JPH06266898A - Character recognition processing method and device for input of hand-written character - Google Patents

Abstract

PURPOSE:To prevent a hand-written character from being swollen from a recognition frame at the time of inputting to reduce erroneous recognition and to improve an input speed, that is, to improve the aspect of a human interface. CONSTITUTION:The 1st character e.g. in a character string inputted by a user is recognized, after the recognized result is displayed, the coordinates of an input character start point and character size data are extracted (9a), and then the size of a recognition frame is determined based upon the extracted data (9b). Namely when x1 and y1 are respectively defined as the initial sizes in X and Y directions, these values are corrected to x2 and y2. The values x2, y2 can be made larger or smaller than the values x1, y1 in accordance with the inputted character size.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character input area in which a position input device and a display device are integrated, and when a character to be input on the position input device is recognized, the character input area is determined by the size of the character input first. Improve the input interface by changing the preset grid-shaped recognition frame and character inspection frame information,
As a result, it relates to improvement of recognition speed and input speed.

[0002]

2. Description of the Prior Art In conventional handwritten character input devices, a method of inputting a character in a free format area without providing a recognition frame in the input area so that the user can easily input the input position, and input speed Alternatively, there has been a method of providing a recognition frame in an input area for the purpose of improving the recognition speed and inputting characters into this recognition frame.

In the former method, an arbitrary size is input in an arbitrary position in a character input area, one character is input, a next input is waited for a while, and a cutoff for character recognition is performed by the waiting time. It is to recognize.

In the latter method, a recognition frame is set in the character input area, a character is input in the recognition frame, and the character is cut off by the frame cutoff for recognition.

[0005]

However, the above-mentioned conventional methods have merits and demerits, and when viewed as a human interface, it is difficult to say that they are all good.

First, a method of recognizing a character input by handwriting on a free format area in which a character input position is not specified is excellent as a human interface in that a user can input an arbitrary size at an arbitrary position. However, because the characters are cut off depending on the time,
Since the next character cannot be input until it is recognized even after the input of one character is completed, there is a problem that it takes time to input, and it cannot be said that the human interface is good after all.

Further, in the method of recognizing by providing a character recognition frame in the character input area, if the character is properly fitted in the frame to input characters, the recognition rate is high and the recognition speed / input speed is improved. However, in other words, since the user cannot input characters at arbitrary positions and the size of the recognition frame is predetermined, the frame set to improve the character recognition processing speed is not set by the user. If the input is larger than the recognition frame, the recognition will start in the middle, which will cause erroneous recognition, and eventually the input speed will be slow and it cannot be said that the human interface is good.

The present invention has been made in view of the above problems of the prior art, and an object of the present invention is to provide a character recognition processing method for handwritten character input and an apparatus thereof which improve both operability and recognition speed. To provide.

[0009]

According to the character recognition processing method for handwritten character input of the present invention, a recognition frame for character inspection is provided in a handwritten character input area of a position input means for inputting coordinate values corresponding to a designated position. A step of displaying the grid-shaped arrangement state on the display means superimposed on the position input means based on the recognition frame information preset for the grid-shaped arrangement;
The step of capturing the size information of the input handwritten character in the handwritten character input area, and changing the recognition frame information so that the recognition frame has a size according to the size of the input handwritten character, the The step of displaying the changed state on the display means is included.

Further, the handwriting character input character recognition processing device of the present invention includes a position input means for inputting coordinate values corresponding to the designated position, and an indication display of the various input operation areas overlapped with the position input means. Display means for displaying various information, handwritten character recognition means for recognizing an input handwritten character consisting of a coordinate value string input by an instruction operation to the handwritten character input area of the position input means, and within the handwritten character input area Recognition frame information setting means for arranging recognition frames for performing character cut-off in a grid shape, and the arrangement state for setting recognition frame information displayed on the display means, and the recognition frames for inputting handwritten characters The recognition frame information changing means for changing the recognition frame information so that the recognition frame information has a size corresponding to the size of the recognition frame, and the recognition frame displaying means for displaying the changed recognition frame on the display means. It is characterized in.

[0011]

According to the present invention, a grid-shaped recognition frame is set and displayed in advance in the handwritten character input area. The user can input based on this first. This allows
As compared with the case of the free format in which the input position is not specified, the user can be prompted to input without feeling uncomfortable, and the input can be performed smoothly by adding the frame inspection without relying on only the time inspection.

Further, as compared with the case where the user inputs a character in which a frame for character inspection (recognition) is set within the normal character recognition frame, the character recognition frame corresponding to the size of the initially input character is recognized. Also, since the recognition frame information is reset, it is possible to provide a recognition frame suitable for the user's input. As a result, it is possible to reduce erroneous recognition, improve the input speed, that is, improve the human interface.

[0013]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a schematic configuration of a handwritten character input device according to an embodiment of the present invention.

As shown in this figure, the device of this embodiment comprises a position input device 1, a display device 2, a processing control device 3 and an external storage device 4. The position input device 1 is for inputting handwritten position coordinates, and the display device 2 is for displaying a coordinate value string (characters) input by the position input device 1 or a recognized character. The processing control device 3 is
A function for presetting a grid-like character inspection (recognition) frame (hereinafter referred to as a recognition frame) in the handwritten character input area and setting recognition frame information corresponding to the recognition frame, and a control function for an online handwritten character recognition method. A function of recognizing a designated coordinate value string handwritten via the position input device 1 as a character and displaying the result on the display device 2, a recognition frame corresponding to the handwritten character size initially input, and It has a function of setting recognition frame information and various other processing functions. The external storage device 4 stores a dictionary for recognizing characters input from the position input device 1, a system application for driving this device, and the like.

FIG. 2 shows an external structure of the handwritten character input device.

As shown in the figure, the position input device 1 is composed of a transparent tablet 1a and a stylus pen 1b, the display device 2 is composed of, for example, a liquid crystal display, and the transparent tablet 1a and the display device 2 are the same. It has the same coordinate plane of dimensions. These are transparent tablets 1a
Are superposed in a positional relationship with the front side of the display device 2, and are integrated into a laminated structure. Therefore, the display device 2
The information displayed on is visible through the transparent tablet 1a.

3 to 5 schematically show the internal structure of the processing control device 3.

First, as shown in FIG. 3, the processing control device 3 includes an input control unit 3a, a recognition processing unit 3b, a recognition frame resetting processing unit 3c, a display control unit 3d, a position data buffer 3e, and an input character size data buffer 3f. And a recognition frame data buffer 3g and an output data buffer 3h. As shown in FIG. 4, the input character size data buffer 3f comprises a starting point X coordinate data buffer 4a, a starting point Y coordinate data buffer 4b, a character size X data buffer 4c and a character size Y data buffer 4d. As shown in FIG. 5, the recognition frame data buffer 3g includes a frame number data buffer 5a, a frame start point X coordinate data buffer 5b, a frame start point Y coordinate data buffer 5c, and a frame end point X coordinate data buffer 5.
A buffer unit for one frame is configured from d and the frame end point Y coordinate data buffer 5e, and the buffer units are provided for the number of frames.

Each constituent element will be described in detail below.

First, FIG. 6 shows the configuration of the relevant portions of the position input device 1 and the process control device 3.

In this figure, the transparent tablet 1a is of an electrostatic type, and as shown in FIG. 6 (a), a pair of transparent glass plates 1ax, 1ay are joined together via an insulating protective layer (not shown). It is a thing. Transparent glass plate 1ax,
A plurality of transparent electrodes 1c and 1d are arranged in parallel at a predetermined interval on 1ay, respectively.
x and 1ay are joined so that the transparent electrodes 1c and 1d are orthogonal to each other. Then, as shown in FIG.
Each of them is connected to shift registers 1f and 1g which operate under the control of the shift controller 1e.

The specific operation is as follows.

First, when measuring the horizontal (X direction) coordinate, a predetermined voltage is sequentially applied to the transparent electrode 1c connected to the shift register 1f from the end. When the electrode to which voltage is applied approaches the stylus pen 1b, the transparent electrode 1
An electrostatic induction voltage is induced in the stylus pen 1b by the electrostatic capacitance formed between c and the stylus pen 1b. This electrostatic induction voltage is amplified by the amplifier 1j,
By removing the high frequency component by the filter 1k, a stable pointing position detection signal is provided and provided to the counter 1h. The counter 1h starts counting when a predetermined voltage is applied to the electrode located closest to the origin among the transparent electrodes 1c, and ends counting when a designated position detection signal of a certain value or more is input. . By performing such an operation, the coordinates of the stylus pen 1b are converted into time and detected.

When the horizontal coordinate measurement is completed, the vertical coordinate measurement is started. In the vertical coordinate measurement, a predetermined voltage is sequentially applied from the end of the transparent electrode 1d connected to the shift register 1g as in the horizontal coordinate measurement.
An amplifier 1j and a filter 1k generate a pointing position detection signal from the electrostatic induction voltage detected by the stylus pen 1b, and the counter 1h detects coordinates.

When writing characters or other patterns on the transparent tablet 1a using the stylus pen 1b,
The information indicating the writing stroke is input as a connection of the designated positions by the stylus pen 1b, that is, a connection of the designated position coordinates (position coordinate series).

As described above, the characters or other pattern information handwritten and input through the position input device 1, that is, the sequence of position coordinates, is displayed at the same position coordinates on the display device 2 through the processing control device 3. It is designed to be driven. By this display control, the pattern input by handwriting on the transparent tablet surface is simultaneously displayed as the handwriting input locus, and characters or other patterns are input as if writing on the paper surface using a pen. .

Next, FIG. 7 shows a processing procedure of the processing control device 3 including character input / recognition in character input processing and resetting of a recognition frame.

In this figure, first, an initial screen as shown in FIG. 10 (a) is displayed (step 7a). Here, the initial screen includes a character input area 10a for inputting handwritten characters and a command area 10b for displaying various command menus in a divided state. Character input area 1
A recognition frame 10d is arranged in a grid pattern at 0a. In the command area 10b, a command menu frame 10e for displaying each command menu and performing selection input is arranged. The grid-shaped array of the recognition frame 10d is based on the frame data stored in advance in the recognition frame data buffer 3g.

When the user first inputs a character in the character input area with the stylus pen 1b, the switch state and the coordinate value of the stylus pen 1b when the user inputs the position input device 1 are detected (steps 7b to 7b). 7d), the character recognition flag is set (step 7e). Also,
At the same time, the recognition frame data stored in advance in the recognition frame data buffer 3g is set (step 7f).

Then, character recognition processing is performed for the character input by the user with reference to the recognition frame set (step 7g).

If the input character is the first character (step 7h), the recognition frame and the frame information are reset for the input character (step 7i). By this process,
Providing a recognition frame that matches the character size of the user's input, the user does not get out of the frame during input, and recognition is not performed unexpectedly, which improves input speed by reducing false recognition. Can be planned.

When the input is completed, the stylus pen 1b is outside the character input area, and the character recognition flag is set, the flag is reset (step 7).
j, 7k).

Here, the above-mentioned character recognition processing and recognition frame resetting processing will be described in detail.

FIG. 8A shows the flow of the character recognition process.
(B) to (d) show the character recognition principle.

In FIG. 8A, when the user performs handwriting input in the character input area on the display screen, the processing control device 3 detects the ON state of the switch of the stylus pen 1b (step 8a). The position data of the designated position detected via the position input device 1 is stored in the position data buffer 3e by the input control unit 3a (step 8b). The character is cut off when the switch of the stylus pen 1b is turned off in the case of the first character, and when the OFF time exceeds a preset fixed time, that is, the time cutoff process is performed (steps 8c and 8d). ). Further, in the case of the second and subsequent characters, the recognition frame is moved out of the frame with the stylus pen 1b or the switch being in the ON state according to the set recognition frame data.
When the character is turned off outside the frame, that is, the character cutting is performed by the frame cutting processing (step 8e). At this time, the position data stored in the position data buffer 3e is simultaneously transferred to the output buffer 3h and displayed on the display device 3 at a position corresponding to the data, thereby being displayed as handwriting pattern information. In this way, when the position data series for one character is obtained in the position data buffer 3e, the recognition processing section 3b analyzes the position data series (step 8).
h), character recognition is performed (step 8i).

In character recognition, the position data sequence is divided for each writing stroke according to the switch state of the stylus pen 1b, and a direction value sequence showing the characteristic is obtained for each writing stroke. Here, the direction values are quantized in eight directions (direction values) as shown in FIG. 8C. In addition to the conversion of the direction value of the writing stroke, a stroke position code indicating the position of the writing stroke is obtained. The recognition result is calculated by referring to the information obtained by the direction value string conversion process and the position code conversion process of the writing stroke and the handwritten character recognition dictionary stored in the external storage device 4.
As shown in FIG. 8D, the handwritten character recognition dictionary used for recognition processing is represented by a direction value string of strokes forming a recognition target character and a position code, classified by the number of strokes of each character, and each character code. It is configured to correspond to. The result recognized as described above is output to the output data buffer 3h.
And display it on the display device 3 (step 8m).
For example, if the number "3" is input as shown in FIG. 8B, the stroke position code is "2523456.
7 ", and based on this, the character code is obtained by referring to the dictionary shown in FIG. 8D, and the character corresponding to this character code is displayed on the display device 2. The first character (first input Character), the starting point coordinates of the input character and the character size data are stored in the input character size data buffer 3f (steps 8k and 8l), and after displaying the recognition result, the process returns to the main routine (step). 8
n).

FIG. 9 shows the flow of the recognition frame resetting process.

In this figure, the first character input by the user is handwritten character recognition by time cutoff, and after the recognition result is displayed, the input character size data buffer 3f is displayed.
The input character starting point coordinates and the character size data stored in (4) are extracted (step 9a), and the size of the recognition frame is determined based on the extracted data (step 9b).

That is, x1, y1 shown in FIG.
Where x is the initial size in the X direction and Y is the initial size, these values are corrected to x2 and y2. These x2, y2
Will be larger or smaller than x1 and y1 depending on the input character size, but in the example shown in the figure, x2 and y2 are larger than x1 and y1. If the character size is x1 = 40 dots and y1 = 50 dots, x2 = y2 = 60
When the dots are formed, the recognition frame 10d is in a state as illustrated.

The frame size is determined in step 9b by the following processing, for example. First, the size of the recognition frame is calculated based on the input character start point coordinates and the character size data extracted in step 9a. At this time, the conditions such as allowing a slight margin from the character size are taken into consideration. next,
The calculated frame size value and the initial frame size value are compared, and if they are different, it is decided to adopt the newly calculated frame size.

Then, the frame size and the recognition data set in advance by the newly obtained frame size are reset (step 9c). That is, in this case, FIG.
Each data shown in FIG. 3 is updated, and the total number also changes if the size of the frame changes. Therefore, the frame number data 5a is updated accordingly, and at the same time, the coordinate data 5b to 5e of the frame start and end points of each frame. Will be updated. For character recognition of the second and subsequent characters, frame inspection is performed based on the recognition frame data that has been reset. After resetting the recognition frame data, the process returns to the main routine (step 9d).

As described above, according to the present embodiment, a grid-shaped recognition frame is set and displayed in advance in the handwritten character input area, and the user first makes an input based on this.
After that, since the recognition frame information and the recognition frame information are reset according to the input character size, it is possible to provide the recognition frame that matches the input by the user. As a result, it will not be pushed out from the inspection (recognition) frame when inputting,
As a result, false recognition can be reduced and input speed can be improved, that is, human interface can be improved.

In the above embodiment, the character input processing is performed as described above, but the present invention is not limited to the above embodiment.

For example, the position input device 1 is not limited to the electrostatic type which utilizes the electrostatic capacitance between the stylus pen and the tablet, but is of the pressure sensitive type in which the tablet detects the contact pressure from the stylus pen. It may be one. Also,
The tablet may be an opaque tablet.

The display device 2 may be composed of a plasma display or the like in addition to the liquid crystal display. Further, regarding the display of the recognition result after the handwritten character is input, the input area and the display area may be the same area as described above, or may be different areas different from each other. The size is not limited to the state as in the above embodiment.

The recognition frame is not limited to the one having the frame pitch as in the above embodiment, but may have the frame pitch set. Further, the recognition frame may be set for vertical writing. Furthermore, the recognition frame on the display and the recognition frame data for inspection may or may not match.

Further, in the above embodiment, the size of the recognition frame is determined on the basis of the input character size of the first character, but the input character size may always be corrected. The main routine in this case is modified as shown in FIG. The routine shown in this figure will be described in relation to that shown in FIG.
11g, 11j and 11k are steps 7a to 7g in FIG.
It is the same as 7j and 7k. In the routine shown in FIG. 7, the recognition frame resetting process of step 7i is executed after confirming that the character recognized in step 7g is the first character in step 7h. However, in the case of always monitoring the character size shown in FIG. 11 and considering resetting, there is no processing corresponding to step 7h, and instead of step 1
The recognition frame resetting process of 1i is inserted. As a result, the size of the recognition frame is examined for each recognition of each input character, and even if there is a change in the input character size due to factors such as changing the user, it is possible to immediately cope with the change.

Alternatively, the resetting may be taken into consideration at an arbitrary interval (the number of input characters) without the need for constant correction. In this case, the main routine is as shown in FIG. In this figure, steps 14a to 14g correspond to steps 7a to 7g in FIG. 7 (steps 14j and 14k correspond to steps 7j and 7k, and step 14m corresponds to step 7i. Therefore, steps 14h, 14i and
14l is different.

That is ,. Each time one character of the variable i is recognized, it is incremented by 1 (ie, the number of recognized characters) is counted: Step 14h), It is determined whether or not i = k, the process proceeds to step 14j when the determination is NO, and the process proceeds to the next process (step 14i) when YES, a branching process ,. When the judgment is YES, "i
← 0 ”(that is, i is reset), and the process proceeds to the recognition frame resetting process 14m (step 14l).
By providing such a process, the user can perform the resetting process at an arbitrary recognized character number interval by setting the k value.

Further, in this case, as shown in FIG. 12, a menu for setting the recognition frame size resetting interval and character input is provided in the command area 10c, k is set by the former selection, and the character after the end is set. Character input can be started by selecting the input menu.

Next, in resetting the size of the recognition frame, the size is obtained by arithmetic calculation in the above embodiment,
If the initial value and the calculated value are compared and the latter is different, the latter calculated value is determined as the new adopted value as it is, but it may be changed as follows. That is, reference numeral 13- in FIG.
, 13n have frame size data corresponding to different frame sizes in stages, and compare the character size data with each frame size data. For example, the frame size data is the character size. A frame size satisfying a predetermined relationship that is larger than the data and closest to each other is determined as the adopted frame size.

In short, the present invention can be variously modified and implemented without departing from the scope of the invention.

[0054]

As described above, according to the present invention, a grid-shaped recognition frame is set and displayed in advance in the handwritten character input area, and thereafter, in the recognition frame corresponding to the size of the input character. Since the recognition frame information is reset, it is possible to provide the recognition frame that matches the input by the user. This allows
It is possible to prevent handwritten characters from protruding from the recognition frame at the time of input, which in turn reduces erroneous recognition and improves input speed, that is, the human interface.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of a handwritten character input device according to an embodiment of the present invention.

FIG. 2 is a three-dimensional block diagram showing an external configuration of the handwritten character input device.

FIG. 3 is a block diagram schematically showing an internal configuration of the processing control device shown in FIGS.

FIG. 4 is a block diagram showing the structure of an input character size data buffer shown in FIG.

5 is a block diagram showing a structure of a recognition frame data buffer shown in FIG.

FIG. 6 is a block diagram showing a configuration of a related portion of a position input device and a processing control device.

FIG. 7 is a flowchart showing the processing contents of a main routine including character input / recognition and resetting of a recognition frame in the character input processing.

8 is a flowchart (a) showing the contents of a subroutine corresponding to the character recognition processing shown in FIG. 7 and principle explanatory diagrams (b) to (d).

9 is a flowchart showing the contents of a subroutine corresponding to the recognition frame resetting process shown in FIG.

FIG. 10 is an explanatory diagram showing a screen display example of the embodiment shown in FIGS.

FIG. 11 is a flowchart showing a modified example of the main routine.

FIG. 12 is an explanatory diagram showing a display example of a command area when the frequency of resetting the recognition frame size can be arbitrarily set.

FIG. 13 is an explanatory diagram for explaining an image of stepwise frame size data corresponding to different frame sizes.

FIG. 14 is a flowchart showing another modification of the main routine.

[Explanation of symbols]

 1 position input device 1a transparent tablet 1b stylus pen 2 display device 3 processing control device 3a input control unit 3b character recognition processing unit 3c recognition frame resetting processing unit 3d display control unit 3e position data buffer 3f input character size data buffer 4a starting point x Coordinate data buffer 4b Start point y coordinate data buffer 4c Character size x data buffer 4d Character size y data buffer 3g Recognition frame data buffer 5a Frame number data 5b Frame start point x coordinate data 5c Frame start point y coordinate data 5d Frame end point x coordinate data 5e Frame End point y coordinate data 3h Output data buffer 4 External storage device 7a Initial screen display step 7g Character recognition step 7i Recognition frame resetting step 8l Starting point coordinate / character size storing step 9a Starting point coordinate / character size extracting step 9b Size decision step 9c frame coordinate data resetting step

Claims (2)

[Claims] 1. A recognition frame information preset for arranging a recognition frame for character inspection in a grid shape in a handwritten character input area of a position input means for inputting coordinate values corresponding to a designated position. And displaying the grid-like arrangement state on the display means superimposed on the position input means, the step of capturing the size information of the input handwritten character in the handwritten character input area, the recognition frame The step of changing the recognition frame information so that the input handwritten character has a size corresponding to the size of the input handwritten character, and displaying the changed state on the display means. Character recognition processing method. 2. A position input means for inputting coordinate values corresponding to a designated position, a display means for superposing the position input means for displaying various information including instruction display of various input operation areas thereof, and the position input means. A handwritten character recognition means for recognizing an input handwritten character consisting of a coordinate value sequence input by an instruction operation to the handwritten character input area, and a recognition frame for performing character cutoff in the handwritten character input area in a grid shape. Recognition frame information setting means for setting recognition frame information displayed on the display means, and the recognition so that the recognition frame has a size corresponding to the size of the input handwritten character. A character recognition processing device for inputting handwritten characters, comprising recognition frame information changing means for changing frame information and recognition frame displaying means for displaying the changed recognition frame on the display means.