JPH0969137A - Method for extracting radical, method and device for recognizing character in radical unit - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a radial extracting method, a radical unit character recognizing method and a device capable of accurately dividing each character into radical units from information such as pen-up time for input, pen holding time and pen pressure by the use of on-line constitution. SOLUTION: A coordinate input device 1, a controller 2 being a device main body, a display device 4, and an external storage device 4 are provided, thereby accurately dividing a character into radical units and improving a character recognition rate.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a radical extraction method, a radical unit character recognition method, and a device for recognizing a character for each radical based on a writing stroke input by handwriting via a coordinate input device. .

[0002]

2. Description of the Related Art A handwriting input device in which a coordinate input device and a display device are laminated and integrated is capable of inputting characters and figures on paper, and can be directly pointed to a desired place. It started to come.

However, current handwritten character recognition technology is not sufficient, and there is a growing demand for more efficient and accurate handwritten character recognition. As one of the methods, a radical-based character recognition method has been considered in which one character is recognized in radical units and the recognition result is output as a character from the radical recognition result. However, it is difficult to divide into radicals with the conventional technology, and the recognition rate as a character has not been satisfactory.

[0004]

As described above, in the conventional technique, it is difficult to divide into radicals, and the recognition rate of characters is not satisfactory. Therefore, the present invention has been made in consideration of the above circumstances, and utilizes the point of being online, from the information such as the pen-up time at the time of inputting, the stagnant time of the pen, and the writing pressure of the pen to the radical unit. An object of the present invention is to provide a radical extracting method, a radical unit character recognizing method, and an apparatus thereof that are accurately divided.

[0005]

In order to achieve the above-mentioned object, the present invention is a method of extracting radicals of Chinese characters, which measures the time during pen-up when one character is input, and A radical extraction method is characterized in that a stroke group up to a stroke with a long up time and a stroke group after that are extracted as two radicals.

In order to achieve the above object, the present invention is a method for extracting radicals of Chinese characters, in which the stagnant time of a pen is measured when one character is input, and the stroke is the longest. In the radical extraction method, the stroke group and the subsequent stroke groups are extracted as two radicals.

Further, in order to achieve the above object, the present invention provides
A method of extracting radicals of kanji, measuring pen pressure when one character is input, and using a stroke group up to the stroke with the highest pen pressure and a stroke group after that as two radicals. The radical extraction method is characterized in that extraction is performed.

Further, the present invention has the following objects to attain the above objects.
When one character is input, the time for pen-up is measured, and the stroke group up to the stroke with the longest pen-up time and the stroke group after that are extracted as two radicals. A radical-based character recognition method is characterized in that neck recognition is performed so that a combination of two radicals is recognized as one character.

Further, in order to achieve the above object, the present invention provides
When one character is input, the pen input speed is measured, and the stroke group up to the stroke with the longest pen stagnation time and the stroke groups after that are extracted as two radicals. There is a radical unit character recognition method characterized by recognizing and recognizing as a single character from a combination of two radicals.

Further, in order to achieve the above object, the present invention provides
When writing one character, the pen pressure is measured, and the stroke group up to the stroke with the highest writing pressure and the stroke groups after that are extracted as two radicals. Then, a radical unit character recognition method is characterized in that a character is recognized as a single character from a combination of two radicals.

Further, in order to achieve the above object, the present invention provides:
A stroke information input means for inputting a stroke group representing a character, a pen-up time measuring means for measuring a time during pen-up, and a stroke group for one input character is obtained from information obtained by the pen-up time measuring means. A radical extracting device comprising a radical extracting means for extracting a stroke group up to a stroke having the longest pen-up time and a stroke group subsequent thereto as two radicals.

Further, in order to achieve the above object, the present invention provides
The stroke information input means for inputting a stroke group representing a character, the pen stagnation time measuring means for measuring the stagnation time of the pen, and the stroke group for one character that has been input are most determined from the information obtained by the pen stagnation time measuring means. A radical extracting device comprising: radical extracting means for extracting a stroke group up to a stroke with a long pen stagnation time and a stroke group subsequent thereto as two radicals.

Further, in order to achieve the above object, the present invention provides
The stroke information input means for inputting a stroke group representing a character, the writing pressure measuring means for measuring the writing pressure of the pen, and the stroke group for one input character are the most writing pressure from the information obtained by the writing pressure measuring means. A radical extraction device comprising: radical extraction means for extracting a stroke group up to a high stroke and a stroke group subsequent thereto as two radicals.

Further, in order to achieve the above object, the present invention provides:
The stroke information input means for inputting a stroke group representing a character, the pen-up time measuring means for measuring the time at the time of pen-up, and the stroke group for one character that has been input are most obtained from the information obtained by the pen-up time measuring means. A radical extraction unit that extracts a stroke group up to a stroke with a long pen-up time and a stroke group subsequent thereto as two radicals, a radical unit dictionary in which radical information is stored, and a radical extraction unit. A radical recognition means for performing radical recognition for each radical by referring to the radical unit dictionary for the two stroke groups obtained in step 1, and a character configuration in which information of characters corresponding to radical combinations is stored. Table means and character recognition means for outputting a character recognition result as one character by referring to the character configuration table means from a combination of two radicals recognized by the radical recognition means. Lying in radical unit character recognition apparatus according to claim comprising.

Further, in order to achieve the above object, the present invention provides:
The stroke information input means for inputting a stroke group representing a character, the pen stagnation time measuring means for measuring stagnation time of the pen, and the stroke group for one character inputted are the most pens based on the information obtained by the pen stagnation time means. A radical extraction unit that extracts a stroke group up to a stroke with a long stagnation time and a stroke group after that as two radicals, a radical unit dictionary in which radical information is stored, and a radical extraction unit. A radical structure recognition unit for performing radical recognition for each radical by referring to the radical unit dictionary for the obtained two stroke groups, and a character configuration table storing character information corresponding to radical combinations. Means and character recognition means for outputting the character recognition result as one character by referring to the character configuration table means from the combination of two radicals recognized by the radical recognition means. The radical unit character recognition apparatus according to claim certain.

Further, in order to achieve the above object, the present invention provides
The stroke information input means for inputting a stroke group representing a character, the writing pressure measurement means for measuring the writing pressure of the pen, and the stroke group for one character that has been input are the most writing pressure based on the information obtained by the writing pressure measurement means. A radical extraction means for extracting a stroke group up to a high stroke and a subsequent stroke group as two radicals, a radical unit dictionary in which radical unit information is stored, and a radical extraction means. A radical recognition means for performing radical recognition for each radical by referring to the radical unit dictionary for the two stroke groups, and a character configuration table means for storing information of characters corresponding to radical combinations. A radical unit statement comprising character recognition means for outputting a character recognition result as one character by referring to the character configuration table means from a combination of two radicals recognized by the radical recognition means. In the recognition device. According to the above configuration, one character can be accurately divided into radical units, and the character recognition rate is also improved.

[0017]

BEST MODE FOR CARRYING OUT THE INVENTION The outline of the present invention is as follows. That is, according to the present invention, a stroke information input means for inputting a stroke group representing a character, a pen-up time measuring means for measuring a time at the time of pen-up, and a stroke group for a stroke of one character that has been input. From the information obtained in step 1, the radical extraction means is provided, which comprises radical extraction means for extracting the stroke group up to the stroke with the longest pen-up time and the subsequent stroke groups as two radicals. Characterize. Further, according to the present invention, a stroke information input means for inputting a stroke group representing a character, a pen stagnation time measuring means for measuring a stagnation time of a pen, and an input stroke group for one character are obtained by the pen stagnation time measuring means. It is characterized by comprising radical extraction means for extracting a stroke group up to a stroke having the longest stagnant time of the pen and a stroke group after that as two radicals from the stored information. Further, according to the present invention, stroke information input means for inputting a stroke group representing a character, writing pressure measuring means for measuring a writing pressure of a pen, and information obtained by the writing pressure measuring means for a stroke group for one character that has been input. To a stroke group having the highest writing pressure and stroke groups subsequent to the stroke group are extracted as two radicals. With such a configuration, one character can be accurately divided into radical units, and the character recognition rate is also improved.

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a main configuration of a character recognition device according to the present embodiment, in which 1 is a coordinate input device, 2 is a control device which is a device body, 3 is a display device, and 4 is a display device.
Is an external storage device.

The coordinate input device 1 comprises, for example, a transparent tablet 1a and a stylus pen 1b as shown in the figure, and is laminated and integrated on a display device 3 such as a liquid crystal display (LCD) to be integrated with the control device 2. It The transparent tablet 1a and the liquid crystal display have the same coordinate plane, so that the coordinate position on the liquid crystal display can be directly designated by the transparent tablet 1a, and the display information on the liquid crystal display can be visually recognized through the transparent tablet 1a. Is configured.

As shown in FIG. 2, the control device 2 includes a stroke information input section 21, a pen-up time measuring section 22, a pen stagnation time measuring section 23, a writing pressure measuring section 24, a character detection cutting section 25, and a radical. It is composed of an extraction unit 26, a radical recognition unit 27, a character recognition unit 28, a coordinate data buffer (not shown), a pen-up time data buffer, a pen stagnation time data buffer, a writing pressure data buffer, and a component radical data buffer. .

The display device 3 is a liquid crystal display or the like as described above. The external storage device 4 is a radical unit dictionary 41.
And a character storage table 42 (see FIG. 2). The external storage device 4 may be integrated with the control device 2 as shown in the figure.

By the way, the stroke information input section 21 is provided.
Stores the coordinate data series (coordinate value sequence) designated and input through the coordinate input device 1 in the coordinate data buffer 2a together with the pen input information.

The pen time measuring unit 22 measures the pen-up time from the on / off information of the pen switch and stores the pen-up time information between strokes of one character sentence in the pen-up time data buffer 2b.

The pen stagnation time measuring unit 23 measures the time during which the pen stays in the ON state without moving and is stagnated, and stores the pen stagnation time information in the pen stagnation time data buffer. When there are a plurality of points that are stagnant in one stroke, the longest stagnant time or the average stagnant time is stored in the pen stagnant time data buffer.

The writing pressure measuring unit 24 measures the writing pressure of the pen for each stroke and for each sampling point in the stroke, and stores the maximum value in the writing pressure data buffer. Or the average at the sampling point of 1 stroke,
It is stored in the writing pressure data buffer.

The character detection / cutout unit 25 determines whether or not one character has been input. Specifically, when it takes more time than the set value since the last input, the character detection and cutout method by the time to make the stroke group up to that time one character, and the current input There is a character detection / cutout method using a frame in which the stroke group until that time is one character when moving from the character input frame to the next character input frame, but any method is also possible. When the character detection / cutout unit 25 determines that one character has been input, the process proceeds to the radical extraction unit 26.

The radical extraction unit 26 extracts, from the information stored in the pen-up time data buffer, the stroke group up to the stroke with the longest pen-up time and the stroke groups after that as two radicals. The data is stored in the component radical data buffer 2e. Alternatively, from the information stored in the pen stagnation time data buffer 2c, the stroke group up to the stroke with the longest pen stagnation time and the stroke groups after that are extracted as two radicals and stored in the component radical data buffer. To do. Alternatively, from the information stored in the writing pressure data buffer, the stroke group up to the stroke with the highest writing pressure and the stroke groups after that are extracted as two radicals and stored in the component radical data buffer.

The radical recognition unit 27 refers to the radical unit dictionary 41, recognizes the radical unit, and stores the result in the radical recognition data buffer. There are various recognition processing methods, such as a method based on the shape of an input stroke (for example, a vertical line or a horizontal line) and a combination thereof, but there is no particular limitation.

The character recognition unit 28 uses the character configuration table 42.
With reference to, the character formed by the radical stored in the radical recognition data buffer 2f is output as the character recognition result.

The radical unit dictionary 41 is not in character units, but in
The information for each radical forming a character is stored, and the radical stored in the constituent radical data buffer is recognized by referring to the radical unit dictionary 41. Although the data structure differs depending on the recognition method, when the data is based on the shape of the input stroke and its combination, the data structure is as shown in FIG.

As shown in FIG. 4, the character structure table 42 stores a combination of radicals and one character in association with each other. The character structure table 42 is referred to and stored in the radical recognition data buffer. The final recognition result as a character is output from the combination of radicals.

Next, the operation of each of these processing units will be described with reference to the flow chart shown in FIG. The stroke information input unit 21 stores the coordinate data string obtained from the coordinate input device 1 in the coordinate data buffer together with the pen input information (step 1). Pen-up time measuring unit 22
Measures the pen-up time from the on / off information of the pen switch (step 2a) and stores the pen-up information between strokes in the pen-up time data buffer (step 3a). 1 by the character detection cutout unit 25
When it is determined that the stroke input for the characters has been completed, the process moves to the radical extraction unit 26 (step 4a).

The pen stagnant time measuring unit 23 measures the time during which the pen remains stagnant in the on state (step 2a),
The stagnation time information at the end of stroke input is stored in the pen stagnation time data buffer (step 3a). When the character detection / cutout unit 25 determines that the stroke input for one character is completed, the process proceeds to the radical extraction unit 26 (step 4a).

The writing pressure measuring unit 24 measures the writing pressure of the pen at each sampling point in the stroke (step 2).
C) The pen pressure information at the sampling point at the end of pen input is stored in the pen pressure data buffer 2d (step 3).
C). When the character detection cutout unit 25 determines that the stroke input for one character is completed, the radical extraction unit 26
(Step 4C).

The radical extraction unit 26 extracts, from the information stored in the pen-up time data buffer 2b, the stroke group up to the stroke with the longest pen-up time and the stroke groups after that as two radicals. Then, the data is stored in the component radical data buffer (step 4a). Alternatively, from the information stored in the pen stagnation time data buffer, the stroke group up to the stroke with the longest pen stagnation time and the stroke groups after that are extracted as two radicals (step 4a), and component radical data is extracted. Store in buffer (step 5a). Alternatively, the stroke group up to the stroke with the highest writing pressure, the stroke group after that and the stroke group after that are extracted as two radicals from the information stored in the writing pressure data buffer (step 4C). , Is stored in the component radical data buffer (step 5C).

The radical recognition unit 27 refers to the radical unit dictionary 41 to recognize the radical unit (step 6) and stores the radical recognition result in the radical recognition data buffer (step 6). 7).

The character recognition unit 28 uses the character configuration table 42.
With reference to, the character formed by the radical stored in the radical recognition data buffer is output as the recognition result (step 8).

6 to 8 illustrate the above-described processing for each operation of the radical extracting unit 26 by using a concrete example. First, in FIG. 6, when the character “mura” is input, the pen-up time between each stroke is measured while the input data is stored in the coordinate data buffer, and is stored in the pen-up time data buffer 2b. . When the stroke for one character is input, pay attention to the stroke between the 4th stroke and the 5th stroke in which the pen-up time is the longest in the pen-up time data buffer, and then enter one radical. Consider it to be. It is divided into a stroke group up to the stroke having the longest pen-up time, that is, a stroke group of the first stroke to the fourth stroke, and a stroke group after that, that is, a stroke group of the fifth stroke to the seventh stroke. One stroke group is stored in the component radical data buffer using a separator or the like so that the two can be distinguished.

The stroke groups from the first stroke to the fourth stroke stored in the constituent radical data buffer 2e are found to be "kihen" as a result of radical recognition, and are stored in the constituent radical data buffer. It can be seen that the stroke group of the fifth stroke to the seventh stroke is "dimension". The radical recognition results are stored in the radical recognition data buffer.

This radical recognition processing is performed based on the stroke shape and its combination, as described in the explanation of the radical recognition unit 27. The character recognition unit 28 refers to the character configuration table 42 and outputs a “village” composed of “tree” and “dimension” stored in the radical recognition data buffer. This "village" is the character recognition result.

Next, in FIG. 7, when the character "mura" is input, the pen stagnation time at the time of inputting each stroke is measured while storing the input data in the coordinate data buffer. Stored in the time data buffer 2c. When the stroke for one character has been input, the pen stagnation time was the longest in the pen stagnation time data buffer. 4
Focusing on the stroke, it is considered that the input of one radical has been completed. The stroke group up to the stroke with the longest pen stagnation time, that is, the stroke group of the first stroke to the fourth stroke, and the stroke group after that, that is, the stroke stroke of the fifth stroke to 7 strokes
The strokes are divided into stroke groups and the two stroke groups are stored in the component radical data buffer using a separator or the like so that the stroke groups can be distinguished from each other.

As a result of radical recognition, the stroke groups of the first to fourth strokes stored in the component radical data buffer are found to be "kihen" and are stored in the component radical data buffer. It can be seen that the stroke group from the seventh stroke to the seventh stroke is "dimension".

The radical recognition results are stored in the radical recognition data buffer. This radical recognition processing is performed based on the stroke shape and its combination, as described in the explanation of the radical recognition unit 27.

The character recognition unit 28 uses the character configuration table 42.
With reference to, a “village” composed of “tree” and “dimension” stored in the radical recognition data buffer is output. This "village" is the character recognition result.

Finally, in FIG. 8, when the character "mura" is input, the writing pressure for each stroke is measured while storing the input data in the coordinate data buffer,
Store in the pen pressure data buffer. Here, the highest writing pressure point at the sampling points of the strokes respectively input is taken as the writing pressure of that stroke. When the stroke for one character has been input, the fourth stroke with the highest writing pressure in the writing pressure data buffer is focused on, and it is considered that the input for one radical has been completed. The strokes up to the stroke with the highest writing pressure, that is, the stroke group of the first stroke to the fourth stroke, and the stroke group after that, that is, the stroke group of the fifth stroke to the seventh stroke, are divided into two strokes. The group is stored in the component radical data buffer using a separator or the like so that the two can be distinguished.

The stroke groups of the first stroke to the fourth stroke stored in the component radical data buffer are found to be "kihen" as a result of radical recognition, and are stored in the component radical data buffer. It can be seen that the stroke group from the seventh stroke to the seventh stroke is "dimension".

The radical recognition results are stored in the radical recognition data buffer. This radical recognition processing is performed based on the stroke shape and its combination, as described in the explanation of the radical recognition unit 27.

The character recognition unit 28 uses the character configuration table 42.
With reference to, a “village” composed of “tree” and “dimension” stored in the radical recognition data buffer is output. This "village" is the character recognition result.

The present invention is not limited to the above embodiment. For example, the radical recognition result is not limited to one, and a plurality of candidates may be prepared, and the character recognition unit 27 may output the result for each combination. In short, the present invention can be variously modified and used without departing from the scope of the invention.

[0050]

As described in detail above, according to the present invention,
It is possible to accurately divide one character into radical units, that is, to accurately extract radicals, and as a result, the radical recognition system is improved and the character recognition system is also improved. It is a thing.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an apparatus according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a main part in the same embodiment.

FIG. 3 is a view showing a data structure of a radical unit dictionary 41 according to the same embodiment.

FIG. 4 is a view showing a data structure of a character structure table 42 according to the same embodiment.

FIG. 5 is a flowchart explaining the operation of the embodiment.

FIG. 6 is a diagram described using a specific example according to the embodiment.

FIG. 7 is a diagram related to the embodiment and described using a specific example in a state different from that of FIG. 6;

FIG. 8 is a diagram related to the embodiment and described using a specific example in a state different from that of FIG. 7;

[Explanation of symbols]

1 ... Coordinate input device, 2 ... Control device, 3 ... Display device, 4 ...
External storage device, 21 ... Stroke information input unit, 22 ... Pen-up time measuring unit, 23 ... Pen stagnant time measuring unit, 24
... pen pressure measuring unit, 25 ... character detection cutout unit, 26 ... radical extraction unit, 27 ... radical recognition unit, 28 ... character recognition unit, 41 ...
Radical unit dictionary, 42 ... Character configuration table.

Claims (12)

[Claims] 1. A method of extracting radicals of Chinese characters, wherein when one character is input, the time during pen-up is measured, and a stroke group up to the stroke with the longest pen-up time and thereafter. And a group of strokes are extracted as two radicals. 2. A method of extracting radicals of Chinese characters, in which when a single character is input, the pen stagnation time is measured, and stroke groups up to the stroke with the longest pen stagnation time and subsequent strokes are measured. A radical extraction method, characterized in that a group and two radicals are extracted. 3. A method for extracting radicals of Chinese characters, wherein the pen pressure is measured when one character is input, and the stroke group up to the stroke with the highest pen pressure and the stroke groups after that are measured. Is extracted as two radicals. 4. When one character is input, the time during pen-up is measured, and the stroke group up to the stroke with the longest pen-up time and the stroke group after that are extracted as two radicals. Then, after this, radical recognition is performed in radical units, and a character is recognized as a single character from a combination of two radicals. 5. When one character is input, the pen input speed is measured, and a stroke group up to the stroke with the longest pen stagnation time and a stroke group after that are extracted as two radicals. After that, radical recognition is performed in radical units, and a character is recognized as a unit of radical character by recognizing a combination of two radicals as one character. 6. When one character is input, the pen pressure is measured, and a stroke group up to the stroke with the highest writing pressure and a stroke group after that are extracted as two radicals. A radical-based character recognition method characterized by performing radical recognition in radical units and recognizing a single character from a combination of two radicals. 7. A stroke information input means for inputting a stroke group representing a character, a pen-up time measuring means for measuring a time during pen-up, and a stroke group for one character that has been input is the pen-up time measuring means. Radical extraction means for extracting a stroke group up to the stroke with the longest pen-up time from the information obtained in step 1 and a stroke group after that, as radical radical extraction means. apparatus. 8. A stroke information input means for inputting a stroke group representing a character, a pen stagnation time measuring means for measuring a stagnation time of a pen, and a stroke group for one character input by the pen stagnation time measuring means. A radical extracting device comprising: radical extracting means for extracting a stroke group up to a stroke having the longest pen stagnation time from the obtained information and a stroke group subsequent thereto as two radicals. . 9. A stroke information inputting means for inputting a stroke group representing a character, a writing pressure measuring means for measuring a writing pressure of a pen, and a stroke group for one input character can be obtained by the writing pressure measuring means. A radical extracting device comprising: radical extracting means for extracting, as two radicals, a stroke group from information to a stroke having the highest writing pressure and a stroke group after that. 10. A stroke information input means for inputting a stroke group representing a character, a pen-up time measuring means for measuring a time during pen-up, and a stroke group for one character input by the pen-up time measuring means. A radical extraction means for extracting a stroke group up to the stroke with the longest pen-up time from the obtained information and a subsequent stroke group as two radicals, and a radical unit in which radical information is stored. A dictionary, a radical recognition unit that performs radical recognition for each radical by referring to the radical unit dictionary for the two stroke groups obtained by the radical extraction unit, and character information corresponding to the radical combination Is stored and the combination of the two radicals recognized by the radical recognition unit is referred to the character configuration table unit to obtain the character recognition result as one character. A character recognition device for each radical, comprising a character recognition means for outputting. 11. A stroke information input means for inputting a stroke group representing a character, a pen stagnation time measuring means for measuring a stagnation time of a pen, and a stroke group for one character input by the pen stagnation time means. A radical extraction unit that extracts a stroke group up to the stroke having the longest pen stagnation time and a stroke group after that as two radicals, and a radical unit dictionary in which radical information is stored. The two groups of strokes obtained by the radical extraction means are referred to the radical unit dictionary to perform radical recognition on a radical-by- radical basis, respectively, and character information corresponding to the radical combination is stored. A character that outputs a character recognition result as one character by referring to the character configuration table means from the combination of the stored character configuration table means and the two radicals recognized by the radical recognition means. A character recognition device for each radical, comprising a recognition means. 12. A stroke information input means for inputting a stroke group representing a character, a writing pressure measuring means for measuring a writing pressure of a pen, and information obtained by the writing pressure measuring means for a stroke group for one input character. To a stroke group having the highest writing pressure, and stroke groups after that, as radical extraction means, a radical unit dictionary in which information on a radical unit is stored, Radius recognition means for performing radical recognition for each radical by referring to the radical unit dictionary for the two stroke groups obtained by the radical extraction means, and character information corresponding to radical combinations are stored. And a character recognition means for outputting the character recognition result as one character by referring to the character composition table means from the combination of the two radicals recognized by the radical recognition means. Character recognition device for each radical.