JPH09185677A - Stroke start and end point decision method, and recognition dictionary registration method and device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To surely recognize a character even when a horizontal bar of the first stroke of HIRAGANA (cursive form of Japanese syllabary) character 'TA', for example, is drawn from the right to the left by deciding the end points having the short and long distances from a point set previously as a start point and an end point respectively. SOLUTION: A point is previously set, and the end points having the short and long distances from the set point are decided as a start point and an end point respectively. A stroke information input part 21 of this device inputs the groups of strokes which form the characters. A distance calculation part 23 calculates the distance between a stroke end point and the set point. An end point decision part 24 decides the end points having the short and long distances from the set point as the start and end points respectively among those end points which are calculated at the part 23. An end point comparison part 25 compares the end points which are decided as the start and end points respectively at the part 24 with the end points which are arranged in the order of input. A coordinate value string operation part 26 reverses the time series of all coordinate value of strokes when no coincidence of comparison is confirmed at the part 25.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention judges the start point / end point of a stroke by the distance from a set set point, rather than the input order of the type of the end point of a stroke group handwritten and input through a coordinate input device. The starting point of the stroke for character recognition based on the stroke group obtained by this
End point determination method, character recognition method, recognition dictionary registration method,
And its device.

[0002]

2. Description of the Related Art Conventionally, for example, a handwriting input device in which a coordinate input device and a display device are laminated and integrated has been capable of inputting characters and figures on paper. Moreover, since it is possible to directly instruct a place to input, it has been recently used in various fields.

[0003]

However, the conventional handwritten character recognition technology is not sufficient, and the demand for efficient and accurate handwritten character recognition is increasing. For example, if the horizontal bar of the first stroke of the character "ta" is written from right to left, in the character recognition method that compares the coordinate points of the dictionary and the coordinate points of the input data by distance calculation, the starting point in the character recognition dictionary is Since the end point and the end point are the right end points, the evaluation value was bad and as a result, it was not recognized.

Therefore, the present invention has been made in consideration of the above-mentioned circumstances. A preset point is set in advance, and an end point having a short distance from the preset point is set as a start point and a far end point is set as an end point. Thus, for example, even when the horizontal bar of the first stroke of "ta" is written from right to left, the stroke start point / end point determination method, the character recognition method, and the recognition dictionary registration that ensure the character recognition are possible. It is an object to provide a method and an apparatus thereof.

[0005]

In order to achieve the above object, the present invention provides a stroke information inputting means for inputting a stroke group forming a character and a distance for measuring a distance between a stroke end point and a set set point. A measuring means, an end point determining means whose end point is a point close to a set point among the end points measured by the distance measuring means and an end point is a far point, and an end point and an input order which are regarded as a start point and an end point by the end point determining means. The end point comparing means for comparing the end points according to the above, and the end point comparing means reverses the time series of all coordinate values in the stroke when the end point of the single point determining means and the start point / end point of the end points in the input order are different. The coordinate value sequence operating means, the recognition dictionary storing character coordinate information, etc., the coordinate value sequence obtained by the coordinate value sequence operating means, and the coordinate value sequence stored in the recognition dictionary are associated with each other. Perform the distance calculation is to include a character recognition section for character recognition result having the highest evaluation value characters.

In order to achieve the above object, the present invention comprises a stroke information input means for inputting a stroke group forming a character, a character code input means for inputting a character code, and a set point set as an end point of the stroke. A distance measuring means for measuring the distance, an end point determining means for starting a point closer to the set point and an end point for a far point among the end points measured by the distance measuring means, and the end point determining means for determining the start point and the end point. End point comparing means for comparing the end point according to the input order with the end point thus determined, and the end point of the end point determining means in the single point comparing means and the start point / end point of the end point according to the input order are different from each other in the coordinate value of the stroke. The coordinate value sequence operating means for reversing the time series, the stroke information input means, and the coordinate information of the stroke group for one character obtained by the coordinate value sequence operating means are used to identify the character. In that it comprises a recognition dictionary registering unit for registering in the recognition dictionary with a character code obtained by the character code input means as a point. According to the above configuration, correct character recognition is possible even if the stroke input direction is reversed.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The outline of the present invention is as follows. (1) When a stroke group that constitutes a character is input, the start point and end point of the stroke are not in the order of input, but the end point close to the set set point is the start point, and the far end point is the end point. Method.

(2) When a stroke group forming a character is input, the start point and the end point of the stroke are not in the order of input, but the end point close to the set set point is the end point and the far end point is the start point of the stroke.・ How to determine the end point.

(3) When a stroke group forming a character is input, the start point and the end point of the stroke are not in the order of input, but the end point close to the set set point is set as the start point, and the far end point is set as the end point, and then recognized. A character recognition method that calculates the distance from the coordinate values of the characters stored in the dictionary and uses the character with the highest evaluation value as the character recognition result.

(4) When a stroke group forming a character is input, the start point and the end point of the stroke are not in the order of input, but the end point close to the set set point is set as the end point, and the far end point is set as the start point, and then recognized. A character recognition method that calculates the distance from the coordinate values of the characters stored in the dictionary and uses the character with the highest evaluation value as the character recognition result.

(5) A recognition dictionary registration method in which the end points of a stroke group forming a character are registered in the recognition dictionary as feature points after the end points near the set point are set as the start points and the far end points are set as the end points, not in the order of input. .

(6) A recognition dictionary registration method for registering the end points of the stroke group forming a character as the feature points in the recognition dictionary after the end points near the set point are set as the end points and the far end points are set as the start point, not in the order of input. .

(7) Stroke information input means for inputting a stroke group forming a character, distance measuring means for measuring a distance from a set end point of a stroke, and end points measured by the distance measuring means. Of these, an end point determining means having a point close to the set point as a start point and a far point as an end point, an end point comparing means for comparing the end points regarded as the start point and the end point by the end point determining means with the input order, and the end point comparing means. A device for determining a start point / end point of a stroke, comprising coordinate value sequence operating means for reversing the time series of all coordinate values in a stroke when the start point / end point of the end point in the input sequence are different from the end point of the end point determining means. .

(8) Stroke information input means for inputting a stroke group forming a character, distance measuring means for measuring a distance from a set point set as an end point of a stroke, and end points measured by the distance measuring means. An end point determining means of which a point close to the set point is an end point and a far point is a start point, an end point comparing means for comparing the end points regarded as the start point and the end point by the end point determining means and the end points in the input order, and the end point comparing means. A device for determining a start point / end point of a stroke, comprising coordinate value sequence operating means for reversing the time series of all coordinate values in a stroke when the start point / end point of the end point in the input sequence are different from the end point of the end point determining means. .

(9) Stroke information input means for inputting a stroke group forming a character, distance measuring means for measuring a distance from a set point set as an end point of a stroke, and end points measured by the distance measuring means. Of these, an end point determining means having a point close to the set point as a start point and a far point as an end point, an end point comparing means for comparing the end points regarded as the start point and the end point by the end point determining means with the input order, and the end point comparing means. The coordinate value sequence operating means for reversing the time series of all coordinate values in the stroke when the end point of the end point determining means and the start point / end point of the end point in the input order are different, and the coordinate information of the characters are stored. The recognition dictionary, the coordinate value sequence obtained by the coordinate value sequence operation means, and the coordinate value sequence stored in the recognition dictionary are associated and distance calculation is performed, and the character with the highest evaluation value is used as the character. Character recognition apparatus comprising a character recognition unit to identify the result.

(10) Stroke information input means for inputting a stroke group forming a character, distance measuring means for measuring the distance from the end point of the stroke and the set set point,
Of the end points measured by the distance measuring means, an end point determining means having a point close to a set point as an end point and a far point as a starting point, and an end point regarded as a start point and an end point by the end point determining means and an end point according to an input order are compared. End point comparing means and a coordinate value sequence operating means for reversing the time series of all coordinate values in the stroke when the start point and the end point of the end point in the input order are different from the end point of the end point determining means in the end point comparing means. And a recognition dictionary in which coordinate information of characters and the like are stored, a coordinate value sequence obtained by the coordinate value sequence operation means and a coordinate value sequence stored in the recognition dictionary are associated, and distance calculation is performed. A character recognition device comprising a character recognition unit that makes a character having a high evaluation value a character recognition result.

(11) Stroke information input means for inputting a stroke group forming a character, character code input means for inputting a character code, and distance measuring means for measuring a distance from a stroke end point and a set set point. , An end point determination means of a point close to a set point among the end points measured by the distance measurement means as a start point and a far point as an end point; An end point comparison means for comparing and a coordinate value sequence operation for reversing the time series of all coordinate values in a stroke when the start point and the end point of the end point in the input order are different from the end point of the end point determination means in the end point comparison means. Means, the stroke information inputting means, and the coordinate information of the stroke group for one character obtained by the coordinate value sequence operating means as the character feature point of the character code input. Recognition dictionary registration apparatus including a recognition dictionary registration unit for registering in the recognition dictionary with a character code obtained by the stage.

(12) Stroke information input means for inputting a stroke group forming a character, character code input means for inputting a character code, and distance measuring means for measuring a distance from a stroke end point and a set set point. , An end point determination means of which the end point is a point close to a set point among the end points measured by the distance measurement means and a far point is a start point, and the end points which are regarded as the start point and the end point by the end point determination means and the end points in the input order are An end point comparison means for comparing and a coordinate value sequence operation for reversing the time series of all coordinate values in a stroke when the start point and the end point of the end point in the input order are different from the end point of the end point determination means in the end point comparison means. Means, the stroke information inputting means, and the coordinate information of the stroke group for one character obtained by the coordinate value sequence operating means as the character feature point of the character code input. Recognition dictionary registration apparatus including a recognition dictionary registration unit for registering in the recognition dictionary with a character code obtained by the stage.

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a main part of an apparatus according to the embodiment of the present invention, and FIG. 2 is a block diagram showing an overall configuration of the apparatus according to the embodiment.

First, the overall configuration of the apparatus according to the embodiment of the present invention will be described with reference to FIG. In FIG. 2, 1 is a coordinate input device, 2 is a control device which is the main body of the device, 3 is a display device, and 4 is an external storage device.

The coordinate input device 1 is composed of, for example, a transparent tablet 1a and a stylus pen 1b, is laminated and integrated on a display device 3 such as a liquid crystal display, and is integrated with the control device. The transparent tablet 1a and the liquid crystal display have the same coordinate plane, and 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 is transparent tablet 1a.
It is configured to be visible through a.

As shown in FIG. 1, the control device 2 includes a stroke information input unit 21, a distance measuring unit 22, and an end point determining unit 2.
3, an end point comparison unit 24, a coordinate value sequence operation unit 25, a character recognition unit 26, an input coordinate data buffer 2a, a distance data buffer 2c, an end point data buffer 2d, a coordinate data buffer 2e, and the like.

The display device 3 is composed of a display such as a liquid crystal display device as described above. The external storage device 4 is composed of a disk or the like for storing the recognition dictionary 41. The external storage device 4 may be integrated with the control device 2 as shown in FIG.

The stroke information input unit 21 stores the coordinate data series (coordinate value sequence) and the stroke shape, which are instructed and input via the coordinate input device 1, in the input coordinate data buffer 2a together with the pen input information.

The character code input unit 22 stores the character code designated and input through the coordinate input device 1 in the character code buffer 2b. The distance measuring unit 23 measures the distance between the end point of the stroke stored in the input coordinate data buffer 2a and the set set point, and stores the measurement result in the distance data buffer 2c.

The end point deciding unit 24 uses the distance data buffer 2
Of the distances between the two end points and the set point stored in c, the single point with the short distance is the start point and the end point with the long distance is the end point. The result is stored in the single point data buffer 2d together with the coordinates. The endpoint comparison unit 24 compares the coordinates of the single point stored in the input coordinate data buffer 2a with the coordinates of the endpoint data buffer 2d.

If the end point comparing section 24 compares the coordinate value strings with the same coordinate value, the coordinate value string operating section 26 directly receives the information such as the coordinate value string stored in the input coordinate data buffer 2a. 2e. As a result of comparison, when the coordinate values are not the same, the time series of the coordinate value sequence of the stroke is reversed and stored in the coordinate data buffer 2e.

The character recognizing unit 27 is provided in the coordinate data buffer 2
The distance between the coordinate value string stored in e and the coordinate value string of the character stored in the recognition dictionary 41 is calculated, and the pattern with the highest evaluation value is output as the character recognition result.

The recognition dictionary registration unit 28 registers the coordinate value sequence stored in the coordinate data buffer 2e in the recognition dictionary as character feature points. The recognition dictionary 41 stores information such as a coordinate value sequence for each character. Although the data structure differs depending on the character recognition method, when the data is formed based on the shape of the input stroke and its combination, the data structure is as shown in FIG.

Next, the operation of each of these processing units will be described with reference to the flow chart of the processing shown in FIG. The stroke information input unit 21 stores the coordinate data string and the stroke shape obtained from the coordinate input device 1 together with the pen input information in the coordinate data buffer 2a for each stroke (step 1).

The distance calculation unit 23 measures the distance between the end point of the stroke stored in the input coordinate data buffer 2a and the set point (step 2), and stores the measurement result in the distance data buffer 2c (step). 3).

The end point deciding unit 24 uses the distance data buffer 2
Of the distances between the two end points and the set point stored in c, the end point having the shorter distance is set as the start point and the end point having the longer distance is set as the end point (step 4). The result is stored in the end point data buffer 2d together with the coordinates (step 5).

The end point comparison section 25 compares the coordinates of the end point stored in the input coordinate data buffer 2a with the coordinates stored in the end point data buffer 2d (step 6).

When the end point comparing section 25 compares the same coordinate value, the coordinate value sequence operating section 26 uses the coordinate value sequence information stored in the input coordinate data buffer 2a as it is. 2e (step 7)
-1). As a result of comparison, if the coordinates are not the same, the time series of the stroke coordinate value sequence is reversed and stored in the coordinate data buffer 2e (step 7-2).

The above process is repeated until the stroke of one character is input (step 8). In the case of character recognition, the character recognition unit 27 causes the coordinate value string of the stroke for one character stored in the coordinate data buffer 2e and the recognition dictionary 41.
The distance of the coordinate value string of the character stored in is calculated, and the pattern with the highest evaluation value is output as the character recognition result (step 9).

When a character is registered in the recognition dictionary, the character code is input in the character code input section 22 (step 1
0), the character code is stored in the character code buffer 2b (step 11).

Further, the recognition dictionary registration unit 28 recognizes the stroke coordinate value sequence and the stroke shape for one character stored in the coordinate data buffer 2e as the character feature amount together with the character code in the character code data buffer 2b. (Step 12).

The above processing will be described using a specific example. In FIG. 5, the set point is set at the upper left of the character, the coordinates of the set point are (0, 0), and the horizontal bar of the first stroke is changed from right to left in the process of inputting "ta". Take the case where it has been written in. Further, there are three characteristic points for one stroke, an end point and an intermediate point.

When the distance between the end point of the stroke of the first stroke and the set point is calculated, the set point (0,0) and the end point A (14,4) are calculated.
Is the square root of 212. The distance between the set point (0,0) and the end point a (2,6) is the square root of 40. Since the end point a is closer to the set point, the end point a is set as the start point and the end point a is set as the end point. Therefore, the coordinate information stored in the coordinate data buffer 2e for the first stroke is (2, 6) (8, 5).
The order is (14, 4).

The above steps are repeated for the number of strokes, and character recognition is performed when all strokes are completed. Although FIG. 3 shows the data structure of "ta" registered in the recognition dictionary 41, the strokes of these dictionaries are associated with the input strokes one by one, and the distance is calculated. As a result of the distance calculation, the pattern having the highest evaluation value is “ta”, and “ta” is output as the character recognition result. If you do not manipulate the coordinate values, the distance will increase,
It does not come up to the candidate.

The present invention is not limited to the above embodiment. For example, the character recognition method may be any method as long as it is a method of calculating the distance between the input coordinate value sequence and the coordinate value sequence of the recognition dictionary. The distance measuring unit 23, the end point determining unit 24, the end point comparing unit 25, and the coordinate value sequence operating unit 26 may be executed after the user finishes inputting one character. Further, the character code input when registering characters in the recognition dictionary may be executed at any time before the recognition dictionary registration unit 28 is executed. In short, the present invention can be variously modified and used without departing from the gist thereof.

[0042]

As described above in detail, depending on a distance from a preset point, an end point having a short distance is set as a start point and a point having a long distance is set as an end point.
Correct character recognition is possible even if the horizontal bar on the screen is written from right to left.

[Brief description of the drawings]

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

FIG. 2 is a block diagram showing the overall configuration of the device according to the embodiment.

FIG. 3 is a diagram showing a data structure of a recognition dictionary 41 according to the same embodiment.

FIG. 4 is a flowchart showing a flow of processing according to the same embodiment.

FIG. 5 is a diagram showing a specific example according to the embodiment.

[Explanation of symbols]

1 ... Coordinate input device, 2 ... Control device, 3 ... Display device, 4 ...
External storage device, 21 ... Stroke information input unit, 22 ... Character code input unit, 23 ... Distance measuring unit, 24 ... End point determination unit, 25 ... End point comparison unit, 26 ... Coordinate value sequence operation unit, 27 ...
Character recognition unit, 28 ... Recognition dictionary registration unit, 2a ... Input coordinate data buffer, 2b ... Character code buffer, 2c ... Distance data buffer, 2d ... End point data buffer, 2e ... Coordinate data buffer, 41 ... Recognition dictionary.

Claims (10)

[Claims] 1. In a character recognition device, when a stroke group forming a character is input, the start point and the end point of the stroke are not in the order of input but the end point close to the set set point is set as the start point or the end point. A method for determining a starting point / ending point of a stroke, characterized in that a far end point is set as an ending point or a starting point. 2. In a character recognition device, when a stroke group forming a character is input, the start point and end point of the stroke are not in the order of input, but the end point close to the set set point is the start point or end point, and the far end point. Is an end point or a starting point, and thereafter, a distance with respect to the coordinate value of the character stored in the recognition dictionary is calculated, and the character having the highest evaluation value is used as the character recognition result. 3. In a character recognition device, the end points of a stroke group forming a character are not in the order of input, but the end points close to the set set point are used as the start point or the end point, and the far end points are used as the end point or the start point. A recognition dictionary registration method characterized by being registered as a feature point in a dictionary. 4. A stroke information inputting means for inputting a stroke group forming a character, a distance measuring means for measuring a distance from a set point set as an end point of the stroke, and an end point measured by the distance measuring means. An end point determining means having a point close to the set point as a start point and a far point as an end point, an end point comparing means for comparing the end points regarded as the start point and the end point by the end point determining means and the end points according to the input order, and the end point comparing means. The stroke of the stroke is characterized by comprising coordinate value sequence operating means for reversing the time series of all coordinate values in the stroke when the start point and the end point of the end point in the input order are different from the end point of the end point determining means. Start / end point determination device. 5. A stroke information input means for inputting a stroke group forming a character, a distance measuring means for measuring a distance from a set point set as an end point of a stroke, and an end point measured by the distance measuring means. An end point determination means having a point close to the set point as an end point and a far point as a start point, an end point comparison means for comparing the end points regarded as the start point and the end point by the end point determination means and the end points according to the input order, and the end point comparison means. The stroke of the stroke is characterized by comprising coordinate value sequence operating means for reversing the time series of all coordinate values in the stroke when the start point and the end point of the end point in the input order are different from the end point of the end point determining means. Start / end point determination device. 6. A stroke information input means for inputting a stroke group forming a character, a distance measuring means for measuring a distance from a set point set as an end point of a stroke, and an end point measured by the distance measuring means. An end point determination means having a point close to the set point as an end point and a far point as a start point, an end point comparison means for comparing the end points regarded as the start point and the end point by the end point determination means and the end points according to the input order, and the end point comparison means. The stroke of the stroke is characterized by comprising coordinate value sequence operating means for reversing the time series of all coordinate values in the stroke when the start point and the end point of the end point in the input order are different from the end point of the end point determining means. Start / end point determination device. 7. A stroke information input means for inputting a stroke group forming a character, a distance measuring means for measuring a distance from a set point set as an end point of the stroke, and an end point measured by the distance measuring means. An end point determining means having a point close to the set point as a start point and a far point as an end point, an end point comparing means for comparing the end points regarded as the start point and the end point by the end point determining means and the end points according to the input order, and the end point comparing means. A coordinate value sequence operating means for reversing the time series of all coordinate values in a stroke when the end point of the end point determining means and the start point / end point of the end point in the input order are different, and character coordinate information and the like are stored. The recognition dictionary that is present, the coordinate value sequence obtained by the coordinate value sequence operation means, and the coordinate value sequence stored in the recognition dictionary are associated and distance calculation is performed, and the character with the highest evaluation value is recognized. A character recognition device, comprising: a character recognition unit for obtaining a recognition result. 8. A stroke information input means for inputting a stroke group forming a character, a distance measuring means for measuring a distance between a stroke end point and a set set point, and an end point measured by the distance measuring means. An end point determination means having a point close to the set point as an end point and a far point as a start point, an end point comparison means for comparing the end points regarded as the start point and the end point by the end point determination means and the end points according to the input order, and the end point comparison means. A coordinate value sequence operating means for reversing the time series of all coordinate values in a stroke when the end point of the end point determining means and the start point / end point of the end point in the input order are different, and character coordinate information and the like are stored. The recognition dictionary that is present, the coordinate value sequence obtained by the coordinate value sequence operation means, and the coordinate value sequence stored in the recognition dictionary are associated and distance calculation is performed, and the character with the highest evaluation value is the character. Character recognition apparatus characterized by comprising a character recognition unit to identify the result. 9. Stroke information input means for inputting a stroke group forming a character, character code input means for inputting a character code, distance measuring means for measuring a distance from a stroke end point and a set set point, Among the end points measured by the distance measuring means, an end point determining means having a point close to a set point as a start point and a far point as an end point, and an end point regarded as a start point and an end point by the end point determining means and an end point according to an input order are compared. End point comparing means and a coordinate value sequence operating means for reversing the time series of all coordinate values in the stroke when the start point and end point of the end point in the input order are different from the end point of the end point determining means in the end point comparing means. And the character code inputting means using the coordinate information of the stroke group for one character obtained by the stroke information inputting means and the coordinate value sequence operating means as a character feature point. A recognition dictionary registration device, comprising: a recognition dictionary registration unit that registers in the recognition dictionary together with the character code obtained by. 10. Stroke information input means for inputting a stroke group forming a character, character code input means for inputting a character code, and distance measuring means for measuring a distance from a stroke end point and a set set point. Of the end points measured by the distance measuring means, an end point determining means whose end point is a point close to a set point and a far point is a starting point, and an end point which is regarded as a start point and an end point by the end point determining means and an end point in the input sequence are compared. End point comparing means and a coordinate value sequence operating means for reversing the time series of all coordinate values in the stroke when the start point and end point of the end point in the input order are different from the end point of the end point determining means in the end point comparing means. And using the coordinate information of the stroke group for one character obtained by the stroke information inputting means and the coordinate value sequence operating means as the character feature point of the character code inputting operation. A recognition dictionary registration device comprising: a recognition dictionary registration unit that registers in a recognition dictionary together with the character code obtained by the step.