JPH05505687A - Handwriting device for force and position based information collection - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] Background of the Invention: Handwriting Device for Force and Position Based Information Gathering This invention relates to manual data collection, specifically identifying who is the originator; Identifying what the said creator is doing in the form of information input and/or The present invention relates to a data collection device for determining the neuromuscular condition of a person. This issue writing, especially through a stylus or similar writing device; It relates to XY position and pressure recognition and information input technology.

In the past, two other techniques have been developed that use some form of data entry needle. one The first technique was developed to recognize the XY position of the needle on one plane. somewhat formal Digitizer pads that interact with the needle belong to this category of technology; Two dimensions of the surface sense the position of the needle.

In the other technique, in order to track the movement and/or forces to which the needle is exposed, Technology has developed. This latter technique is primarily used for property verification and recognition. , is not used for data entry, because - if the information collected through the needle is If you want to translate it into information, you need to decipher the character's ground type. This is because data entry becomes relatively inaccurate. To be more specific, on one plane It is confirmed that it is virtually impossible to derive the true position of the needle from only the force information of the needle. It has been certified. Because, inter alia, the interaction between the needle and the writing surface, the writing The orientation of the needle with respect to the surface is important for roll, pitch, and yaw positioning, as well as in-needle positioning. This is because it includes positioning of the sensor. Therefore, the needle is used as a pure data entry device. Its use was somewhat unpopular.

On the other hand, needles are extremely useful for certain categories of property verification.

For example, the 3-axis line developed by one of the co-inventors of this invention along with others An automatic characteristic verification device that uses a force-control pen is an extremely effective tool for characteristic verification. I realized something. [Automatic characteristic verification using 3-axis linear force sensing pen] matic Signature Verification Using a Under the title “Three-Axis Force-5 sensitive Pen”) , 7~　′　6　Pasum゛　KU　Iha'　-　ΔSME-3No.　3 PP, 3 376 9 B (IE Transactions on Systems+Man & Cybernetics, Volume SME-13, No. 3. PP, 329-337 (May/June 1 The technical content of the paper published in 983)) is incorporated herein by reference. is a part of it. This paper describes the triaxial force used in connection with the present invention. Explains the sensing pen. This paper is also extracted from the above-mentioned power anger pen converter. It also describes how to process the emitted signals.

Features of this traditional pen that have long been recognized as useful for property verification One of these is its ability to sense force at the tip of the needle on the writing surface. The above page Although unique to the pen's configuration, this pen has been in development for over 10 years and is technically known. However, the utility of this feature is not recognized in other applications discussed herein. . For example, attempts have been made to use the pen described above for data entry, but The plan to use a pen as described above for data entry was abandoned. Abandoned. You2Pyou Magazine (Computer Magazine) 19 PP published in March 1977.

43-50 is the ``Operator for Manually Printed Characters'' by Crane et al. An 0n-Line Data Entry S system for Hand-Printed Characters) It refers to papers that This paper is incorporated herein by reference and is one of the is part of This paper describes the actual physical properties of the above-mentioned pen-sensing mechanism, and and describes early attempts at using this pen for character recognition. There is. This plan was later abandoned as it was impractical. The above plan was abandoned One of the reasons is that its use gives all users the ability to roll, pitch and yaw. This is because it required a constraint to hold the pen in only one way.

There are other characterization devices known in the art that are based on needle construction. Representative of the above technologies Patent issued to Herbst et al., U.S. Pat. No. 128,829 lists early works by Herbst and others. Her Early work by Bust was based on a single acceleration parameter for dynamic properties. In contrast, later Herbst patents described the axial force through the accelerometer and the axis of the needle. describes the multi-axial acceleration parameters associated with strain gauge measured forces along ing. An important but difficult to compensate feature of this prior invention is the connection between the writing surface and the needle. The accelerometer measures the force of acceleration at a point other than the junction.

Therefore, either the absolute position or the relative position at the intersection of the needle and the above writing surface It is virtually impossible to obtain any information that does. Furthermore, if the above intersection It is true that if acceleration can be measured using However, direct information regarding the relative position of the points on the writing surface may be obtained.

The acceleration measured at the above point is an indication of the true position of the intersection of the needle and the writing surface. Must not be. The vFf of this pen as actually configured is the point of acceleration and Due to the lack of information regarding the nature of the interaction of the needle intersections on the It is impossible to stabilize the fixed position. The interaction of these two positions is or depending on the user.

The xY position table left and its equivalent are known, and the XY position can be sensed by absolute sensing. and the two positions contain the force and are sensed on the surface of the table left. In any case, the forces measured in this way are Derived from the downward force of action. These structures prevent contact between the needle and the writing surface. Do not use the force at the tip for any reason other than sensing. It is considered that

Important references include H. D. Frein and J. Nis-Ostre. “3-axis force sensation” by H. D. Crane and J. S. Ostrew Automatic characteristic verification using a knowledge pen (’AutomaticSignature Verification Llsing a Three-Axis For ce-3ensistivePen"), = $4167w / Δ System system and system μ system Δ, Volume SMC43, No. 3. May / June 1983, PP, 329-337. (I E E E Transa tions on Systems.

Man and Cybernetics, Vol, SMC-13, No. 3. May/June 1983゜pp, 329-337. ) is a paper by . This document describes, among other things, a preferred technique for recognizing characteristics as counterfeits. death However, the technology described there could not sense the position relative to a fixed reference position. Therefore, there is only a limited possibility of character recognition.

The following patents were identified in a search of prior art documents: U.S. Patent No. 3,767 ．． No. 907 Radcliffe, Jr., needle flap A needle is used in conjunction with a film strip to achieve mathematical cross-correlation. Disclose technology.

U.S. Patent No. 3,906,44444 Characteristic Verification System Using a Hand with Four Position Outputs Associated with and Clock Outputs Disclosed.

U.S. Pat. No. 3,915,015 Crane et al. Strain Cage Discloses a writing instrument that uses a writing device and can sense force at the tip.

Discloses circuits potentially useful in connection with the present invention.

U.S. Patent No. 3,930.22929 (frain rane) and others, A set of digital to identify hands and characters with associated 4-position 1 output Describe the circuit in which the signal is generated.

U.S. Patent No. 3,988,934 Kawphoefner et al. In addition, a spatial sensing device for handwriting analysis is described.

U.S. Pat. No. 4,040,01010) and a method and apparatus for identifying handwriting using pressure input.

U.S. Patent No. 4,040,011 Crane et al. and other devices for identifying handwriting using pressure input.

U.S. Patent No. 4,040,01212 and other devices for identifying handwriting using pressure input.

U.S. Patent No. 4,078,226 EerNisse et al., writing Description and writing regarding the needle used for characteristic verification that produces a representative signal of force in response to pressure The surface is subjected to a Z force, which is a force in the upward/downward direction only, and is related to the force pulled from the needle. and is sensed.

U.S. Patent No. 4,086,56767 Yet another apparatus for identifying handwriting using pressure input and pressure input is described.

U.S. Pat. No. 4,122,435 Greenaway, It is written between the position sensor that detects the contact force between the needle and the filling surface.

No. 4,128,829 Herbst et al., supra. It's in the text.

U.S. Patent No. 4,190,82020 describes yet another device for identifying handwriting using pressure input.

U.S. Patent No. 4,234,86868), orthogonal to the writing surface Piezoelectric materials used to collect information related to forces Describes a gender verification device or similar pattern verification device.

U.S. Patent No. 4,240,065 Howbrook, mounted on the needle. The attached coil is used to signal the approach of the needle to the surface. Describes position sensing device.

U.S. Patent No. 4,281.313 to Baldrig Jr. e, Jr.), the structure inside the surface due to the characteristic ml derived from the pressure on the surface. Describes the position sensing device that was created.

U.S. Patent No. 4,475,235 to Graham, for Tablets Yes, rank! Also describes the characteristic verification technology that relies on downward force.

U.S. Patent No. 4,475,240 Brogardh et al. Surface Optical, following precise information derived from markings made by the needle on top Describe the devices configured around the sensor.

U.S. Pat. No. 4,495,644 Parks et al. A variable method in which two resistive films are brought into contact to sense position information in X and Y coordinates. Described as a device for real-time characteristic verification using converter pads. This device is characterized by and character recognition.

UK patent application GB2.039.118A Standard Telephone d Te1ephone) describes a characteristic recognition device based on waveform identification technology verification.

Menlo Park, California a) Niss R.I. International (SRI Inter-na) Research by one of the co-inventors and others in Co-author Crane, along with the above documents and others. Discussed in a patent issued to Frein.

Summary of the invention According to the invention, a point on a surface senses a force along at least one axis within that surface. a stylus with a device for a position sensing pad with a device for sensing the true position along at least one axis of the surface; A device is provided which collects information directly at a single point on a surface, consisting of a combination of -

According to the method of the invention, the occurrence of characteristics or other information such as written cryptographic form simultaneously with the dynamic forces involved in making the code by measuring the force at the tip of the needle. while at the same time sensing the position of the needle on the surface and from this position a fixed reference drawn relative to the position.

This invention provides simultaneous registration of spatial patterns and forces associated with the establishment of written cryptography. Can be done. The above pattern can also be registered in electronic form and then compared with the electronic template. compared.

Generally, there are three categories of information collected using devices according to the invention. information The first category is the identity of the writing program. The second category of information is written materials and the third category of information is the state of the writing program. same One basic process is used for all three categories of information up to the point of analysis of the information. oh In each case, compare the information you enter with the standard and confirm the information you want to obtain. Or a mechanism is needed to add an appropriate distance for identification.

For the characteristic verification device, the dynamics of the characteristic is a dynamic pattern related to its establishment. Confirmed by correlation with

This device is useful in medical applications. Information regarding the force at the tip of the needle above is , created in advance by the same user in order to analyze the state of the user's neuromuscular equipment. compared with the pattern.

Symptoms of conditions like Parkinson's disease are particularly obvious, such as tremors and the subterranean variety. white, which is diagnosed by analysis of electronic registration of spatial and force patterns. Ru. This device also includes photomicrographs and unique hand and arm velocities and accelerations. , also used to study other conditions.

What is particularly useful about recognizing certain forms of handwritten information is that the XY tablet alone is It is understood that additional combinations of registration or force inputs associated with cryptographic opening are Useful for recognizing individual characters. In particular, the downward pointing arrow at the tip of the cursor Analog forces are used in conjunction with character or stroke segmentation. When using the XY tablet, the conventional Can be used to replace and enhance up/down switches .

The invention will be further described by reference to the following specific embodiments in conjunction with the accompanying drawings. It will be well understood.

Brief description of the drawing FIG. 1 shows the present invention using a first embodiment of an input device for use in accordance with the present invention. 1 is a perspective view and a partially schematic view of a first embodiment of the invention; FIG.

FIG. 2 is a perspective view of a second embodiment of an input device suitable for use in accordance with the present invention. Ru.

FIG. 3 is a flowchart of a particular embodiment of a method according to the invention.

Description of specific embodiments Referring to FIG. 1, an information gathering device 10 according to the present invention is shown. This information Collection devices are of two basic anger detector configurations.

It is equipped with an XY tablet or digitizer 12 and a 3-axis needle or pen 14. Ru. The digitizer 12 is connected to a first preprocessing device 16 via a suitable electronic connection. The needle 14 is then connected to a second pretreatment device 18 via a suitable electronic connection. Ru. These preprocessing devices are based on the circuit disclosed in U.S. Pat. No. 3,915.015. It can also be formatted. As an additional input, each of the above pretreatment devices: A timing signal can be received from the timer 20.

The timer 20 measures the time between the outputs of the first preprocessing device 16 and the second preprocessing device 18. At that time, the function will be described later, but the outputs 24, 26 and 2 are respectively is applied to the processing device 30. The output of the processing device 30 is processed by the end user. It is supplied to the output device 32 to be utilized.

Above needle 14. The combination of preprocessing devices 18 and interconnections between them is 19 U.S. Patent No. 28, issued Oct. 28, 1975, to Crane et al. 3,915,015, which is included in this book for reference. It is incorporated into and forms a part of the specification. Other input devices can also be used, but The reference needle is not a position displaced from the needle tip, but a position added to the needle tip. Since it has the advantage of being able to detect force, it is possible to detect force at a certain point displaced from the needle tip 22. Instead, accurate information about the dynamics at the needle tip can be collected. Up The output of the pre-processing device 18 corresponds to the above-mentioned reference patent, the force F in the X direction, the force F in the X direction, a force F7 in the Y direction that is substantially perpendicular to the writing surface, and a pressure F2 that is perpendicular to the writing surface.

(However, it is not critical that the directions of the three forces are exactly orthogonal to each other. , ) The digitizer 12 and related preprocessing device 16 are manufactured by SUMMARY GRAPHICS (Su *magraphics tablet or cross (Talog) tablet Both can also be digitizers that are readily available on the market. Digi Mll in Tyza's selection! Parameters include tablet resolution and sample rate It is. resolution of 100 lines per inch, and preferably 200 lines per inch; and a sampling rate of 100 samples per second is expected to be generally applicable. According to the invention, position information as well as force information is transmitted to the first preprocessor 16 and the second preprocessor. information that is synchronously extracted from both the control equipment 18 and is time-correlated. .

Figure 2 shows another combination of needle and digitizing tablet as shown in Figure 1. FIG. 2 shows a five-dimensional platen 34. As shown in FIG.

The five-dimensional platen 34 includes a one-plane digitizer, which includes a top surface 36 and a one-plane digitizer. and a force-sensing platen on a second plane, the bottom surface 38 including a bottom surface in the second plane. , published in Kamphoefner et al., November 2, 1976. The handwriting sensing and analysis device described in U.S. Pat. No. 3,998,934 It can also be derived from the design. Basically, three converters are provided, F 8 force transducer 40. The FV force transducer 42 and the F2 force transducer 44 are connected to the X direction suspension line 4. 6. attached to a suitable suspension line, such as the Y-direction suspension line 48 and the Z-direction suspension line 50. Ru.

The top surface 36 can also be a membrane with a sensor embedded therein, or The X and Y positions are extracted as outputs to appropriate preprocessing equipment. Above top surface 3 Pressure from a conventional needle or writing instrument on the membrane of 6 is applied to the bottom where the membrane lies below. The purpose is to contact surface 38 over the entire active area of its bottom surface. , the pressure generates both position information and force information.

Each of the embodiments of FIGS. 1 and 2 has particular advantages.

The apparatus of Figure 2 allows the user to use virtually any writing instrument. deer However, other pressures, such as the pressure of the palm of your hand against the writing surface, can cause unwanted artificial distortion ( artifacts). Furthermore, the apparatus of FIG. This creates a more consistent signal.

The above output signal may be in the form of an analog voltage or, depending on the characteristics of the above converter, It should be understood that these are directly correlated digital signals.

FIG. 3 is a flowchart showing the basic processing operation of the processing device 30 of FIG. as above The information obtained can be processed in real time or stored for later analysis.

The first stage analyzes the signal values Fx, Fy, Fz so that all information is time-correlated. , X, Y, and t, where t is time (step A), then the force vector in the Z direction is is tested to determine whether it is greater than a threshold (step B), as described above. The threshold value indicates that contact has occurred between the needle and the tablet, and the parameters What if the power of Z shows that you have to start collecting trick information? If the threshold is not exceeded, the first stage is repeated until the threshold is exceeded. -Tan, Once the threshold is exceeded, the above data is indexed by time using a sampling method. stored in a suitable temporary storage register or storage device (i.e., step C), (alternatively, the information is stored in a cumulative analysis of input data until a termination signal is received. directly into the processing equipment that performs the processing. However, the process described here is difficult to analyze. At least a segment of values that can be ranged is collected before This is based on an assumption. In fact, among the steps remembered by the time index One is based on signals indicating whether the needle is up or touching the tablet. This is to divide the information into segments, and the above-mentioned contact divides the information with strokes. That's what I do. ) Step C is the timeout after "Ben 7 Knob", set in advance is completed when a termination signal is received, such as the last occurrence of F2 below the threshold Ru.

While strokes divide information into segments, each writing stroke is segments are analyzed individually or multiple segments are analyzed as a group. 2 formats Processing is then applied according to the invention. The first type of processing is a five-dimensional feature extraction process. be.

Alternatively, a five-dimensional correlation process is applied. Finally, a simple yes/no decision. A yes/no decision is derived from the applied processing of the analytical form that the Ru. Moreover, both processes can be done in parallel, and if one or the other If the output from each is found not to provide satisfactory information for the intended purpose, the output from each properly mixed.

Parametric processing, and the first processing here called No. 4,040,010, issued August 2, 1977, is an extension of three-dimensional processing, the contents of which are incorporated herein and portions thereof are incorporated herein by reference. I am doing it. According to processing step A, the analog signal is digitized and here " A set of descriptors called feature J is , the extraction process is performed from five signals. Average force in each of the force vectors, needle up and needle down and the same Further details, including timing parameters such as number of types, Then, process A starts from the start point of X (X, step D) and the end point of X (X, 1 step E), the start point of Y (Y, step F) and the end point of Y (YN step G). This information is drawn from previously collected segment information .

For parametric calculations, it is often necessary to establish a bottom line in FX and FA. useful, but the "average" parameters, i.e. throughout the stroke duration , the average force in the X direction is most suitable. The bottom line function g(X, t) is the length of processing time. Through this, an equal to the mean of X is established (step H). Similarly, the average A value is calculated for Fv.

Once we have established the base lines for the values of X and Y, the five axes of force and place 1 are For each, the process described above is an IIFX or The characteristics suggested for both FY and FY are: measured average dimensions, standard Standard deviation.

Minimum, maximum, mean absolute value, mean positive value, number of positive samples, mean negative value, number of negative samples, zero crossing number, maximum minus measured average dimension, maximum value minus minimum value, and measured average dimension map. Minus minimum value. Additionally, the following characteristics can be recorded: total number of hours, segments number, number of up hours and number of down hours, all these characteristics are not necessarily included in the analysis. It should be noted that although not necessary, it has been found useful for property verification.

Prior to any decision-making application according to the invention, it is necessary to create a shape plate. (Step K). The shape plate is created by collecting the features recorded above and this Useful for distinguishing shape plates from specific standards. If the above standard is a characteristic of the user For example, then the standard is identified with the user. What if the standard is normal? If the user's writing sample indicates that the sample is in a normal state, the sample is identified as being in a normal state. if Also, if the above shape plate is used for character recognition, this shape plate can be used to identify specific characters. or need to be identified as a segment of a character or a set of characters.

Following the shapeplate creation step above, a test is performed to determine whether this shapeplate is suitable. This step is based on the accumulation of subjective factors. This is done in non-real time, that is, off-line. shape plate is under control Therefore, this shape plate may be changed by replacement or addition to improve its credibility. You can also

Finally, the step of identifying states, properties, or units of information is step J. (Step M).

For this purpose, by fitting an appropriately weighted Euclidean distance , calculations such as root mean square analysis are performed for each feature of each axis.

Cumulative distance can be calculated by summing the features over all cumulative distances.

For example, determining whether a particular test property is a true property or an attempted falsification. The basic criteria used for Let S be a feature vector representing the test characteristics. The component of S is determined by the method described above. is the value of the determined “best” set of features. L is the computer-memory cedar board or reference base. vector, and T is the standard deviation vector associated with it. t or γ The determination of the registered set of known true properties and the degree of similarity between the test properties and the shape plate and the above shape plate is based on, for example, the weighted Euclidean equation given by determined by the distance between the nodes: d (s, t) = (1/f ((s i-t i)/r i)'')>i=1 Here, f is the number of characteristics, and si is the ith component, that is, the value of the feature in the feature vector S. , ti is the ith component of the shape plate vector, and ri is from the set of registered characteristics This is the standard deviation of 1the total at the time of calculation. The calculated value of d (s, t) becomes smaller. Indeed, between S and t, and therefore the test characteristic represented by S and the calculator - The degree of similarity between the memory cedar board t and the memory cedar board t increases more and more.

The decision rules for determining whether a particular test characteristic satisfies the verification criteria are as follows: It is as follows: If d (s, t)\d thrss, then this property is a true property. It will be judged.

If d (s, t) > d"'61, this property is determined to be an attempted forgery. determined.

The quantity d 0, which is a pre-assigned value chosen to obtain dthr@I is relatively low for fraud underwriting to minimize the rate A banking application involving a relatively small amount, while likely to be set to a small value For dthr@l, business usually minimizes the inconvenience to users. Larger values are appropriate. More information about these tests can be found at H.D. Crane and Jay Nis Ostrem (H, D, Crane and "Automatic characteristic verification using 3-axis linear force sensing pen" by J, S, 0stress) ("Automatic Signature Verification Ll sing a Three-Axis Force-5 intensive, P $ IIfp # I Δ's system and cyberspace entitled en'') Rumu, Volume SMC-13, No. 3. May/June 1983゜PP, 329- 337. (IE Transactions on SystemI ms, Man and Cybernetics, Vat, SMC-13, No, 3. May/June 1983. PP, 329-337,) paper is shown.

If said distance is less than said predetermined threshold or When determined to be equal, the property, information, or condition is determined to be normal or true. It will be done. If the above distance is greater than the above predetermined threshold. For example, a characteristic, information, or condition is determined to be abnormal. The threshold amount is The level of safety should be set to eliminate unacceptable unfair exclusion that would result from a value that is too strict. Pre-assigned thresholds selected using empirically gathered information indicating or the above thresholds, depending on the significance of the particular 'treatment'; Can be a different value for each "handling".

Process B is based on Crane et al., published on August 2, 1977, Using sampled data techniques similar to those described in Patent No. 4.040,012 is a correlation process, which is incorporated by reference into this specification and forms a part thereof. There is.

The first step in this correlation process is to record a sample of the full force waveform being tested. (Step N), each of the above waveforms is an input that has a correlation with time. There is a suitable cedar board of order 0 that represents all 5-dimensional independent components of the data. A test is carried out to determine whether the For example, since a cedar board is required, the waveform pattern of the stroke information (i.e., each axis waveform pattern), or instead of a set of dimensions or features as described above. A set of cedar boards is created consisting of the average of several waveforms drawn from each waveform. . Therefore, several samples of input data are needed to obtain a reliable set of cedar planks. It is understood that this will be tested empirically (Step P).

The above set of cedar planks was tested to determine whether this set of cedar planks is suitable or not. (Step Q), this test verifies that, for example, cedar planks are within the acceptable range. In order to It may also include testing the standard deviation value obtained by passing the cedar plank. I can do it (Step R).

- Once sufficient information has been collected to establish a cedar plank, this cedar plank can be used for property verification. , perhaps because of the character recognition or writing program status, the set of cedar boards is not recorded. compared with the recorded specimen (from step N). Once the correlation step is performed , this means that ``the elastic correlation J is valid for each time sample for each of the five dimensions.'' (Step S). The one-dimensional elastic correlation was cited as a reference above. 0 described in detail in connection with the calculations from U.S. Patent No. 4,040,012. Elastic correlation is a technique that allows the standard mathematical correlation to be applied to the cedar board above. dynamic time warping and/or or scaling applied to each time sample multiplex time. The difference between this statement is , the elastic correlation process described above is applied to five orthogonal dimensions. Faith The characteristics of the signal, i.e. the pressure signal and the spatial information signal, are in fact independent characters.

However, by using a set of five-dimensional shapes, the combination of information can be character creation or verification of the shape and dynamics of alphabetic characters, or allows the user to It is possible to identify the user's condition, such as whether they are suffering from neuromuscular dysfunction. ing.

Some examples of neuromuscular dysfunction that can be detected include Parkinson's disease, or hypnosis. Or there is an impairment of muscle response due to excitability or other reasons of its subterranean nature. the latter The use of the present invention for the testing of Particularly useful.

- Once the elastic correlation step is completed, five independent values are obtained, and each independent value represents the numerical value of each of the five dimensions. The above five values are a single number of strengths In order to get go/no-go) i.e. compared to get a yes/no output (step PuS). Therefore, reducing the extremely complex calculations for one number of merits, the above number becomes is passed to an output device (step T).

A purely objective test captures the kinetic and spatial information generated by the needle on the tablet. It is used for analysis.

It is understood that subjective evaluations are also applicable to the information collected by the device according to the invention. For example, under the first processing above, record for each dimension The features identified are individually examined by skilled analysts based on their experience and judgment, and Determine features that are unlikely to be analyzed elsewhere. This means that a character's structure in a medical setpoint or property validation context where unique properties related to This is particularly useful.

Although the present invention has been described with respect to specific embodiments, those skilled in the art will appreciate that other embodiments may be readily understood. It will therefore be obvious that there is nothing to this invention other than as set forth in the appended claims. It is not limited.

FIG., J.

FIG, -2° abstract A needle with a device that senses a force along at least one axis within the surface at a point on the surface , and sensing the true position along at least one axis of the surface of the point in question. The device consists of a combination of position-sensing platens or pads - directly at a point on a surface. A device that directly collects information. The collected information is processed to obtain the characteristics of the merits. . Extracting features or elastic correlations within the five dimensions is a method of extracting features within the five dimensions. Used to obtain properties.

international search report

Claims (12)

[Claims] 1. A device for collecting handwriting information based on the force and position of a writing instrument, the device comprising: A first position component is sensed along a first axis of a Cartesian coordinate system, and the first axis is lateral to the first axis. a tablet that senses a second positional component along a second axis of the Cartesian coordinate system; device; a force sensing device that senses the force around the intersection of the first axis and the second axis; the force; The sensing device: a device for sensing a first force component in a third axis; a device for sensing a first force component in a fourth axis transverse to said third axis; a device for sensing two force components, and a device for sensing a third force component in a fifth axis that intersects the third axis and the fourth axis; including; To create the recorded position and force information, time, the first position component, the first a recording device that records the second position component, the first force component, the second force component, and the third force component; position; and in response to the recorded position and force information to create output information. a device that combines position and force information; 2. a feature recording wherein said combining device receives said recorded position and force information; A device for creating a feature board, a device for storing the feature board, and a device for storing the feature record and and said feature plate memory to create a characteristic of said feature plate. 2. The apparatus of claim 1, comprising a means for comparing with said feature record. 3. The set of shape plates in the feature shape plate storage device is stored together with the recorded specimen and the position information. The average value and standard deviation value for each time segment of force information, as well as each position and The above set of plates was correlated with the recorded specimens in order to characterize their strength against force and force. 2. The apparatus of claim 1, comprising an apparatus for causing. 4. A device that combines the characteristics of the above position and force advantages into a single numerical value and compares it with the standard. 4. The apparatus of claim 3, further comprising: 5. A method of collecting information based on force and position: a) Descartes with respect to time; sensing a first positional component of the needle tip along a first axis of the coordinate system; b) along a second axis of said Cartesian coordinate system that intersects said first axis with respect to time; , sensing a second position component of the needle tip; c) at the intersection of said first axis and said second axis in a third axis with respect to time; sensing a first force component on said needle tip; d) aligning said first axis with respect to time; of the needle tip at the intersection with the second axis in the fourth axis that crosses the third axis; sensing a second force component; e) a fifth axis that intersects the first axis and the third and fourth axes with respect to time; sensing a third force component of the needle tip at a point of intersection with the second axis in a line; f) to obtain force and position information with respect to a common needle tip in time correlation; a common time reference, the first position component, the second position component, the first force component, the first force component; recording for the second force component and the third force component; g) the first and second position components, and the first and second position components to create output information; combining the second and third force components; and h) providing said output information to an output device. 6. The above combination steps record the positions in order to obtain the characteristics of the merits with respect to the criteria. 6. The method of claim 5, comprising comparing the position information and the input force information to a set of shape plates. Method. 7. Further, the output information includes obtaining one number of merits. Method described. 8. The above comparison step is performed to obtain the 5 numbers of merits. Applying elastic correlation processing within the five dimensions corresponding to the fourth and fifth axes. 7. The method of claim 6, comprising: 9. The comparison step corresponds to the first, second, third, fourth and fifth axes. The extracted features are included in the five dimensions mentioned above, and the features are and the average force along the fifth axis and the average distance along the first and second axes. 7. The method according to claim 6. 10. Furthermore, in order to establish standards for the form of the shape plate set, under controlled conditions 6. The method according to claim 5, comprising repeating steps a) to g) a plurality of times. How to put it on. 11. The comparison step is performed between the recorded information and the information of the shape plate set. 7. The method of claim 6, comprising applying a mean square analysis. 12. The comparison step is performed between the recorded information and the information of the shape plate set. 10. The method of claim 9, comprising applying a mean square analysis.