recognition type character input devices that recognize character information input online via a position input device such as a tablet, display the recognition result on a display device, and input characters. When there are multiple recognition candidates for a given character, the recognition result with the next candidate is displayed in a different display format from the other recognition results, and the recognition result with the next candidate is indicated via the position input device. displaying other recognition candidate characters sought for the input character for which the recognition result has been obtained in a different display form near the designated recognition candidate character,
A position input device is also being considered in which the input character is determined by selecting another recognition: & candidate character.

(Problems to be solved by the invention) In this conventional method of selecting recognition candidates, there are a large number of recognition candidates, and some of them are unsuitable as input characters. When making a decision, there is an inconvenience in that all input characters for which there is a next candidate character must be selected, making it very inconvenient to use.

The present invention has been made in consideration of such circumstances, and its purpose is to exclude characters unsuitable for recognition candidates from the recognition candidates and display the minimum necessary number of recognition candidates.

[Structure of the Invention] (Means for Solving the Problems) The present invention recognizes character information input online via a position input device such as a tablet, displays the recognition result on a display device, and inputs characters. In a character input device that uses an online cursive character recognition method, when there are multiple recognition candidates for a handwritten character, the candidate characters are divided by IJ+ based on a rejection condition, and as a result, there are multiple recognition candidates. Only in certain cases, the recognition result with the next candidate is displayed in a different display format from other recognition results,
When a recognition result having this next candidate is specified via the position input device, the other minimum necessary recognition candidates required for the input character that obtained the recognition result are This is a system in which characters are displayed in different display formats near the characters.

(Operation) According to the present invention, recognition candidates having a next candidate are displayed in a different display form from other recognition candidates, so that it is possible to clearly understand recognition results having a next candidate. .

When a recognition result having a next candidate is specified, the minimum necessary next candidates that the recognition result has are displayed in different display formats in the vicinity.

Then, by selectively specifying the desired recognition candidate from among these minimum necessary next candidates, the specified recognition candidate (
(next candidate) will be given as the selected recognition result.

Therefore, according to this device, the positioning can be performed without the complicated operation procedure seen in the procedure of selecting the next candidate by inputting information via the keyboard, and without having to select from among a large number of next candidates. It becomes possible to easily and efficiently perform the procedure for selecting the next candidate simply by directly specifying the position coordinates via the input device.

Therefore, it is possible to make effective use of the characteristics of information input via a position input device to significantly improve its operability, improve the usability of the device, and improve the man-machine interface.

(Example) Hereinafter, an example of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic configuration diagram of an example device.

This character input device using the online handwritten character recognition method basically includes a position input device 1 such as a tablet, into which characters are input by hand, and a position input device 1 such as a tablet that sequentially inputs position coordinates indicating the strokes of the input character pattern. A device main body (control unit) that recognizes input characters from a series of position coordinates indicating the strokes of characters input by hand via the device 1;
2, and a display device 3 consisting of, for example, a liquid crystal display, which displays character information etc. that are recognized and input by this device main body 2.
It is composed of:

The external storage device 4 stores, for example, recognition dictionary data necessary for processing the literary magazine 1 in the device main body 2.

Here, the position input device 1 and the display device 3 have the same dimensions and the same coordinate plane, and in particular, the position input device 1 is superimposed on the display device 3 as shown in FIG.
It is structured as a transparent tablet that is laminated and integrated with.

The information displayed on the display device 3 can be visually recognized via the position input device 1 consisting of a transparent tablet.

The device body 2 includes an input section 2a, a recognition processing section 2b, a next candidate processing section 2 (, a display control section 2d1, a position data buffer 2e, a tj document data buffer 2f, a recognition information buffer 2g, an output It is configured with a buffer 2h and the like.

First, the transparent tablet that is the position input device 1 will be explained.
As shown in the figure, it is realized as an electrostatic transparent tablet.

As shown in FIG. 3(a), this electrostatic type transparent tablet has a plurality of strip-shaped transparent electrodes 1c on one side.

A pair of transparent glass plates 1a and 1d each arranged in one row at a predetermined interval are transparent electrodes 1c.

1d are aligned perpendicularly to each other and bonded via an insulating protective film (not shown) to form a tablet surface.

As shown in FIG. 3(b), shift registers If and I operate under the control of the shift controller 1e.
A predetermined voltage is sequentially applied to the transparent electrodes 1c and 1d by g, and the entire area of the tablet surface is scanned as the intersection of the transparent electrodes 1c and 1d to which the voltage is applied.

Specifically, a predetermined voltage is selectively applied to one of the transparent electrodes 1c in order, and during a period when a predetermined voltage is applied to one of the transparent electrodes IC, the other transparent electrode 1d perpendicular to the predetermined voltage is applied to one of the transparent electrodes 1c. This is done by applying a predetermined voltage to all of them in turn.

That is, by sequentially applying a predetermined voltage to all of the other transparent electrodes 1d and each time scanning the transparent electrode 1d, by sequentially shifting the transparent electrode 1c to which a predetermined voltage is applied, a matrix formed by the transparent electrode IC%1d is created. The entire plane (tablet surface) is sequentially scanned.

The counter 1h performs a counting operation in synchronization with such scanning, and its counted value represents the position coordinates on the matrix plane.

The pen 11 that indicates the matrix plane (tablet surface) configured as described above detects the voltage applied by scanning the transparent electrode IC 11d at the indicated position by electrostatic coupling.

The scanning signal detected by this pen 11 is amplified via the amplifier 1j and then filtered (FIL).
The miscellaneous components are filtered out through lk,
It is given to the counter 1h as a designated position detection signal.

The counter 1h receives this indicated position detection signal and outputs the counted value at that time as information on the indicated position on the tablet surface indicated by the pen 11.

According to the position input device 1 configured in this way, the pen 1
By specifying a position on the tablet surface at 1, a scanning signal (applied voltage) that scans the transparent electrode +1i 1 c, 1d is detected by the pen 11 at the specified position.

The counter 1h at the time of detecting the scanning signal
The count value is detected as information on the position coordinates corresponding to the indicated position.

When a character pattern is written on the tablet surface using the pen 11, the information indicating the writing stroke is expressed as the connection of the positions indicated by the pen 11, that is, the connection of the indicated position coordinates (series of position coordinates). It will be entered.

Information on the character pattern input by hand through the position input device 1 in this manner, that is, a sequence of position coordinates, is provided to the display device 3 through the device main body 2.

According to this sequence of positional coordinates, the same positional coordinates on the display device 3 are driven to be displayed, as will be described later.

By this display control, the character pattern input by hand on the tablet surface is simultaneously displayed as the trajectory of the handwriting, and the character information is inputted as if writing characters on paper using a pen. .

Next, regarding the device main body 2 that recognizes and processes character information input by hand via the position input device 1 in this way, the sixth section
The explanation will be given according to the processing procedure shown in the figure.

The device body 2 first initializes various necessary data buffers, etc. (step a), then detects the on state of the pen 11 (step b), and reads the indicated position detected via the position input device 1. Coordinate data is taken into the input section 2a.

Then, refer to the recognition information buffer 2g (step C),
It is determined whether a recognition candidate having a next candidate is displayed at the position indicated by the coordinate data (step d).

If the input coordinate data does not indicate a recognition candidate with the next candidate, the pen-on check is performed again (step e), and the position data buffer 2e stores the coordinate data input in the pen-on state. (step f).

This coordinate data is stored when the pen 11 is turned off and the information is stored in the position data buffer 2e as character delimiter data.
(step g) until it is confirmed that the entire sequence of positional data representing a character pattern for one character has been input, that is, until the pen is turned off and a character break is detected. (step h)
. At this time, the coordinate data stored in the position data buffer 2eI is simultaneously transferred to the output buffer 2h, and is simultaneously displayed as information on the writing pattern on the display device 3 as described above.

When the sequence of coordinate data for one character is obtained in the position data buffer 2e, the recognition processing unit 2b analyzes the sequence of coordinate data and recognizes the input character (step i).

In this character recognition, for example, a series of coordinate data is divided into each written stroke by determining the stroke end from the distance between each position coordinate data, and is converted into a series of direction values indicating the characteristics of each written stroke.

Here, the direction value is quantized into eight directions (direction values) and given, for example, as shown in FIG. Then, determine what kind of direction value series represents the series of coordinate data indicating each writing stroke, and if the same direction (direction value) continues, those direction values are represented by one direction value. By doing so, the data is compressed, and the position/locus one-vote data indicating each writing stroke is converted into a direction value string.

Note that here, the direction of the stroke is quantized in eight directions and converted into a direction value sequence, but quantization in four directions or quantization in 16 directions may also be performed.

Through this process, the character pattern input by hand via the position input device 1 is expressed as a direction value string for each handwritten stroke.

Then, the input character is recognized by comparing the information of the direction value string thus obtained with the character recognition dictionary stored in the external recording device 4.

The character recognition dictionary used for this recognition process, for example, as shown in FIG. It is organized by classifying each character code and corresponding to each character code.

The recognition processing unit 2b refers to such a character recognition dictionary and obtains, as a recognition result, a character code that corresponds to the stroke characteristics of a handwritten character.

At this time, if a plurality of recognition candidates exist, the degree of difference between these recognition candidate characters and the input character (such as the degree of similarity indicating the degree of difference in character patterns) is determined at the same time.

In order to reduce the plurality of recognition candidates obtained in this way to the minimum necessary number, we set the rejection condition as an II + violation (degree of difference in character pattern) between the first recognition candidate character and the other recognition candidate characters. If the recognition candidate character has a difference of 5 or more in terms of class U, degree, etc. indicating ``class U, degree, etc.'', it is deleted from the recognition and a candidates (step j).

With this process, characters that were previously processed as shown in Figures 7 and 8 are now reduced to the necessary minimum by removing characters that are not suitable for recognition candidates from recognition candidates as shown in Figures 9 and 10. Recognition candidates are now displayed.

When the writing pattern for one character is analyzed in this way and the character is recognized, it is then determined whether or not the recognition result has a next candidate (step k). In other words, it is determined whether only one recognition candidate for the input character has been found, or whether a plurality of recognition candidates have been found.

If there is only one recognition candidate, this is determined as the recognition result, and the display control unit 2d normally displays the character pattern corresponding to that character code at the character input position on the display device 3 (step l). .

On the other hand, when multiple recognition candidates are narrowed down, the recognition candidates and their dissimilarity information obtained as described above are associated with the input character, and recognition is performed, for example, as shown in Fig. 10. The information buffer 2gt8 is stored (step m).

For example, the size of the character written at the coordinates (80, 32) is (16, 16), and the mosquito catch character is "ha".
"Ke" and "A" are respectively obtained, and information indicating that the degree of difference for these input characters is rlJ r3J "5" is stored in the recognition information buffer 2g as unselected information for the input characters.

However, in this case, the display control unit 2d controls the recognition information buffer 2.
According to the dissimilarity information of multiple recognition candidates stored in g,
As a recognition result, the most likely first recognition candidate is sought.

Then, this first recognition candidate is displayed as a recognition result with the next candidate in a display form of other recognition results, for example, as shown in Fig. 9 (a).
) is displayed with an underline as shown in (step n).

In other words, the display formats of recognition results that have a next candidate and recognition results that do not have a next candidate are made different, and in particular, the display format is changed and controlled so that the recognition result that has the above-mentioned next candidate is highlighted. There is.

By controlling the display format for the recognition results of input characters in this manner, the present device can easily recognize recognition candidates (recognition results) having the next Nhfi.

By the way, in the above-mentioned determination process (step d), it was shown that the coordinate data input from the position input device 1 is a recognition result with a next candidate whose characters have already been input and displayed on the display device 3. If so, it is determined that selection of the next candidate for that recognition result has been instructed.

For example, the display format of characters can be appropriately adopted other than the display formats described above.

Further, the display format may be changed by, for example, changing the display brightness or display color, or it is also possible to adopt a display format such as inversion or blinking.

In this case, first, the character pattern displayed at the position specified by the input coordinate data is displayed in a display format different from the previous display format (step 0
). For example, the character "wa" shown in FIG. 9(b) is displayed in reverse video.

Then, the next candidate sought for this character is obtained from the recognition information buffer 2g, and the character pattern corresponding to the next candidate is displayed near the highlighted character and in a display format different from the previous display format. (Step p)
.

For example, as shown in FIG. 9(b), the next candidate character is surrounded by a next candidate window frame and displayed near the highlighted character.

By the way, by displaying this next candidate character, the characters that were already displayed in the vicinity of the reversed character are temporarily not displayed, but since the character of interest here is the reversed character to be selected as the next candidate, it is not displayed at all. There will be no hindrance.

With the next candidate character displayed in this manner, the pen-on check is performed again (step q), and it is determined whether the next candidate character has been selected or not (step ``)''.

When one of the next candidate characters is selected, first, step S) of deleting unselected information consisting of the recognition candidate and dissimilarity information regarding the corresponding character from the recognition information buffer 2g; The display of the window and its next candidate character is canceled (step t).

Then, the highlighted character is replaced with the next [ti selected character, which is normally displayed, and the next candidate selection process ends (step U).

Note that if the next candidate character is not selected, for example, if coordinate data other than the position coordinates specifying the next candidate character is input by pen-on, this is determined to be an abort of the next candidate selection procedure.

Then, the display of the next candidate window and the next candidate character is stopped (step v), and the display form of the highlighted character is changed to the state before the next candidate selection procedure is started, that is, the next candidate character is recognized as having a next candidate. In step w) of returning to the underlined display indicating the result, the mode again operates as a character input mode.

In this way, this device recognizes characters input by hand via the position input device in real time, and when inputting characters while displaying the recognition results, if more than one recognition candidate for the input character is lost. , the first candidate character is displayed as a recognition result with a next candidate, distinguishing it from other recognition results.

Therefore, since the display form of the recognition result having the next candidate is different from other recognition results, it is possible to easily understand this.

Then, when a recognition result character with a next candidate is selected by the input coordinate data, this is interpreted as a designation to immediately select the next candidate, and the display form of that character is changed.
In addition, the minimum necessary next candidate characters for that character are displayed in the vicinity of the character.

When the next candidate character is selected, the recognition result is determined based on the selected character.

Therefore, as a candidate character selection procedure, the character input device 1
The operation procedure has been greatly simplified, with only the operation of specifying the displayed character to be selected above, and the operation of selecting and specifying the character to be input from among the next candidate characters displayed near the specified character. It becomes possible to aim for.

Therefore, the input of character information and the selection of the next candidate via the position input device I can be performed very easily and efficiently, and the operability can be greatly improved.

Note that the present invention is not limited to the embodiments described above.

Although the rejection condition is set to 5 or more in the embodiment, this value may also be changed as necessary.

Furthermore, regarding the input character recognition method, etc., various methods that have been proposed in the past may be adopted depending on the specifications.

Furthermore, the position input device 1 may be a pressure-sensitive tablet, and the display device 3 may be configured by a plasma display or the like.

In short, the present invention can be implemented with various modifications without departing from the gist thereof.

[Effects of the Invention] As explained above, according to the present invention, it is possible to perform character input processing via a position input device and subsequent candidate selection procedure in a very convenient way, and the operability is greatly improved. It has great practical effects, such as being able to achieve the following.

[Brief explanation of the drawing]

The figures show one embodiment of the character input device according to the present invention. FIG. 1 is a schematic configuration diagram of the embodiment device, FIG. 2 is a diagram showing the structural features of the embodiment device, and FIG. 3 is a position diagram. FIG. 4 is a diagram showing an example of the configuration of an input device, FIG. 4 is a diagram showing an example of a direction value expressing the direction of a writing stroke, FIG. 5 is a diagram showing an example of the configuration of a recognition dictionary, and FIG. 6 is a diagram showing a character input process. A diagram showing the flow of selecting the next candidate, FIG. 7 is a diagram showing an example of controlling the conventional character display form,
FIG. 8 is a diagram showing an example of unselected information stored in a conventional recognition information buffer, FIG. 9 is a diagram showing an example of controlling the display form of characters according to the present invention, and FIG. 10 is a diagram according to the present invention. It is a figure showing an example of unselected information stored in a recognition information buffer made by. 1... Position input device, 2... Control unit, 2a...
・Input section, 2b... Recognition processing section, 2C...
Next candidate, 2d...Display control unit, 2e...Position data buffer, 2f...Dictionary data buffer, 2g...Recognition information buffer, 2h...Output buffer,

Claims (2)

[Claims] (1) A position input device that inputs coordinate values corresponding to the indicated position, and analyzes a series of coordinate values sequentially input through this position input device, and generates recognition candidates for characters written on the position input device. , a character recognition device that obtains attribute information regarding the recognized character; a storage device that compares the recognition candidates and the attribute information obtained by the character recognition device with rejection conditions and stores recognition candidates that are satisfied; , a display device that displays a character pattern corresponding to a recognition candidate obtained by the character recognition device according to its attribute information; and a display device that displays the character pattern displayed on the display device at a position specified via the position input device. and means for specifying from the coordinate values, and the display control means selects one of the recognition candidates when a plurality of recognition candidates are obtained for the input character, and displays the recognition candidate differently from the recognition candidate having no next candidate. displaying a character pattern corresponding to the recognition candidate in a form, and when a character pattern on the display device of a recognition candidate having a next candidate is specified via the position input device;
A next candidate for the input character for which a recognition candidate corresponding to the character pattern has been obtained is obtained from the storage device, and a character pattern corresponding to the next candidate obtained from the storage device is separately displayed in the vicinity of the specified character pattern. A character input device characterized in that it is displayed in a form. (2) The character input device according to claim 1, wherein the position input device is laminated and integrated with the display surface of the display device and its coordinate area equal to each other.