JPH02143317A - Signature input method - Google Patents

Abstract

PURPOSE:To exactly decide the completion of signature by always supervising the floating time of a pen from pen pressure data, and comparing the floating time with prescribed reference floating time, and deciding it to be the completion of the input of the signature when the floating time exceeds the reference floating time, and fetching the pen pressure and coordinate data up to a final contact point as signature data. CONSTITUTION:The signature is signed on a signature board 10 by the input of a pen 20, and the pen pressure data and the coordinate data are taken in as the data at that time. These data are taken in continuously whether during a period in which the pen is in contact with the signature board 10 or during a floating period in which it is separated form the signature board 10, and are stored in a general memory. Then, when the floating time of the pen exceeds the prescribed reference floating time determined beforehand, it is decided that the input of the signature is already completed, and the pen pressure data and the coordinate data up to the final contact point before that are taken in as the signature data. Thus, only the correct signature data can be received electrically regardless of signing time.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a signature input method, and particularly to an improved signature input method that can correctly determine the completion of signature input.

[Prior Art] A person's own signature has long been used as a means of verifying the identity of the person, and the signature is used as highly reliable proof for contracts, approval documents, receipts, or confirmations of financial transactions.

In recent years, such signatures have also been used to verify the identity of payment cards at various banks and retail stores.

Normally, such signatures are visually checked to eliminate false signatures, but it has been proposed to read them as electrical signals and use them to determine authenticity.

Conventionally, as a signature input device, for example, JP-A-62-
No. 287387 discloses a system for online recognition of a wide range of handwritten characters, including signatures, and also proposes performing character recognition from both coordinate data and pen pressure data of handwritten characters.

However, signature recognition differs from normal character recognition in that it requires an extremely strict determination of authenticity, so it is not only necessary to recognize each individual character, but also the overall flow of the brush when signing. It is preferable to import the data as a group of data.

Normally, in Western European countries where signatures are popular, most signatures are written in a form close to a single stroke, and it is possible to recognize this as a single character in a - group, but in Japan, where signatures are written separately and independently, In order to see the stroke movements during the series of signatures described above, it is preferable to also capture the stroke movements during the blank period between the letters as data.During the blank period, the pen pressure data is approximately However, only the coordinate data is taken in as necessary data as much as possible.

[Problem to be solved by the invention] However, when importing data during the blank period between characters, especially in Japanese signatures, blanks between characters occur relatively frequently and a non-negligible period of time occurs. As a result, it is not possible to distinguish between this blank period and data generated after the signature is completed, which generates similar data, and as a result, determining the completion of signature input has become a major problem.

In other words, in Japanese signatures, there is usually a blank period between characters, so it is possible to determine the start of the signature based on the occurrence of pen pressure data, but conversely, when the pen pressure data disappears, it is possible to determine whether the blank period or the signature There is a problem that it is not possible to judge whether the process is complete or not.

Therefore, in the past, a method was considered in which the data acquisition time was fixed in advance and the pen pressure data and coordinate data were acquired during this time, but this resulted in a large waste of data acquisition time and also In many cases, the movement of the pen after completion is captured as data, and such data that is essentially unrelated to the signature generates a large amount of noise, which poses a problem of lowering the recognition rate.

Furthermore, fixing the data capture time or restricting the signature area forces the signer to make an unnatural signature, and there is a problem in that good data cannot be captured because the signature that is familiar to people cannot be performed. .

The present invention has been made in view of the above-mentioned conventional problems, and its purpose is to provide an improved signature input method that can accurately determine whether a signature has been completed.

[Means for Solving the Problem] In order to achieve the above object, the present invention constantly monitors the floating time of the pen from pen pressure data, compares it with a predetermined standard floating time, and calculates the actual pen floating time. The present invention is characterized in that it is determined that the signature input is completed when the floating time exceeds this reference value, and furthermore, according to the present invention, the data acquired up to the last touch point is excluded from the data acquired before determining the completion of the signature input. , is characterized by reducing noise mixed in for verification by removing subsequent data.

[Operation] Therefore, according to the present invention, three signatures are performed on the signature board by pen input, and the pen pressure data and coordinate data are captured as data at that time. These data are continuously captured and stored in general memory while the pen is in contact with the signature board and during floating periods away from the signature board.

When the floating time of the pen exceeds a predetermined reference floating time, it is determined that the signature input has already been completed, and the pen pressure and coordinate data up to the last contact point before that are converted into signature data. It becomes possible to electrically accept only correct signature data regardless of the signature data.

Further, according to the present invention, the input of signature data is automatically limited to the correct data being signed, so there is no restriction due to time or the framework of the board to be signed.
The signer can input a natural signature similar to a normal signature.

[Embodiments] Hereinafter, preferred embodiments of the present invention will be described based on the drawings.

FIG. 3 shows the relationship between an eight-person board to which the present invention is applied and an input pen. can be taken in electrically.

The signature board 10 is generally known as a digitizer used for CAD input, etc., and particularly in recent years, various forms have been put into practical use as tablets and others for inputting various computer graphics.

Typical common methods of such digitizers include electromagnetic induction, capacitive coupling, magnetostrictive, resistance, optical/photoelectric, and acoustic methods.

In particular, in recent years, even if the pen is not in complete contact with the signature board 10, even if the board surface and the pen are separated to a certain extent, the electromagnetic circuit formed on the top surface of the board 10 or the like allows the pen to reach a certain height. It is possible to capture the movement of the pen 20 as data almost in the same way as a contact.

In FIG. 3, this valid data acquisition area is indicated by a broken line, and it is possible to convert the movement of the pen 20 within the area as shown in the illustrated area into an electrical signal with respect to the coordinates x, y, force direction, and height h direction. can.

As described above, according to the signature board 10 and stylus pen 20 of this embodiment, it is effective even when the signature input is performed with the pen 210 in contact with the board 10 or when the pen 20 is slightly lifted from the board 10. If it is within the area, an electrical signal can be captured as an input of the contact state, and thereby, even when the pen is in the air, it is possible to capture the movement of the pen as coordinate data.

FIG. 4 shows an example of the signature "Japanese fat", and the solid line indicates the document that remains on the paper, and according to FIG. 3, the stylus pen 20 is in complete contact with the signature board 10. Shows the handwriting of the time.

The broken line indicates the stylus pen 20 within the effective area 100.
Indicates the coordinate data when pen 2 passes.
0 is not in contact with the signature board 10, so the pen pressure data is "0", but the coordinate data is captured as shown by the broken line. Therefore, according to such a signature board 10, since the handwriting of a genuine signer follows an almost constant trajectory even when the pen is floating in the air, such invisible movements of the pen can also be registered as a signature. This is extremely useful as it can be used as a criterion for determining authenticity.

In recent years, the magnetostrictive plate method has been suitable as a digitizer that can read the coordinate data of such spatial trajectories.The principle is that a cut is made in a part of the magnetostrictive diaphragm, and the phase is reversed at each cut. When an excitation coil is wound so that the input cable is added, the magnetostrictive vibration waves are transmitted as if they were propagating through a magnetostrictive wire stretched parallel to the magnetostrictive plate.

Therefore, the coordinate values can be determined by determining the time required to detect this signal with the detection coil.

This magnetostrictive vibration wave can be stably generated several 11 meters above the digitizer surface of the board, thereby making it possible to obtain the hollow dynamic region 100.

As mentioned above, the data to be detected as a signature is from the beginning of the signature shown in FIG. is formed.

However, the data obtained by adding the valid area has movement other than the signature shown by broken lines 210 and 211 before and after it, and it is necessary to remove this.

In particular, it is known that the movement 211 after completion shows an extremely unspecified pattern, and removal of this data is extremely important in signature recognition.

However, as mentioned above, there are gaps unique to the Japanese language between each character shown in FIG. It is difficult to accurately determine the difference.

The present invention is characterized in that the movement of the pen in the floating state is determined separately for when signing and when not.
For this purpose, a comparison is made with the basic floating time.

FIG. 2 shows an example of the writing pressure data of the signature shown in FIG. 4 described above, and in the present invention, this writing pressure data is used to determine the floating time of the pen 20.

The horizontal axis in FIG. 2 is t, which is the elapsed time, and the vertical axis is the pen pressure P. It is understood that the contact pressure when the stylus pen contacts the signature board is electrically detected from the signature board sequentially at P, and P-0 indicates the floating state of the pen.

The first occurrence of pen pressure P (tl) in FIG. 2 indicates the starting point of the signature on the board, and it is possible to recognize that the signature has started from this tl.

This starting point 1. From then on, the pen pressure data P sequentially records the pen pressure by the signer, and some zero points of pen pressure occur in the middle of writing and after the signature is completed, and the pen pressure characteristics in Fig. 2 are from time t2 to t3.
(Δt1), t4~ts (Δt=), ts~Lt
(Δt3), t8~tw (Δtn) and tlO
~t1. (Δt5) is the pen floating period after the signature starts. Of course, O'=t+(Δto) before the start of the signature is also the floating period of the pen.

As is clear from FIG. 2, the floating times Δtl, Δt2 . It is clear that Δt3+Δt4 is sufficiently shorter than the floating time Δt5 after the signature is completed.

Of course, the floating time during a signature differs depending on the signer, but it is statistically clear that the floating time for genuine signers is extremely short, while the floating time for fake signers is generally long. It is being

Therefore, in the present invention, the floating time during signature for authentic signatures can be suppressed within a certain standard.

FIG. 1 shows a flowchart of a preferred embodiment of the celebrity method according to the present invention.

When the preparation for signature input is completed, coordinate data
(0), y (0) and pen pressure data P(0) are input. Then, it is determined whether the pen is within the effective area 100 based on the coordinate data x, y (step 302), and further, in step 303, the pen pressure data P
It is determined whether or not (0) has occurred. That is, in step 303, the pen moves to the starting point 200 in FIG.
It is determined whether or not it has been reached. Steps 301, 302, and 303 are then repeated until the pen reaches the starting point 200 and the signature begins.

When the pen reaches the start point in step 303, the sampling number n is set to 1 (step 304
), thereafter, coordinate data and pen pressure data are sequentially captured at every predetermined sampling interval Δ.

A characteristic feature of the present invention is that the floating time of the pen is constantly monitored, and in FIG. 1, the pen floating time is t.
, and in step 305 this floating time tp
is initially set to O, and this floating time is divided into numbers in each of the following entry steps.

Furthermore, according to the embodiment, the invalid time t when the pen leaves the valid area 100 is also constantly monitored, and although this invalid time t is extremely short for a genuine signer, it is very short for a fake signer. This is relatively long, which makes it possible to obtain good data when determining the authenticity of the signature.

Then, in step 305, this invalid time t is initialized to zero.

When the above initial settings are completed, coordinate data x (n), y (n) and pen pressure data P are entered in step 306.
(n) is input. Then, in step 307, it is determined whether these inputs are within the valid area.

If the input data is outside the valid area, its invalid time is counted in step 308, and it is determined in step 309 whether this exceeds a predetermined fundamental invalid time T.

Of course, the invalid time is also 2! If the invalidity time T is exceeded, this indicates either that the pen has left the valid area during the signature process or that the pen has been quickly ejected from the valid area by the authentic signer. Since this is extremely difficult in itself, according to this embodiment, the process proceeds to step 314, which will be described later, assuming that the signature has been completed.

In the normal case, the signature is made within the valid area, step 30
If it is determined in step 7 that the area is a valid area, step 31
At 0, the invalid time t is reset to 0 again, and the sampling number n is incremented by 1.

Then, in step 311, it is determined whether or not the immediately preceding data P(n-1) with respect to the captured pen pressure data p(n) is 0.

That is, if the pen pressure data immediately before the captured data is O, the sampling interval Δt is added to the floating time t in step 312, and on the other hand, if the immediately preceding data is not 0, the process returns to step 305. The next data acquisition is repeated.

As described above, when the time when the pen pressure data P is 0 is sequentially added by the sampling interval Δt in step 312, this floating time added value is compared with the reference floating time Tp in step 313, and the floating time is determined as the reference floating time. Until the time is reached, data is repeatedly captured, assuming that the pen is temporarily floating during the signature.

On the other hand, if the floating time t exceeds the reference floating time T, the process proceeds to step 312 and it is determined that the signature input is complete.

The reference floating time T can be set arbitrarily, and the optimum reference value can be set to a different value depending on the actual signer.

As described above, according to the present invention, the stylus pen 2
The floating time t of 0 is constantly monitored, and it is possible to determine the completion of signature input based on whether or not this exceeds the reference floating time T. With an extremely simple configuration, only the valid signature shown in Fig. 4 can be input as data. can do.

According to the present invention, unnecessary data is further removed after it is determined that the input is completed. That is, the coordinate data X captured as described above.

y and pen pressure data P, the data 231 advanced by the standard floating time from the completion 201 in FIG. 4, for example, as indicated by the reference numeral 230 in FIG. 4, is stored as useless data,
Removing this is useful for noise-free and accurate data input.

Therefore, according to the present invention, after it is determined that the signature input is completed, an operation is performed to return the captured input data to the data up to the last touch point, which according to the embodiment is performed in steps 314 and 315 This is done by repeating.

That is, in step 314, the sampling number n
is subtracted by 1, and it is determined in step 315 whether the pen pressure data at that time is 0, and the sampling number n is subtracted until the pen pressure data P becomes O.

Therefore, if we explain FIG.
The sampling number n goes backwards until it becomes .

Although the return data up to the last contact point is stored in the memory, it is not used for signature recognition, and only the data from the start point 200 to the completion point 201 in FIG. 4 is used as signature input data.

Therefore, according to such an embodiment, it is possible to capture only the data from when the stylus pen contacts the signature board to the last contact point as coordinates and writing pressure data.

Although not shown in detail in the figure, coordinate data and writing pressure data are each stored in separate memories, and the addresses determined by the finally determined number n of sunblinks are taken in as data.

[Effects of the Invention] As explained above, according to the present invention, only the data that is actually being signed out of the input famous data is imported, and the accuracy of signature registration and signature recognition can be significantly improved. It becomes possible.

Further, according to the present invention, as described above, only the input data of the correct signature operation is automatically captured, and the data before the start and after the end is automatically removed, so that the signer can However, there is no unnatural restriction on signature time or signature space area, and the signer has the advantage of being able to freely input a signature in a natural atmosphere.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing a preferred embodiment of the signature input method according to the present invention, FIG. 2 is an explanatory diagram showing an example of pen pressure data of the signature input method using the present invention, and FIG. 3 is a flow chart showing a preferred embodiment of the signature input method according to the present invention. FIG. 4 is an explanatory diagram showing an example of a signature. 10...Signature board 20...Stylus pen t,...Floating time T,...Standard floating time Applicant: Kenji Yoshida, patent attorney, Cadex Co., Ltd. [D-4]

Claims (1)

[Scope of Claims] A signature input method in which a pen is input on a signature board and the input signature is converted into a signal from pen pressure data and coordinate data at this time, the floating time of the pen is constantly monitored from the pen pressure data, and a predetermined standard is set. A signature input method characterized by comparing the floating time and determining that the signature input is complete when the time exceeds the floating time, and importing the pen pressure and coordinate data up to the last contact point as signature data.