JPH10240941A - Input data recognizing device and its method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the reliability and the safety of personal identification. SOLUTION: To after finishing the pen input of handwritten data (S1), this handwritten data and pen characteristic data are inputted to an additional information obtaining means. Then pen characteristic data is fetched from an additional information obtaining means (S2) and when this pen characteristic data coincides with pen characteristic data previously stored in a storage means, the shape feature value of handwriting data is extracted (S3→S4). Then, when this shape feature value and a shape feature value stored in the storing means in advance are highly similar to each other, the equality of the inputting person and the registered person is judged to execute personal identification. On the other hand, when at least one answer of the step 3 and a step 5 is negative (No), the inputting person and the registered person judged to be different from each other and personal identification is not executed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an input data recognizing device and an input data recognizing method, and more particularly to an input data recognizing device and an input data recognizing method for performing personal authentication of signature data input by handwriting with a pen.

[0002]

2. Description of the Related Art Conventionally, when personal authentication is performed by an input data recognition device that is input by handwriting with a pen, a stroke of input data or a shape feature amount of a plurality of stroke groups (for example, a stroke input start point and a stroke input end point, (A curved point, a minimum point, a maximum point, etc.) are extracted in advance and registered in the storage unit, and the shape characteristic amount of a stroke (or stroke group) input when personal authentication is required and registered in the storage unit. The similar writer is determined to be the same writer if the similarity between the two is high.

[0003]

However, in the above-described conventional input data recognition method, since personal authentication is performed only with the above-mentioned shape feature, an input person who inputs a pen to the input device is input by some means. If the signature data of a third party different from the third party is obtained and handwritten with the same stroke as the signature data of the third party, the input person and the third party may be determined to be the same writer. There is a problem that personal authentication lacks reliability and security. In particular, an input device that inputs handwritten signature data and an authentication device that performs personal authentication are connected to a communication network (network).
In the case where the input device and the authentication device are physically separated, such as when the connection is performed via a network, there is a possibility that stroke data for personal authentication is stolen and used by copying in the middle of the network, In such a case, since the input person and the third party are determined to be the same writer, there has been a problem that the reliability and security of personal authentication are lacking.

[0004] The present invention has been made in view of the above problems, and has as its object to provide an input data recognizing apparatus and an input data recognizing method with improved reliability and security of personal authentication.

[0005]

According to a first aspect of the present invention, there is provided an input means for inputting handwritten data, and an extraction means for extracting a shape feature of the handwritten data input by the input means. And an information adding unit for adding additional information to the handwritten data input by the input unit, and the additional information added by the information adding unit and the shape extracted by the extracting unit. Storage means for storing a feature quantity, wherein when handwritten data is input to the input means, the shape feature quantity and additional information of the input handwritten data and the shape feature quantity previously stored in the storage means It is characterized in that it has a judging means for judging whether or not the additional information has the sameness.

According to a seventh aspect of the present invention, there is provided a method for recognizing input data, comprising: an input step of inputting handwritten data; and an extracting step of extracting a shape feature of the input handwritten data. The additional information and the shape / shape feature amount are stored in advance, and when handwritten data is input, the shape feature amount and the additional information of the input handwritten data and the pre-stored information are added. It is characterized in that it is determined whether or not the input feature data and the additional information have the sameness to recognize the sameness of the input data.

[0007]

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

FIG. 1 is a block diagram showing an embodiment (first embodiment) of an input data recognition device according to the present invention. The input data recognition device is an input device for inputting position coordinate data. Coordinate input means 1 composed of a pen, information adding means 2 for adding additional information such as pen-specific data for specifying the input pen and input time, and additional information acquisition for extracting the additional information (including position coordinate data). Means 3,
Feature extraction means 4 for extracting a shape feature amount for personal authentication based on handwriting data (stroke data) from the coordinate input means 1 outputted from the additional information acquisition means 3
A storage unit 5 for preliminarily storing the shape feature amount extracted by the feature extraction unit 4 and the additional information; and the additional information and the feature extraction unit 4 extracted by the additional information acquisition unit 3. The extracted shape feature amount is obtained, and the additional information and the shape feature amount are compared with the additional information and the shape feature amount stored in the storage unit 5 in advance. When the similarity of the shape feature amount is determined to be high, a determination unit 6 for determining the same writer is provided.

In this embodiment, the input pen constituting the coordinate input means 1 is constituted by an ultrasonic digitizer. Since the ultrasonic frequency digitizer has different parameters for calculating the drive frequency and coordinate values according to the input pen, the input pen is specified based on the difference between the drive frequency and the parameter. Therefore, the position coordinate data from the coordinate input unit 1 and the pen-specific data from the information adding unit 2 are input to the additional information acquisition unit 3 as a set. The extraction of the shape feature amount and the calculation of the similarity of the shape feature amount by the feature extraction means 4 described above are performed by a well-known method used in online handwritten character recognition.

FIG. 2 is a flowchart showing a first embodiment of the input data recognition method according to the present invention.

First, in step S1, it is determined whether or not handwriting pen input by the coordinate input means 1 has been completed. The handwritten data input by the pen is sequentially stored in the information adding unit 2, and when the data input is completed, the process proceeds to step S2. In step S2, the input data stored in the information adding unit 2 is transferred to the additional information obtaining unit 3 together with the pen-specific data as additional information.
To extract the pen-specific data, and transfer the pen-specific data to the determination means 6.

Next, in step S3, the pen-specific data sent from the additional information acquisition means 3 is compared with the pen-specific data stored in the storage means 5 in advance. And
If they match, the process proceeds to step S4, in which the handwriting data is transferred to the feature extracting means 4, the feature extracting means 4 extracts the shape feature amount of the handwriting data, and the result is transferred to the determining means 6. Then, in the subsequent step S5, the feature extracting means 4
Is compared with the shape feature amount stored in the storage means 5 in advance, and when the similarity is high, it is determined that they match, and the writer is the same writer in step S6. And end the process.

On the other hand, if at least one of step S3 and step S5 is negative (No), it is determined that the input person and the registered person are different, and the flow advances to step S7 to authenticate with the same writer. The processing is terminated as not to be performed.

As described above, in the first embodiment, not only the shape characteristic amount of the shape but also the pen-specific data of the input pen as additional information is added to the criterion for personal authentication. Safety can be improved.

FIG. 3 is a block diagram showing a second embodiment of the input data recognizing device according to the present invention. In the second embodiment, a coordinate input means 1 and an information adding means 2 are provided. The input device 7 is connected to a host computer 8 including an additional information acquisition unit 3, a feature extraction unit 4, a storage unit 5, and a determination unit 6 via a predetermined communication network (network) 9 such as a LAN. Additional information such as position coordinate data, pen-specific data and input time, and the result of determination by the determination means 6 are transmitted and received on the network 9.

FIG. 4 is a flowchart showing a second embodiment of the input data recognition method according to the present invention.

In step S11, it is determined whether or not there has been an input from the coordinate input means 1. If an input has been made, the input time is indicated as output information from the coordinate input means 1 in addition to the position coordinate data and the pen-specific data. The time data is added, the position coordinate data, the pen-specific data, and the time data are stored in the information adding means 2, and the process returns to step S11.
When the answer to step S11 is negative (No), it is determined whether the input has been completed (step S1).
3) If the input has not been completed, the process returns to step S11. On the other hand, if the input has been completed, the input data stored in the information adding means 2 is transferred to the additional information acquiring means 3. Then, the additional information acquiring means 3 extracts the time data T1 at the start of input and the time data T2 at the end of input from the additional information (step S14), and further extracts all the input time data t (step S15). ), The current time tn is obtained in step S16, and thereafter, the formula (1) is obtained.
It is determined whether or not (2) is satisfied.

Tn−t <γ (1) T1 <t <T2 (2) Here, γ is a predetermined value in consideration of a time delay and a processing time for network transfer.

If all of the handwriting data satisfy the above formulas (1) and (2), the flow advances to step S18 to extract pen-specific data. After that, as in the first embodiment, the comparison between the input data of the pen-specific data and the storage data stored in the storage unit 5 is performed.
The shape feature of the input data is compared with the stored data stored in the storage unit 5 (step S19 → step S2).
0 → Step S21) When both are determined to match, it is determined that the writer is the same writer, personal authentication is performed (Step S22), and the process ends.

On the other hand, if at least one of steps S17, S19 and S21 is negative (No), it is determined that the input person and the registered person are different, and the process proceeds to step S23. The process ends without performing personal authentication without being determined to be a writer.

As described above, in the second embodiment, the input device 7 for inputting handwriting and the host computer 8 (authentication device) for performing personal authentication are installed at locations physically separated via the network 9. Even in this case, since personal authentication is performed using real-time information, even if a third party tries to copy and use handwriting data for personal authentication on the network, those copy data will be invalidated be able to. Further, similarly to the first embodiment, even when the same shape feature as the personal authentication handwriting data is input, the pen unique data of the input pen matches the pen unique data stored in the storage unit 5 in advance. If it is determined not to do so, it is determined that the input person and the registrant are different, so that personal authentication is not performed in such a case, and the reliability and security of the personal authentication can be improved.

In the first and second embodiments, an ultrasonic digitizer is used as the coordinate input means 1.
The input pen used by the input person is identified by the driving frequency or the like of the digitizer. However, a unique identifier (ID) is assigned to each pen and the ID is stored in the storage unit 5 in advance, and the ID is stored in the storage unit 5 in advance. May be used as additional information, for example, a unique value may be given to the input pen to use the unique value as additional information.

Further, the position coordinate data and the additional information output from the input device 7 (coordinate input means 1, information adding means 2) are encrypted and stored in the storage means 5 of the recognition device 8,
When handwritten data is input, the encrypted data stored in the storage unit 5 is decrypted by the additional information acquisition unit 3, the feature extraction unit 4, and the determination unit 6, and then stored in the storage unit 5 with the input data and the like. The network 9 is determined by comparing the stored contents with the stored contents to determine whether or not the authentication is possible.
The above-mentioned sign data can be more securely prevented from being stolen, and the reliability and security of authentication can be further improved.

FIG. 5 is a flowchart showing a third embodiment of the input data recognizing method according to the present invention. In the third embodiment, position coordinate data is added as information added by the information adding means 2. Use the input time. That is, the storage means 5 stores the required input time S0 (= input time of input end point−input time of input start point) of a stroke constituting handwriting data together with the shape feature amount extracted by the feature extracting means 4 in advance. ing.

In step S31, it is determined whether or not an input has been made from the coordinate input means 1. If an input has been made, time data indicating the input time is added to the position coordinate data as output information from the coordinate input means 1. Then, the position coordinate data and the time data are stored in the information adding means 2, and the process proceeds to step S
Return to 31. If the answer to step S31 is negative (N
When it becomes o), it is determined whether or not the input has been completed (step S33). When the input has not been completed, the process returns to step S31. When the input has been completed, the same as in the second embodiment. Then, the time data T1 at the start of the input and the time data T2 at the end of the input are extracted (step S34), and all the input time data t are extracted (step S35). After that, it is determined whether the above-described formulas (1) and (2) hold (step S37).

If all of the handwriting data satisfy the above formulas (1) and (2), step S38 is executed.
Then, the input time S1 of the input start point of the stroke of the handwriting data and the input time S2 of the input end point of the handwriting data are extracted by the additional information obtaining means 3, and these data are determined
Transfer to Next, the required time S0 of the stroke previously stored in the storage means 5 is extracted, and it is determined whether or not Expression (3) is satisfied (step S39).

| (S2−S1) −S0 | <δ (3) where δ is a predetermined value in consideration of an error or variation in input time that may occur at the time of input.

When the expression (3) is satisfied for all the handwriting data, the shape feature amount of the input data and the shape stored in the storage means 5 are stored in the same manner as in the first and second embodiments. Comparison with the feature amount is performed (step S
40 → step S41), and when the similarity between the two is high, it is determined that they are the same writer and personal authentication is performed (step S22).
The process ends.

On the other hand, if at least one of step S37, step S39, or step S41 is negative (No), it is determined that the input person and the registered person are different, and the flow advances to step S43 to execute the same handwriting. The process ends without authenticating with the user.

As described above, in the third embodiment, even when a digitizer (input pen) that cannot generate pen-specific data is used as the coordinate input means 1, position coordinate data is input. By calculating the stroke writing speed using the input time, the same effect as in the first and second embodiments can be obtained, and the reliability and security of personal authentication can be improved. Becomes

[0031]

As described above, according to the input data recognizing apparatus and the input data recognizing method of the present invention, not only the shape characteristic amount but also the pen specific data unique to the input pen, the input time data, and the handwriting. By adding additional information such as the time required for inputting data to the criteria for personal authentication, the reliability and security of personal authentication can be improved. Further, by encrypting the data obtained by the coordinate input means and the information adding means and the data stored in the storage means and decrypting the data by the feature extracting means and the judging means, the security can be further improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of an input data recognition device according to the present invention.

FIG. 2 is a flowchart showing a first embodiment of a method for recognizing input data according to the present invention.

FIG. 3 is a block diagram showing a second embodiment of the input data recognition device according to the present invention.

FIG. 4 is a flowchart showing a second embodiment of the input data recognition method according to the present invention.

FIG. 5 is a flowchart showing a third embodiment of the input data recognition method according to the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Coordinate input means (input means) 2 Information addition means 4 Feature extraction means (extraction means) 5 Storage means 6 Judgment means

Claims (11)

[Claims] 1. An input data recognizing device comprising: input means for inputting handwritten data; and extracting means for extracting a shape feature of the handwritten data input by the input means. An information adding unit for adding additional information to the data; and a storage unit for storing the additional information added by the information adding unit and the shape feature quantity extracted by the extracting unit. Handwritten data is input to the input unit. A determination unit that determines whether the shape feature amount and the additional information of the input handwritten data and the shape feature amount and the additional information previously stored in the storage unit have the sameness. An input data recognizing device, comprising: 2. An input device including the input unit and the information adding unit, and a recognition device including at least the extraction unit, the storage unit, and the determination unit are connected via a predetermined communication network. 2. The input data recognition device according to claim 1, wherein: 3. The input data recognition device according to claim 1, wherein the additional information is identification information of the input unit. 4. The input data recognition device according to claim 1, wherein the additional information is an input time at which the handwritten data was input. 5. The input data recognition device according to claim 1, wherein the additional information is a time required for inputting handwritten data. 6. An encryption unit for encrypting the shape feature and the additional information, and a decryption unit for decrypting the encrypted information, wherein the storage unit is configured to store the encrypted information. The input data recognizing device according to any one of claims 1 to 5, wherein the input data recognizing device stores the encrypted data. 7. A method for recognizing input data, comprising: an inputting step of inputting handwritten data; and an extracting step of extracting a shape characteristic amount of the input handwritten data. The additional information and the shape / shape feature amount are stored in advance, and when handwritten data is input, the shape feature amount and the additional information of the input handwritten data and the previously stored shape feature amount and additional information 2. A method of recognizing input data, comprising determining whether or not the input data has the same identity and recognizing the identity of the input data. 8. The apparatus according to claim 7, wherein the additional information is identification information of an input pen for inputting handwritten data.
How to recognize the input data described. 9. The method according to claim 7, wherein the additional information is an input time when the handwritten data is input. 10. The input data recognition method according to claim 7, wherein the additional information is a time required for inputting handwritten data. 11. An encryption step means for encrypting the shape feature quantity and the additional information, and a decryption step for decrypting the encrypted information, wherein the encryption is performed in the encryption step. The method for recognizing input data according to any one of claims 7 to 10, wherein the stored information is stored in advance.