WO2002052493A1 - Procede et dispositif de reconnaissance de caracteres traces manuellement sur une zone de saisie - Google Patents
Procede et dispositif de reconnaissance de caracteres traces manuellement sur une zone de saisieInfo
- Publication number
- WO2002052493A1 WO2002052493A1 PCT/CH2001/000710 CH0100710W WO02052493A1 WO 2002052493 A1 WO2002052493 A1 WO 2002052493A1 CH 0100710 W CH0100710 W CH 0100710W WO 02052493 A1 WO02052493 A1 WO 02052493A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- character
- segment
- pass
- nature
- characters
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
- G06V10/17—Image acquisition using hand-held instruments
Definitions
- the present invention relates to a method for recognizing characters drawn manually on an input area as well as to an electronic device making it possible to implement this method.
- character will be understood to mean any alphanumeric character (A, B, C 1, 2, 3 %), punctuation (?,!,
- One solution consists in providing complex recognition algorithms based on a point grid or "bitmap" image of the drawn character, point grid which is compared with the model data of a predetermined character set in order to identify the most faithful character. to the line made.
- This solution is impractical and often much too heavy and greedy in time and in computing power to be applied, in particular in portable electronic devices.
- Another disadvantage of such a solution lies in the fact that it requires a large memory space for the definition of the characters to be recognized. Such a solution also requires learning on the part of the user so that he traces each character according to a determined pattern.
- the current practice is to propose recognition devices and algorithms according to which the movement carried out during the drawing of the character on the input area is examined.
- signals from the area of input, or more precisely of the plurality of sensors forming this input zone are sampled in order to determine the history of the evolution of the plot of the character.
- Document EP 0 695 979 in the name of the Applicant describes for example a device for identifying a character transcribed on a surface as well as a method for identifying such a character, in particular for a watch.
- a series of identifiers representative of the changes of state ("ON" or "OFF") of the sensors forming the entry area is produced and compared with at least one series of reference identifiers in order to identify the transcribed character.
- Document EP 0 632 401 proposes to model each character, not only on the basis of models made up of rectilinear segments, but also on the basis of models taking into account other parameters such as lifting ("PEN UP”) or lowering (“PEN DOWN”) of the tracing instrument on the entry zone, the duration of a pause during the tracing, the center of gravity of the tracing, the size of the tracing, the direction of rotation of a curved segment , the position of the start and end points of a curved segment as well as the orientation of such a curved segment.
- This last document teaches that such a solution is implemented on a system with parallel processors using a specific type of memory called associative. Such a parallel processor structure is obviously not conceivable for application to portable electronic devices where the data processing capacities, in particular, are limited.
- a first object of the present invention is therefore to propose a solution which requires virtually no learning on the part of the user, namely a solution which is capable of recognizing characters drawn by the user without the latter does not have to significantly change his habits.
- Another object of the present invention is to propose a solution which is capable of recognizing alphanumeric, symbolic, command or other characters, traced in various ways by the user.
- Yet another object of the present invention is to propose such a solution which is however not too complex nor too cumbersome to implement, so that it can be applied in low-volume electronic devices, such as electronic diaries, computers pockets, timepieces, portable telephones, remote controls, or any other portable electronic device of reduced size, in particular portable electronic devices powered by battery.
- Yet another object of the present invention is to limit the number of errors of interpretation of the plotted character, that is to say to propose a solution which is capable of rejecting certain plots of character which would in no way correspond to a predetermined character. .
- the present invention thus relates to a method for recognizing characters drawn manually on an input area of an electronic device, the characteristics of which are set out in independent claim 1.
- the present invention also relates to an electronic device making it possible to implement the method according to the invention, the characteristics of which are set out in independent claim 10.
- An advantage of the present invention lies in the fact that the characters are not modeled as characters as such, but that each character is defined by at least one predetermined model representative of the way in which this character can be traced. Likewise, the manual drawing of the character made by the user on the input area is represented by a model representative of the way in which this character was drawn. Typically, a character will be defined by more than one model in order to take into account the various ways of manually drawing it.
- the models include at least topological parameters relating to the topology of the character, namely, in particular: the number of passes used to trace the character; the nature of each of said passes, namely whether it is a pass of one-dimensional (one point) or two-dimensional (one or more segments) nature; and the composition of each pass of two-dimensional nature, namely: - the number of segments making up the pass of two-dimensional nature; and the nature of each segment, namely whether it is a straight or curved segment.
- topological parameters can be supplemented by an indication concerning the "closed” nature (for example characters “B”, “D”, “O”, “8”, “0”, etc.) or "open” ( eg characters "A”, “C”, “E”, etc.) from a pass of the character plot.
- “topology” is meant, in the context of the present description, all the characteristics of a character independent of the exact size and dimensions of the character, that is to say independently of the exact geometry of the character.
- topology is meant, in the context of the present description, all the characteristics of a character independent of the exact size and dimensions of the character, that is to say independently of the exact geometry of the character.
- the number of passes used to trace the character as well as the nature and, if necessary, the composition of each of said passes as mentioned above.
- We will also hear other parameters such as the direction of rotation of a curved segment as mentioned below.
- pass is meant, in the context of the present description, the tracing carried out by the finger or the tracing instrument on the input zone from the start until its interruption, each character being able to be traced in a variable number of passes, typically less than or equal to four with respect to an alphanumeric or symbolic character from a set of Latin characters.
- segment is meant a portion of a pass which is distinguished from another segment of the pass by its straight or curved nature and / or by a significant change of direction adopted by the finger or the instrument of tracing during the execution of the pass in question.
- the models also include, where appropriate, for each segment of rectilinear or curved nature, geometric parameters representative of the evolution of the angular orientation of the path of the segment.
- the geometric parameters representative of the evolution of the angular orientation of the segment layout include: for each segment of rectilinear nature, the angular orientation (designated ⁇ j>) of the segment; or - for each segment of curved nature, the angular orientation of departure or arrival (designated respectively ⁇ -j, ⁇ 2) at one end of the segment (whose positions are designated respectively P- *, P2).
- angular orientation one will understand the orientation and the direction of a rectilinear segment, or of a tangent to a curved segment taken at a point of this curved segment, indicated by an angle value , as shown schematically in Figures 12 and 13a to 13c.
- the angular orientation of a segment or of a tangent will preferably be defined, solely for the purposes of the explanation, as being between the values -180 ° and 180 °. It will however be understood that these values are defined to the nearest n * 360 ° (n being an integer), -180 ° and 180 ° thus designating the same angular orientation.
- the parameters concerning the topology of the character also include, for each segment of curved nature, the direction of rotation (designated w) of the segment, this direction of rotation being at least trigonometric (in the opposite direction clockwise) or non-trigonometric (clockwise).
- a third direction of rotation called mixed or variable, that is to say a direction of rotation which varies between trigonometric directions and non-trigonometric, such a characteristic being for example present in certain character plots, such as character plots "N”, “M”, “W”, “S”, “2”, “5", “8” , etc.
- mixed or variable a third direction of rotation which varies between trigonometric directions and non-trigonometric
- character plots "N”, “M”, “W”, “S”, “2”, “5", “8” character plots "N”, “M”, “W”, “S”, “2”, "5", “8” , etc.
- the geometric parameters representative of the evolution of the angular orientation of the segment can also advantageously include, for a segment of curved nature, the total development angle (designated ⁇ ) of the segment of curved nature.
- this total development angle is preferably defined as being the sum, in absolute value, of the total development angles of each portion of trigonometric and non-trigonometric direction of rotation.
- the total development angle ⁇ simplifies the identification of certain particular characters.
- each model also includes geometric parameters representative of the position of each pass on the input area, in particular and advantageously the position of each pass of one-dimensional nature (a point) or the position of the ends of each two-dimensional nature pass.
- these geometric parameters can be supplemented by the position of the ends of each segment making up a pass of two-dimensional nature.
- the position of a pass of one-dimensional nature or of the ends of a pass of two-dimensional nature can be defined with respect to the input area on which the pass was made or, alternatively, in relation to the layout of the character concerned.
- use will be made essentially of indications of position relative to the outline of the character, that is to say indications of position making it possible to define whether such a pass is in a upper or lower part, and / or left or right of the character outline.
- indications of relative position with respect to the input zone on which the pass was drawn to differentiate characters whose lines do not distinguish only by their position on the input area it will also be understood that all the models do not necessarily require a geometric definition of their characteristics. Thus, certain characters can be unequivocally defined only by means of topological parameters.
- FIG. 11 is an illustration of a one-dimensional pass, that is to say a point, defined by its position P *
- FIG. 12 is an illustration of a segment of rectilinear nature defined by its angular orientation ⁇ and the positions P- ⁇ , 2 of its ends.
- FIGS. 13a, 13b and 13c are illustrations of segments of curved nature each defined by the angular orientations of departure and arrival ⁇ i, ⁇ 2 at the ends of the segment, the positions P * ⁇ , P2 of the ends, as well as the direction of rotation w, respectively non-trigonometric, trigonometric and mixed in Figures 13a to 13c.
- the total development angle ⁇ of each curved segment has also been indicated.
- the total development angle ⁇ is defined, in this case, as the sum in absolute value of the total development angles, designated ⁇ -i, c.2, of the trigonometric portions (s ) and non-trigonometric (s) of the curved segment.
- a vertical straight segment drawn from top to bottom will be defined as having an angular orientation ⁇ of -90 °
- a vertical straight segment drawn from bottom to top will be defined as having an angular orientation ⁇ of + 90 °
- a straight segment drawn from left to right will be defined as having an angular orientation of 0 °
- a straight segment drawn from right to left will be defined as having an angular orientation of ⁇ 180 °.
- these angular orientations are defined to within n-360 °.
- a curved segment representing a "C" drawn from top to bottom or from bottom to top will, by way of example only, be defined as having an angular orientation of departure ⁇ i of approximately ⁇ 180 ° and an angular orientation of arrival ⁇ 2 of approximately 0 ° (the total development angle ⁇ is 180 °). These two curved segments differ only in their direction of rotation.
- the angular orientation of the segment will be developed from -170 ° to 0 ° (the total development angle ⁇ is here 170 °), etc.
- this curved segment could be broken down into two successive curved segments of respectively trigonometric and non-trigonometric direction of rotation (each having a total development angle ⁇ of 180 °).
- topological criteria that is to say criteria independent of the exact shape and dimensions of the character, makes it possible to carry out a significant sorting and preselection of the candidates likely to correspond to the traced character.
- the angular orientations of the segments making up the traced character are preferably first compared with the angular orientations of the candidate segments and the most faithful candidate (s) is selected on this basis. If this selection results in a set of several candidates, then use is made of the position information of the passes (namely the position of a one-dimensional pass and / or the position of at least one point representative of a two-dimensional pass) of the character plot to identify the best candidate. Thanks to the invention, it is thus possible to model a large number of characters (at least 80 characters without counting the lower case letters) as well as a large number of variants of plots for the same character, learning for the user being thus limited and the use of the device implementing this process facilitated.
- the recognition rate resulting from the implementation of the recognition method according to the present invention is also very high and the risk of introducing erroneous data is therefore limited. It is also very easy to add new models, the approach used being applicable regardless of the number or complexity of the models defined.
- Yet another advantage of the present invention is to allow the taking into account of writing characteristics specific to a right-handed person and / or a left-handed person.
- FIG. 1 is a schematic sectional view of a timepiece comprising a character recognition device according to the present invention
- FIG. 2 shows a spatial arrangement of sensors forming an input area of the character recognition device
- FIG. 3 shows a block diagram of a character recognition device according to the present invention employing, without limitation, a network of capacitive sensors as illustrated in Figure 2;
- FIGS. 4a to 4d illustrate in a non-exhaustive manner a set of models of alphanumeric, symbolic and control characters which can be defined in the - 1 U -
- Figure 5 illustrates a possible plot of the character "A" on the input area of Figure 2;
- FIG. 6a to 6c show three examples of plotting the character "R” and a diagram illustrating the evolution of the plot of this character, in terms of angular orientation, for each pass used to perform this plot;
- Figures 8a and 8b show a flow diagram of a variant implementation of the method illustrated in Figures 7a and 7b;
- FIG. 9a and 9b show, by way of illustration, respectively, a possible plot of the character "R" and of the candidates having a similar topology
- FIG. 12 is a representation of the definition of a rectilinear segment
- Figures 13a to 13c are representations of the definitions of a segment of curved nature respectively having a non-trigonometric, trigonometric and mixed direction of rotation.
- the recognition device according to the present invention will be described in the context of a particular application, namely an application in a timepiece. It will nevertheless be understood that the invention applies in an equivalent manner to any other electronic device into which it is desired to enter data by means of a finger or of a tracing instrument. It will also be understood that the nature and the composition of the input area of the recognition device are not decisive.
- the input area is defined in the context of the present description as comprising sensors of the capacitive type, it will easily be understood that any other type of sensor can be used, such as optical, ohmic, inductive sensors, propagation of surface acoustic waves (SAW) or any other type of sensor capable of being actuated by a finger or a tracing instrument so as to produce a signal which can be used to extract topology and geometry information from the plot made on the input area.
- SAW surface acoustic waves
- FIG. 1 represents a sectional view of a timepiece 1 comprising a housing 2 and a crystal 3.
- An electronic circuit 4 is arranged in the housing 2. - not -
- conductive electrodes preferably transparent, are arranged on the inner face 31 of the glass 3, only four of these electrodes, respectively referenced 15B, 15F, 15J and 15N, having been shown in FIG. 1.
- the conductive electrodes 15B, 15F, 15J and 15N are respectively connected to the electronic circuit 4 by conductors 6, 7, 8 and 9.
- a battery 11 or other source of electrical energy (such as a rechargeable battery) is also placed in the housing. 2 and is connected to a supply pole of the electronic circuit 4 by a conductor 12.
- Each of the conductive electrodes 15B, 15F, 15J, 15N (15A to 15P in FIG.
- This input area 10 comprises in this example a plurality of sensors 15, in this case sixteen in number, arranged in the form of a matrix four by four as illustrated.
- a parasitic capacitance is, by construction, present between each of the conductive electrodes 15A to 15P and the case 2 of the timepiece 1.
- Each of the capacitive sensors 15, of which the conductive electrodes form part, and the parasitic capacitance associated with this capacitive sensor are connected in parallel between ground and an input of the electronic circuit 4.
- each capacitive sensor is connected to the input of a voltage controlled oscillator, the frequency of this oscillator varying as a function of the total capacity present between its input and the mass.
- each capacitive sensor can be connected to its own voltage-controlled oscillator or, alternatively, all can be connected to a single voltage-controlled oscillator via a multiplexer.
- a multiplexer obviously requires a multiplexing period of substantially short duration less than the movement of the finger or of the tracing instrument over the input area and the period of sampling of the signals, so that the signals emanating from each sensor are considered, at a given instant, to be simultaneous measurements of the state of the sensors.
- the frequency of oscillation of the signal produced at the output of the voltage-controlled oscillator is proportional to the inverse of the total capacity of the two capacities connected in parallel.
- the finger 13 of the user of the timepiece 1 is not placed on the crystal 3 opposite a particular electrode, one of the armatures of the corresponding capacitive sensor is therefore not formed.
- the total capacity present between the input and the mass of the oscillator is equivalent to the value of the parasitic capacity, and the frequency of oscillation of the output signal of the voltage-controlled oscillator is proportional to l 'inverse of this.
- the frequency of oscillation of the output signal of the voltage-controlled oscillator is proportional to the inverse of the sum of the capacities. It will also be noted that the variation in frequency caused by the presence of finger 13 is dependent on the surface of the electrode covered by finger 13. Thus, it is possible to detect not only the presence or absence of finger 13 opposite of the electrodes but also the degree of coverage of these electrodes by the finger 13.
- the information desired in the output signal is contained in the frequency of the signal produced by the voltage-controlled oscillator. It suffices to use digital means designed in a simple way to extract this information.
- the resolution of the input area is sufficient to obtain a sufficiently precise image of the manual plot performed on this input area.
- the resolution of the input zone also depends on the tracing tool used and on its contact surface with the input zone, the contact zone of a finger is obviously larger than that of a stylus for example.
- the arrangement illustrated in FIG. 2 (and in FIG. 5) is particularly suitable for allowing the use of a finger as a tracing tool, each sensor having an area of the order of 1 cm 2 less than the size of the area. typical finger contact.
- the timepiece or, more generally, the electronic device in which the recognition device according to the present invention is incorporated, with display means (not shown in the figure 1) making it possible to provide the user with an indication of the trace he has made on the input area 10.
- display means consisting for example of a liquid crystal display, can advantageously be arranged in the room clockwork of FIG. 1 so as to facilitate the use of the recognition device, the user benefiting from directly readable confirmation information opposite the input area 10.
- Figure 3 shows a particular block diagram of a character recognition device according to the present invention employing a network of capacitive sensors. This diagram is only illustrative.
- display means generally indicated by the reference numeral 5, and comprising in particular a display control circuit 50 as well as a display device 55 such as a liquid crystal display, the latter typically being in the form of a matrix display allowing the display of a plurality of lines of characters and, where appropriate, the display of an image of the manual trace produced on the input area 10.
- the input area 10 and its capacitive sensors are connected to one (by multiplexing) or several voltage-controlled oscillators 41 delivering at output, at a given instant designated t ⁇ , frequency signals fyrtk), the indices i and j being indicative of the position of the sensor in the matrix forming the input area 10.
- a frequency detector 42 connected to the output of the voltage-controlled oscillator (s) 41 makes it possible to determine the frequency of the signal produced for each sensor, for example by determining the number of zero crossings of the signal during a determined measurement period, this number, for each sensor, being indicated by the reference NyOk).
- the frequency detector 42 typically receives a clock signal produced by clock means 43 and defining the measurement period for determining the frequency of the signals produced for each sensor.
- clock means 43 produce a second clock signal determining the sampling frequency of the signals produced by the sensors of the input zone 10 during the drawing of the character.
- the number Njj (tk) representative of the frequency of the signal produced for each sensor of the input zone 10 at the instant tk is transmitted to a microcontroller 45 responsible for processing the input signals of the trace of the character and for recognizing this character.
- the microcontroller 45 uses data stored in storage means 46 typically comprising a random access memory (RAM), a read only memory (ROM), and / or a reprogrammable memory (EEPROM).
- RAM random access memory
- ROM read only memory
- EEPROM reprogrammable memory
- the data used by the microcontroller 45 notably includes data relating to the various models of alphanumeric, symbolic and control characters, these data being stored in the form of a library of characters 460 as shown diagrammatically in FIG. 3.
- FIG. 3 the introduction of an alphanumeric text has also been illustrated (in this example the introduction of the term "HELLO").
- each character is successively introduced by means of the input area 10 and recognized by the microcontroller 45 according to a process which will be fully described below.
- the manual drawing of the character, designated 100 carried out on the entry zone 10 (here the character "O") is preferably displayed on the display 55 (as illustrated in a). in order to offer the user confirmation of the route taken.
- the character is validated (or rejected if necessary) and displayed on the display device 55 as illustrated in b) in FIG. 3.
- the operation of the character recognition device will now be described more in detail with reference to Figures 4 to 10.
- FIGS. 4a to 4d schematically and partially illustrate a set of models representative in various ways with which various alphanumeric, symbolic and command characters can typically be traced on the input area 10 within the framework of the present invention.
- the illustrated assembly is in no way exhaustive or limiting and that other routes can easily be envisaged.
- the drawing of many alphabetic characters such as the characters "B”, “D”, “E”, “F”, “M”, “N”, “P” or “R”
- FIGS. 4a and 4b in particular illustrate character models of the alphabet, FIG. 4c of the numerical and symbolic character models and FIG. 4d of the possible models of control characters.
- the various passes necessary for tracing the character designated A to D, have been illustrated with different colors or different patterns.
- the first pass A is shown in black
- the second pass B in gray
- the third pass C in white
- the fourth pass D with hatching as indicated in the legend of FIG. 4a.
- the arrow in each pass indicates the direction of tracing of this pass, two opposite arrows indicating that the pass in question can be traced independently in one direction or another.
- the various models will preferably be grouped together first according to the number of passes. Each of these categories is then broken down according to the nature and composition of these passes.
- the model also preferably includes geometric parameters relating to the position of this one-dimensional pass.
- the model also preferably includes geometric parameters relating to the angular orientation or orientations of each segment, and, for each curved segment, at least one topological parameter relating to the direction of rotation of the curved segment (trigonometric or non-trigonometric, even mixed).
- a tree structure is thus formed which ultimately leads to a model particular of a character.
- the line drawn by the user on the input area does not correspond to an alphanumeric, symbolic or command character but corresponds to an undefined character.
- the drawing of a character comprising a pass formed by a number of segments greater than a determined limit corresponds to an undefined character or, advantageously, to a specific control character. It is for example possible to define a certain number of specific control characters having a unique topological definition, that is to say characters which do not require geometric parameters to allow their recognition.
- first specific control character by a model formed, for example, of a single pass comprising more than five successive segments, or a second specific control character by a first pass comprising more than five segments successive followed by a second pass comprising a point as illustrated in the right part of FIG. 4d.
- the distinction is advantageously made between the one- or two-dimensional nature of each pass.
- a point can thus quickly be differentiated from a segment or from several successive segments (rectilinear or curves) if a very localized trace contained in an area of reduced area is detected.
- it will be possible to quickly reduce the search field of a traced character as soon as a pass recognized as a point appears, this characteristic being present only in a reduced number of characters, the latter being essentially punctuation characters, symbolic characters, or control characters.
- a first aspect of the invention thus requires the constitution of a catalog or set of models for each character capable of being traced by the user, each model being representative of at least one way of tracing the character in question, a such a game being for example illustrated in FIGS. 4a to 4d.
- FIG. 5 shows the arrangement of the sensors of FIG. 2 in which, by way of example, the manual layout, designated 100, produced by the finger 13 representing the character "A" has been symbolized by two arrows.
- the finger 13 has first made a first pass, designated A, to perform a first part of the character, that is to say the part of concave shape of the character "A" (here in the form of an inverted U-shaped curved segment), then a second pass, designated B, for drawing the substantially horizontal bar of the character "A".
- A a first pass
- B for drawing the substantially horizontal bar of the character "A”.
- This example corresponds to a natural and typical manual layout of a right-hander with the character "A”.
- this same character can be traced in other ways, for example in a first pass in the form of an inverted V then a second pass consisting of a straight rectilinear segment, or even in three passes as illustrated in Figure 4a.
- the input area is formed by a matrix of four by four discrete capacitive sensors each connected to a voltage controlled oscillator (as described above) and producing an electrical signal whose frequency varies depending on the degree of coverage of each sensor by the finger.
- a voltage controlled oscillator as described above
- an average position of the finger on the input area is thus determined in real time by calculating the center of gravity or barycenter of the activated sensors, preferably by assigning them a greater or lesser weight depending on the degree of coverage of these sensors by the finger, this weight being maximum when the finger covers a sensor over its entire surface, and minimum when only a small part of this sensor is covered.
- the first step is to detect the number of passes used by the user to trace the character.
- This number of passes is easily determined by detecting the number of interruptions during the drawing of the character, that is to say the appearance of a period of determined duration (typically several successive periods of inactivity) during which none of the sensors was not activated by the presence of the finger, the end of the character tracing being detected in a similar manner but on the basis of a longer period of inactivity. A first selection of candidates can then be made on the basis of the number of passes.
- each segment is for example determined by linear approximation of each segment by a conventional method of least squares known to those skilled in the art. If the difference between this linear approximation model is greater than a determined difference, it is assumed that the segment in question is of a curved nature and, for example, the starting and ending angular orientations are determined ⁇ * ⁇ , ⁇ 2 of this segment. In this case, the direction of rotation w of the curved segment is also preferably determined. Otherwise, it is assumed that this segment is rectilinear in nature and its angular orientation ⁇ is determined. It is also preferable to determine the position of the extreme points of each segment, designated P * ⁇ and 2.
- FIGS. 6a to 6c show, by way of comparison, three examples of plots of the character "R", FIG. 6a illustrating a plot in two passes and FIGS. 6b and 6c illustrating a plot in a single pass.
- FIGS. 6a to 6c are not exhaustive of the possibilities that can be adopted for drawing the character "R".
- the first pass A consists of a substantially vertical rectilinear segment A1 traced from top to bottom and consequently having an angular orientation ⁇ of -90 ° approximately.
- the second pass B consists of two successive segments, respectively a curved segment B1 followed by a straight segment B2 both drawn from top to bottom.
- the curved segment B1 has angular orientations of departure ⁇ - * and arrival ⁇ 2 respectively equal to 0 ° and -180 ° approximately, the direction of rotation being non-trigonometric.
- the rectilinear segment B2 has an angular orientation ⁇ of approximately -45 °.
- the differentiation between the two segments B1 and B2 of the second pass B can easily be carried out because there is an abrupt variation in the angular orientation (from -180 ° to -45 °) at the end of the drawing of the part. "R" curve before drawing the last straight portion.
- Pass B results in its first portion similar, as a first approximation, to a straight portion of negative slope between 0 ° and -180 ° approximately and corresponding to the curved segment B1, followed by a substantially horizontal straight portion located around -45 ° and corresponding to the rectilinear segment B2.
- the sudden change in angular orientation between segments B1 and B2 is clear from this diagram.
- FIG. 6b illustrates a plot of the character "R" produced in a single pass A, that is to say a rectilinear segment A1 at approximately + 90 °, followed by a curved segment A2 varying from 0 ° to -180 ° approximately and finally a rectilinear segment A3 at around -45 °.
- a first substantially straight line portion at approximately + 90 °
- a second straight portion of negative slope between 0 ° and -180 ° approximately
- a third straight portion substantially horizontal at approximately -45 ° In the diagram of the evolution of the angular orientation of pass A, there is thus a first substantially straight line portion at approximately + 90 °, followed, as a first approximation, by a second straight portion of negative slope between 0 ° and -180 ° approximately, and finally a third straight portion substantially horizontal at approximately -45 °.
- FIG. 6c illustrates another possibility of drawing the character "R" in a single pass A, this pass A comprising a first rectilinear segment A1 at approximately -90 °, followed by a curved segment A2 varying from + 90 ° to -45 ° approximately with a mixed direction of rotation, that is to say a curve whose angular orientation decreases first of all from + 90 ° to -180 ° approximately then increases from -180 ° to -45 ° approximately.
- the curved segment A2 of mixed direction of rotation can alternatively be broken down into two successive curved segments of non-trigonometric then trigonometric direction of rotation.
- FIG. 7a describes a first part of a flowchart illustrating this mode of implementation of the method according to the invention, the second part of this flowchart being shown in FIG. 7b.
- step S100 begins at step S100 by sampling the plotted character, this sampling comprising a possible prior processing of the signal if necessary.
- step S102 the number of passes made to trace the character is determined, this number allowing a selection of a first set of candidates comprising a number of identical passes as shown in step S104.
- step S106 we begin to determine the nature and then the composition of each pass as shown schematically by the steps - -
- step S108 the one-dimensional or two-dimensional nature of each pass is first determined, namely whether it is a point or a number of successive segments. If this pass is a point, the position of this point is determined in step S109, otherwise, the segmentation of the pass in question is started as illustrated by steps S110 to S118 which are repeated as many times as there are of segments in the pass in question.
- step S110 the number of segments making up the pass is thus determined, for example by determining sudden changes in angular orientations in the pass as already mentioned, the determination of the positions P- *, P2 of the ends segments being performed at this stage.
- the recognition process can advantageously be interrupted by steps S140 and S142 if the number of segments is greater than a determined limit which shows that the plot does not correspond to any defined character. If this is not the case, in step S114, the rectilinear or curved nature of each segment is determined, for example by linear approximation of each segment and selection as a function of the difference between this linear approximation and the plot of the segment in question. Depending on the rectilinear or curved nature of the segment, the angular orientation ⁇ of the rectilinear segment in step S1 5 or the angular orientations of departure and arrival ⁇ i, ⁇ 2 of the curved segment in step S116 are determined respectively. If the segment is of curved nature, the topological information is supplemented by the determination of the direction of rotation w of the curved segment as shown schematically by step S117. This process is repeated for each segment and for each pass as illustrated by steps S118 and S120.
- step S124 a second set of candidates is thus selected from among the candidates of the first set determined in step S104 whose nature and composition of the passes are identical to those of the model of the plotted character. It is not necessary, at this stage, to consider the information of a geometrical order of the models, namely the parameters of evolution of the angular orientation or of position of the segments.
- step S126 it can then be identified, in step S126, if the character traced corresponds to a defined model or if this model is unknown in which case - ----- -
- the second set contains no candidate.
- step S 1208 a third set of candidates is selected from the candidates of the second set whose geometry in terms of angular orientation is the closer to that of the model of the drawn character.
- the difference is determined for example in terms of angular orientation between the segments constituting each pass and the best candidate (s) with the smallest difference are selected. In practice, we allow a certain tolerance on these angular orientations (for example ⁇ 10 °).
- step S129 it is advantageous to reject the candidate (s) if the difference in terms of angular orientation is too large, this signifying that the character traced is not defined. The process then ends with steps S140 and S142.
- step S130 If the candidate (s) are judged to be sufficiently faithful to the plotted character, it is determined first of all, at step S130, if more than one candidate has been selected at this stage, in which case one selects, at step S132, the candidate from among the candidates of the third set whose geometry in terms of position (absolute or relative) of the segments is closest to that of the model of the character plotted, then this candidate is validated in step S134. If the third set includes only one candidate in step S130, the latter is directly validated in step S134 as illustrated.
- step S134 the character plotted has thus been recognized and validated and it is therefore possible to display this character if it is an alphanumeric (or symbolic) character or to execute the associated command if it is a control character as illustrated by steps S136, S137 and S138.
- steps S136, S137 and S138 the flowchart illustrated in FIGS. 7a and 7b constitutes only a particular mode of implementation of the invention and that various modifications, such as the redistribution of certain tests upstream or downstream in the flowchart, and / or improvements can be made.
- the mode of implementation illustrated in FIGS. 7a and 7b can advantageously be modified so as to include additional steps allowing the identification of specific command characters with which particular commands are associated.
- These specific control characters are advantageously described unequivocally only by topological information or parameters.
- a trace of a pass comprising more than a certain number of successive segments (for example at least five) can be considered as the trace of a specific command character with which a specific command is associated as illustrated in FIG. 4d (for example the erasure of a character previously entered or the complete reinitialization of the data entered). Since no other character can be traced in this way (provided, however, that it has not been defined in this way), it is not necessary to determine the straight or curved nature of segments, nor to the determination of the geometric parameters of these segments.
- step S 13 following step S112 aimed at determining the number of segments per pass
- this additional step S113 consisting in restricting the first set of candidates determined at the step S104 has a specific set of control characters, in which case neither the straight or curved nature of the segments, nor the geometry of these segments are determined in steps S114 to S116.
- Another additional step S125 is then provided following the selection step S124 of the second set of candidates, this step S125 aimed at identifying whether the candidate corresponds to a specific command character, in which case the command associated with this command character. specific is directly executed in step S138.
- step S118 it is also possible to provide an additional step following step S118 in FIG. 7a (or 8a) whose purpose is to filter segments whose length is not sufficiently significant, these segments being able to be considered as hesitation on the part of the user.
- Figure 9a shows an example of the character "R" plotted on the input area.
- the character is drawn in a single pass A in the form of a first rectilinear segment A1 followed by a curved segment A2 and a second rectilinear segment A3.
- a first set of candidates capable of being drawn in one pass will thus be selected in step S104.
- the rectilinear or curved nature of the three segments A1 to A3 of this single pass as well as their geometric data of angular orientation and position will then be determined in steps S110 to S120 in accordance with what has already been mentioned.
- the first rectilinear segment A1 has an angular orientation ⁇ of approximately + 80 °
- the curved segment A2 has a starting angular orientation ⁇ ** of approximately + 10 ° and an angular orientation of arrival ⁇ 2 of approximately -170 °
- the second rectilinear segment A3 has an angular orientation ⁇ of approximately -60 °.
- the curved segment A2 is also drawn with a non-trigonometric direction of rotation.
- a model of the traced character formed of these characteristics is then developed in step S122.
- the angular orientations of the segments of the traced character can be summarized as follows: -. b -
- step S124 all the candidates whose nature and composition are identical to the model of the traced character are selected, that is to say all the models defined as being produced in a pass formed by three successive segments comprising a first segment of rectilinear nature, a second segment of curved nature and a third segment of rectilinear nature, in that order.
- the models of the character "M", of the character "R” and of the character "W” defined in the character set of FIGS. 4a and 4b are shown in FIG. 9b (these models have obviously been previously defined in this way).
- These candidates have the same topology (without considering the direction of rotation of the curved segment) as the character plotted in FIG. 9a, ie a first segment of rectilinear nature, followed by a second segment of curved nature, then a third segment of nature straight.
- Other models could exist if necessary, but these have not been represented for the sake of simplification.
- the candidate "M” is rejected, the latter having a curved segment of direction of trigonometric rotation, unlike the candidates "R” and "W".
- step S128 The recognition process then continues in step S128 with the __t
- Figure 10a shows an example of the character "T" plotted on the input area.
- the character is drawn in a first pass comprising a first rectilinear segment A1 followed by a second pass B comprising a second rectilinear segment B1.
- a first set of candidates capable of being traced in two successive passes will thus be selected in step S104.
- the rectilinear nature of the two segments A1 to B1 as well as their geometric data of angular orientation (and of position) will then be determined in steps S110 to S120 in accordance with what has already been mentioned.
- the passes can obviously be reversed.
- the first rectilinear segment A1 has an angular orientation ⁇ of approximately + 5 °
- the rectilinear segment B1 has an angular orientation ⁇ of approximately -95 °.
- a model of the traced character formed of these characteristics is then developed in step S122.
- the angular orientations of the segments of the traced character can be summarized as follows:
- step S124 all the candidates whose nature and composition are identical to the model of the traced character are selected, that is to say all the models defined as being carried out in two passes each formed by a - __e
- FIG. 10b shows five candidates meeting this definition (these models have obviously been defined in this way beforehand).
- Other models could exist, if necessary, but these have not been represented for the sake of simplification.
- step S132 of FIG. 7b the process continues at step S132 of FIG. 7b during which the candidate whose geometry in terms of position is closest to that of the model of the plotted character is selected.
- a possible solution, applicable to all characters, can consist in determining the position of each point or of each end of the segments forming the character in a normalized Cartesian frame of reference, that is to say a frame of reference Cartesian type whose vectors x and y coordinates are normalized to the width (along the x coordinate axis) and height (along the y coordinate axis) of the plotted character. The position of the ends of the segments and / or points forming the character can then be determined in the normalized Cartesian frame of reference thus defined.
- the candidate "T” can thus easily be recognized as the character most faithful to the character plotted by determining the position difference of the various ends of the segments, the character “T” being in particular characterized by the fact that the ends of the first segment A1 are situated substantially in the upper part of the character.
- the case which has just been presented and which arises when drawing the character "T” (or alternatively when drawing the character "+") is similarly presented for other characters.
- Such an approach will for example be adopted to differentiate the traces of the characters "X” and "Y", or else the traces of the characters "D” and "P".
- the models will thus be grouped into a plurality of families according to their topological parameters and, where appropriate, their geometric parameters of evolution of the angular orientation, each family comprising one or more models which cannot be distinguished from other models of the family only as a function of their geometric position parameters, this being the case of the characters "T” and "+”, “X” and “Y”, “D” and "P”, etc.
- the method consists in (i) identifying the family of models which best corresponds to the model of the manual line drawing of the character, then in (ii) identifying the model in this family which has the geometric position parameters closest to the model manual character tracing as described above.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Character Discrimination (AREA)
Abstract
Description
Claims
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020037008754A KR100860923B1 (ko) | 2000-12-27 | 2001-12-13 | 입력 존에서 수작업으로 그려넣어진 문자들을 인식하기위한 방법 및 장치 |
JP2002553717A JP4122226B2 (ja) | 2000-12-27 | 2001-12-13 | 入力ゾーン上で手動トレースされた文字を認識するための方法およびデバイス |
EP01271938.1A EP1350218B1 (fr) | 2000-12-27 | 2001-12-13 | Procede et dispositif de reconnaissance de caracteres traces manuellement sur une zone de saisie |
US10/451,534 US7042809B2 (en) | 2000-12-27 | 2002-11-14 | Method and device for recognizing manually traced characters on an input zone |
HK04104194A HK1061098A1 (en) | 2000-12-27 | 2004-06-10 | Method and device for recognising manually traced characters on an input zone |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP00204756A EP1220140A1 (fr) | 2000-12-27 | 2000-12-27 | Procédé de reconnaissance de caractères tracés manuellement sur une zone de saisie et dispositif électronique permettant de mettre en oeuvre ce procédé |
EP00204756.2 | 2000-12-27 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2002052493A1 true WO2002052493A1 (fr) | 2002-07-04 |
Family
ID=8172530
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CH2001/000710 WO2002052493A1 (fr) | 2000-12-27 | 2001-12-13 | Procede et dispositif de reconnaissance de caracteres traces manuellement sur une zone de saisie |
Country Status (7)
Country | Link |
---|---|
US (1) | US7042809B2 (fr) |
EP (2) | EP1220140A1 (fr) |
JP (1) | JP4122226B2 (fr) |
KR (1) | KR100860923B1 (fr) |
CN (1) | CN1252634C (fr) |
HK (1) | HK1061098A1 (fr) |
WO (1) | WO2002052493A1 (fr) |
Families Citing this family (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2847996B1 (fr) * | 2002-11-28 | 2005-10-07 | Thales Sa | Dispositif d'affichage dedie a detection de proximite |
JP2005135268A (ja) * | 2003-10-31 | 2005-05-26 | Internatl Business Mach Corp <Ibm> | 紙面をレイアウトするための装置、方法、およびプログラム |
CN100405389C (zh) * | 2004-08-06 | 2008-07-23 | 摩托罗拉公司 | 从划线标记识别字符的方法和装置 |
JP4648045B2 (ja) * | 2005-03-17 | 2011-03-09 | 株式会社リコー | 文字入力装置、プログラムおよび文字入力方法 |
EP2196944B1 (fr) * | 2005-10-18 | 2014-10-01 | Authentec Inc. | Capteur d'empreintes digitales incluant un circuit flexible et procédés associés |
US8280169B2 (en) * | 2005-12-21 | 2012-10-02 | Michael Linderman | Recordation of handwriting and hand movement using electromyography |
JP4607797B2 (ja) * | 2006-03-06 | 2011-01-05 | 株式会社東芝 | 行動判別装置、方法およびプログラム |
CN101256624B (zh) * | 2007-02-28 | 2012-10-10 | 微软公司 | 建立适用于识别手写东亚字符的hmm拓扑结构的方法及系统 |
US20100020033A1 (en) * | 2008-07-23 | 2010-01-28 | Obinna Ihenacho Alozie Nwosu | System, method and computer program product for a virtual keyboard |
US8769427B2 (en) | 2008-09-19 | 2014-07-01 | Google Inc. | Quick gesture input |
TWI444858B (zh) * | 2009-05-27 | 2014-07-11 | Htc Corp | Method and system for updating the gesture track of hand - held touch device |
CN101923635B (zh) * | 2009-06-16 | 2014-11-26 | 宏达国际电子股份有限公司 | 手持触控装置的笔势轨迹更新方法及其系统 |
US8824245B2 (en) * | 2010-10-25 | 2014-09-02 | Advance Watch Company, Ltd. | Touch screen watch |
US9104306B2 (en) | 2010-10-29 | 2015-08-11 | Avago Technologies General Ip (Singapore) Pte. Ltd. | Translation of directional input to gesture |
CN103425425B (zh) * | 2012-05-22 | 2018-09-04 | 北京蒙恬科技有限公司 | 手写输入选字系统及方法 |
EP2784607B1 (fr) * | 2013-03-28 | 2017-01-18 | ETA SA Manufacture Horlogère Suisse | Montre comprenant un rehaut |
JP6180888B2 (ja) * | 2013-11-07 | 2017-08-16 | 株式会社東芝 | 電子機器、方法およびプログラム |
CN104537343B (zh) * | 2014-12-22 | 2018-11-16 | 中山大学 | 一种基于网格多级精确度递进式比划识别方法 |
JP6481513B2 (ja) * | 2015-05-28 | 2019-03-13 | アイシン精機株式会社 | 操作装置 |
US10503271B2 (en) * | 2015-09-30 | 2019-12-10 | Apple Inc. | Proximity detection for an input mechanism of an electronic device |
JP7171500B2 (ja) | 2018-06-25 | 2022-11-15 | 株式会社クボタ | 作業車両 |
KR20220112368A (ko) * | 2021-02-04 | 2022-08-11 | 주식회사 네오랩컨버전스 | 필기 데이터 문자 인식 방법 및 그 장치 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2538581A1 (fr) * | 1982-12-27 | 1984-06-29 | Casio Computer Co Ltd | Appareil de reconnaissance de caracteres |
WO1990015399A1 (fr) * | 1989-06-08 | 1990-12-13 | Midhat Gazale | Dispositif electromecanique de reconnaissance de caracteres executes manuellement |
EP0632401A2 (fr) * | 1993-06-28 | 1995-01-04 | International Business Machines Corporation | Procédé et appareil de reconnaissance en ligne et en temps réel de caractères manuscits avec des traits |
EP0695979A1 (fr) * | 1994-08-02 | 1996-02-07 | Asulab S.A. | Dispositif d'identification d'un caractère transcrit sur une surface et procédé d'identification d'un tel caractère, notamment pour une montre |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CH599599B5 (fr) | 1974-09-04 | 1978-05-31 | Centre Electron Horloger | |
CH615520A5 (fr) | 1977-06-08 | 1980-01-31 | Centre Electron Horloger | |
US4477797A (en) | 1980-12-12 | 1984-10-16 | Citizen Watch Company Limited | Data input device for electronic device |
US6366697B1 (en) * | 1993-10-06 | 2002-04-02 | Xerox Corporation | Rotationally desensitized unistroke handwriting recognition |
US5596656B1 (en) * | 1993-10-06 | 2000-04-25 | Xerox Corp | Unistrokes for computerized interpretation of handwriting |
FR2717917B1 (fr) | 1994-03-24 | 1996-05-03 | Asulab Sa | Montre comportant un dispositif de commande manuelle. |
DE69533479T2 (de) * | 1994-07-01 | 2005-09-22 | Palm Computing, Inc., Los Altos | Zeichensatz mit zeichen aus mehreren strichen und handschrifterkennungssystem |
FR2755269B1 (fr) | 1996-10-25 | 1998-12-04 | Asulab Sa | Dispositif d'identification d'une action manuelle sur une surface, notamment pour une piece d'horlogerie |
US5940535A (en) * | 1996-10-31 | 1999-08-17 | Industrial Technology Research Institute | Method and apparatus for designing a highly reliable pattern recognition system |
-
2000
- 2000-12-27 EP EP00204756A patent/EP1220140A1/fr not_active Withdrawn
-
2001
- 2001-12-13 JP JP2002553717A patent/JP4122226B2/ja not_active Expired - Fee Related
- 2001-12-13 CN CNB018215815A patent/CN1252634C/zh not_active Expired - Fee Related
- 2001-12-13 WO PCT/CH2001/000710 patent/WO2002052493A1/fr active Application Filing
- 2001-12-13 EP EP01271938.1A patent/EP1350218B1/fr not_active Expired - Lifetime
- 2001-12-13 KR KR1020037008754A patent/KR100860923B1/ko active IP Right Grant
-
2002
- 2002-11-14 US US10/451,534 patent/US7042809B2/en not_active Expired - Lifetime
-
2004
- 2004-06-10 HK HK04104194A patent/HK1061098A1/xx not_active IP Right Cessation
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2538581A1 (fr) * | 1982-12-27 | 1984-06-29 | Casio Computer Co Ltd | Appareil de reconnaissance de caracteres |
WO1990015399A1 (fr) * | 1989-06-08 | 1990-12-13 | Midhat Gazale | Dispositif electromecanique de reconnaissance de caracteres executes manuellement |
EP0632401A2 (fr) * | 1993-06-28 | 1995-01-04 | International Business Machines Corporation | Procédé et appareil de reconnaissance en ligne et en temps réel de caractères manuscits avec des traits |
EP0695979A1 (fr) * | 1994-08-02 | 1996-02-07 | Asulab S.A. | Dispositif d'identification d'un caractère transcrit sur une surface et procédé d'identification d'un tel caractère, notamment pour une montre |
Also Published As
Publication number | Publication date |
---|---|
KR100860923B1 (ko) | 2008-09-29 |
KR20030066774A (ko) | 2003-08-09 |
EP1350218B1 (fr) | 2014-03-26 |
CN1484810A (zh) | 2004-03-24 |
EP1220140A1 (fr) | 2002-07-03 |
US20040042346A1 (en) | 2004-03-04 |
EP1350218A1 (fr) | 2003-10-08 |
US7042809B2 (en) | 2006-05-09 |
CN1252634C (zh) | 2006-04-19 |
HK1061098A1 (en) | 2004-09-03 |
JP4122226B2 (ja) | 2008-07-23 |
JP2004528620A (ja) | 2004-09-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2002052493A1 (fr) | Procede et dispositif de reconnaissance de caracteres traces manuellement sur une zone de saisie | |
EP2235614B1 (fr) | Circuit electronique d'analyse a alternance de mesure capacitive/ resistive pour capteur tactile multicontacts a matrice passive | |
EP2310932B1 (fr) | Procédé d'acquisition et d'analyse d'un capteur tactile multicontacts suivant un principe dichotomique, circuit électronique et capteur tactile multicontacts mettant en oeuvre un tel procédé | |
WO2011010037A1 (fr) | Procede et dispositif de localisation d'au moins un toucher sur une surface tactile d'un objet | |
EP3109797A1 (fr) | Procédé de reconnaissance d'écriture manuscrite sur une surface physique | |
FR2925716A1 (fr) | Circuit electronique d'analyse a modulation de caracteristiques de balayage pour capteur tactile multicontacts a matrice passive | |
FR2925715A1 (fr) | Circuit electronique d'analyse a alternance axe d'alimentaiton/axe de detection pour capteur tactile multicontacts a matrice passive | |
EP2671140B1 (fr) | Procede et dispositif d'acquisition de donnees d'un capteur tactile matriciel multicontacts | |
EP0094867B1 (fr) | Procédé et dispositif pour la saisie et le traitement de données tracées en clair | |
EP2988249B1 (fr) | Procédé de détermination, dans une image, d'au moins une zone susceptible de représenter au moins un doigt d'un individu | |
EP2742409A1 (fr) | Procédé de caractérisation de toucher sur un écran tactile | |
EP3252563B1 (fr) | Détermination d'un contexte de mobilité d'un utilisateur porteur d'un équipement muni de capteurs inertiels | |
EP2474937A1 (fr) | Procédé d'authentification électronique d'une signature manuscrite, module et programme d'ordinateur correspondants | |
EP0993650A1 (fr) | Procede de segmentation et de reconnaissance d'un document, notamment d'un plan technique | |
CH686601B5 (fr) | Dispositif d'identification d'un caractère transcrit sur une surface et procédé d'identification d'un tel caractère, notamment pour une montre. | |
WO2024146995A1 (fr) | Système de saisie tactile d'un code à destination des personnes malvoyantes | |
WO1979000363A1 (fr) | Dispositif pour mesurer des parametres caracteristiques de la vitesse d'une ecriture manuscrite | |
EP2895939A1 (fr) | Systeme et procede de detection de l'actionnement d'une fonction materialisee par un trajet predetermine sur une surface tactile d'un objet | |
FR2878343A1 (fr) | Dispositif d'entree de donnees | |
FR3017470A1 (fr) | Procede de saisie sur un clavier numerique, interface homme machine et appareil mettant en œuvre un tel procede | |
EP0892360A1 (fr) | Procédé de reconnaissance de caractères manuscrits | |
FR3050540A1 (fr) | Procede d'identification des gestes elementaires d'un mouvement et systeme mettant en œuvre un tel procede | |
FR3041446A1 (fr) | Bouton de selection a detection multiple de doigts |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CR CU CZ DE DK DM DZ EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ PL PT RO RU SD SE SG SI SK SL TJ TM TR TT TZ UA UG US UZ VN YU ZA ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 2001271938 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 10451534 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2002553717 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1020037008754 Country of ref document: KR Ref document number: 018215815 Country of ref document: CN |
|
WWP | Wipo information: published in national office |
Ref document number: 1020037008754 Country of ref document: KR |
|
WWP | Wipo information: published in national office |
Ref document number: 2001271938 Country of ref document: EP |
|
REG | Reference to national code |
Ref country code: DE Ref legal event code: 8642 |