JP6275840B2 - How to identify handwritten characters - Google Patents

Description

The present invention relates to a heat-menu man computer interactor sucrose emissions technology, and more particularly to identification methods and systems for handwriting.

With the development of computer visual applications, the demand for human computer interaction technology is increasing widely and demands are increasing. Handwriting recognition technology is also part of human computer interaction . Many handwritten character recognition systems perform identification after the user finishes writing all strokes of a character, but this type of system is slow in inputting characters with a large number of strokes. It's difficult to input all strokes of complex characters. In order to solve this problem, several identification methods for predicting handwritten characters have been proposed. The characters can be identified by inputting a part of the strokes of the characters. Some of these methods are based on a character prediction database, and the system is given one based on the frequency of use of all characters stored and stored in a database containing a given stroke. A character candidate including one or a plurality of strokes is provided. In another type of method, a prediction result is output when a user inputs a radical of a character. Similarly, this type of method also provides character candidates based on frequency of use.

  Patent Document 1 entitled “Rapid Input Method Based on Incomplete Identification of Words” proposes a method for identifying handwritten characters for generating prediction candidates based on a prediction rate. The method generates a complete template and a series of derived sub-templates for each Kanji character. The derived sub-template may be the character head of a Chinese character or any other incomplete part. Each sub-template has a “completeness weight” based on its relative level of completeness relative to the complete Kanji character. Template matching is performed between the input stroke and all sub-templates for each Kanji character, and then the obtained matching rate is multiplied by the integrity weight of the corresponding sub-template, followed by the weighted matching rate. Is the prediction rate of complete kanji characters, and finally, a prediction candidate is represented based on the prediction rate.

Chinese patent application no. 201210284415 gazette

  However, it is inaccurate for the conventional predictive handwritten character identification method to display the character candidates based on the frequency of use. For example, when the user inputs a non-usual character, the system cannot obtain a correct identification result from the prediction candidates because the non-usual character is used less frequently.

  The problem of Patent Document 1 is that no clear rule is set for how to measure the “integrity weight”. The “integrity weight” is very important for producing a correct prediction candidate in the method. However, the patent application does not describe how to measure and determine the weight. Defining rational rules for the weight measurement is ambiguous and complex, but this has a huge impact on the prediction effect. The patent application also proposes dividing the complete template for each Kanji character into a plurality of first-order derived templates. The primary derivation template is further divided into a number of secondary derivation templates, and a new derivation template or the like can be generated by combining several primary derivation templates and some secondary derivation templates. According to this type of scheme, the quantity of all Kanji character templates becomes very large, which requires a large storage space to store all the templates of a large character set, and the template matching process Becomes very cumbersome and complicated to match all templates.

  Accordingly, an object of the present invention is to provide a method capable of accurately predicting a handwritten input result.

  The handwritten character identification method and system according to the present invention proposes an identification method and system for predicting a handwritten character by a part of a stroke, and can predict prediction candidates accurately, so that a user can write a part of the character by handwriting When the stroke is input, the character to be input can be identified.

One aspect of the present invention proposes a handwritten character identification method. The method first receives a handwritten trajectory input by the user, and then matches the handwritten trajectory to at least one stroke template to determine the degree of matching of the stroke template. Finally, according to the degree of matching, a standard character corresponding to the matched stroke template is output . Before SL stroke template is a standard character matching templates, matching templates one standard character even without least one且, wherein the full stroke templates of at least one standard character, incomplete stroke standard of the at least one standard character has a incomplete strokes template character, the corresponding incomplete stroke standard character before Symbol least one standard character incomplete stroke template does not constitute a part or all of any other standard character .

  Preferably, the matching and output steps are executed each time one stroke is input.

  Preferably, at least one stroke template matched with the handwriting trajectory is all matching templates of all standard characters.

  Preferably, the step of matching the handwriting trajectory with at least one stroke template to determine the degree of consistency of the stroke template further includes: the number of strokes already input and the stroke count of a standard character corresponding to the stroke template to be matched. And weighting the degree of matching based on the difference.

  Preferably, a standard character corresponding to the stroke template having the highest degree of matching is further displayed on the background of the handwriting locus.

Preferably, the incomplete stroke template is
Defining basic strokes and radicals;
Assigning one index number for each basic stroke and each stroke / radical in the radical;
Displaying each standard character using an index number corresponding to each stroke / radical of the standard character and an index number sequence obtained according to the stroke order combination of the standard character;
Ordering the standard character index number sequence,
Identifying index numbers that differ between the at least one standard character and other standard characters;
Determining the unique index number sequence of the at least one standard character, which is the index number sequence of the portion of the at least one standard character from the starting index number to the different index number in the index number sequence;
Generating incomplete stroke standard characters corresponding to the unique index number sequence to be incomplete stroke standard characters of the at least one standard character;
A step of, based on the full stroke sample before Symbol least one standard character, acquires an incomplete stroke samples of the standard character corresponding to said incomplete stroke standard characters,
Obtaining an incomplete stroke template of the at least one standard character based on the incomplete stroke sample of the at least one standard character.

  Preferably, the unique index number sequence further has an index number after the different index number.

  Preferably, when a stroke / radical corresponding to the different index number has at least two strokes, an incomplete stroke standard character corresponding to the unique index number sequence is further set to one of the different index numbers from the start index number. Incomplete stroke standard having strokes / radicals corresponding to the index number sequence of the portion up to the previous index number, and in addition each stroke / radical stroke corresponding to the different index number Character.

  Preferably, from the acquired incomplete stroke template, the incomplete stroke template whose stroke / radical corresponding to the different index number is one of the specific strokes / radical is discarded.

  Preferably, the specific stroke / radical may be one of the following letter N and the following letter O.

Preferably, the at least one standard character incomplete stroke template further comprises:
Determining at least two standard characters, the same stroke portion being each incomplete stroke portion;
The same stroke portion as an intermediate standard character, and for the intermediate standard character, generating an incomplete stroke template;
The generated incomplete stroke template and the stroke template corresponding to the intermediate standard character as an incomplete stroke template for each standard character in the at least two standard characters, and a template generated by Have.

  Preferably, the standard character may be one of Chinese, Japanese, and Korean.

According to another aspect of the present invention, a handwritten character identification system is further proposed, a handwriting input unit for receiving a handwritten locus input by a user, a template database for storing and storing a standard character matching template, and a handwritten locus Is matched with at least one template, and a template matching unit arranged so as to determine the matching degree of the template and a standard character corresponding to the matched template are output according to the matching degree determined by the template matching unit. has an output unit arranged to, one imperfect standard character matching template is incomplete strokes standard character of the at least one of the standard character full stroke template said at least one standard character even without least Stroke ten And a rate, incomplete strokes standard characters incomplete stroke template corresponding prior Symbol least one standard character, does not constitute a part or all of any other standard characters.

  According to the embodiment of the present invention, for a Chinese, Japanese or Korean character, each character is displayed by a series of “stroke / radical index numbers”, and each character's “unique index (stroke / radical)” ”Is recorded as a unique stroke or radical for displaying the character. Then, an incomplete stroke template of characters is generated based on the “unique index”. In this way, when the user inputs one or more strokes of one character, template matching between the input stroke and the incomplete stroke template is executed, and the identification result is accurately acquired. The

  Unlike the prior art, according to the handwritten character identification method and system of the present invention, a prediction candidate is issued based on the “unique index” of the character, so that the prediction candidate for each character is not confused with other characters. . Moreover, if the user inputs one or more strokes of the character, the character is identified. In this way, the work burden is reduced and the input efficiency is improved.

  The above and other objects, features, and advantages of the present invention will be further clarified by describing preferred embodiments of the present invention in conjunction with the accompanying drawings.

1 is a block diagram illustrating a handwritten character identification system according to an embodiment of the present invention. 3 shows a flowchart of a method for generating an incomplete stroke template according to an embodiment of the present invention. 3 shows an example of basic strokes and radicals according to an embodiment of the present invention. A part of the “stroke index table” according to the embodiment of the present invention is shown. The example of the character stroke information arrangement table by the Example of this invention is shown. The same index number and unique index number in the character stroke information arrangement table shown in FIG. 5 are shown. An exercise sample of a complete stroke character of "Whip" is shown. FIG. 8 shows an exercise sample of an incomplete stroke template of “whipping” obtained from the complete stroke template of “whipping” shown in FIG. 7. 5 shows a flowchart of a handwritten character identification method according to an embodiment of the present invention. FIG. 5 shows one example of the operation result of the handwritten character identification method according to the embodiment of the present invention, where (a) shows a handwritten locus inputted by a user, and (b) shows an incomplete stroke template that highly matches the handwritten locus (C) shows that the identified character is output in the background of the handwritten trajectory. The example of the incomplete stroke template and the perfect stroke template of the following character D by the Example of this invention is shown. An example of intermediate character construction according to an embodiment of the present invention is shown. One stroke sample for illustration is shown.

  Reference will now be made in detail to the exemplary embodiments of the present invention, examples of which are illustrated in the accompanying drawings. In the following description, some specific examples are used for descriptive purposes only and should not be understood as any limitation to the present invention. It is merely an example of the present invention. Where the understanding of the present invention may be chaotic, normal mechanisms or structures are omitted.

図１は本発明実施例による手書き文字識別システム１００を説明したブロック図を示している。該システム１００は、手書き入力ユニット１１０、テンプレートデータベース１２０、テンプレートマッチングユニット１３０、出力ユニット１４０、の４ユニットを有する。各ユニットの機能は以下に記述するとおりである。

FIG. 1 is a block diagram illustrating a handwritten character identification system 100 according to an embodiment of the present invention. The system 100 includes four units of a handwriting input unit 110, a template database 120, a template matching unit 130, and an output unit 140. The function of each unit is as described below.

  Handwriting input unit 110: This unit is used to receive handwritten input data from the user. For example, the unit may be a sensor that identifies contact between the user's finger and the touch pad when the user inputs handwriting into the touch screen electronic device. The unit may be an infrared sensor when the user inputs to the electronic device using infrared or the like.

  Template database 120: The unit is used to store and store matching templates for each standard character. According to the embodiment of the present invention, the standard character complete stroke template and the incomplete stroke template are stored in the template database. First, incomplete stroke standard characters are defined for each standard character. Some complex standard characters may have multiple incomplete stroke standard characters. However, an incomplete stroke standard character of one standard character does not constitute a part or all of any other standard character. Thereafter, exercises are performed on incomplete stroke standard characters and complete stroke standard characters for each standard character, and a complete stroke template and an incomplete stroke template for each standard character are acquired. A method for generating an incomplete stroke template will be described in detail later in connection with FIG.

Template matching unit 130: The unit matches the handwritten input data to an incomplete stroke template and / or a complete stroke template to determine the matching degree of the matching template.

  Output unit 140: This unit outputs the identification result to the user. According to one embodiment, the unit can output standard characters corresponding to the template with the highest degree of matching. The unit can also perform optimization of any matching result and select to order candidates based on a predetermined rule. For example, for an electronic device of the mobile phone type with a limited display screen size, the output unit 140 outputs only the top 10 matching results with a high degree of matching. In one embodiment, for example, the data may be output in order according to the degree of matching.

  FIG. 1 shows that the system may further include a preprocessing unit 150.

  Preliminary processing unit 150: This unit is used to perform processing such as smoothing, straight line standardization and / or non-straight line standardization on handwritten input data. The purpose of the unit is to make handwritten input data smoother and to make matching within the template matching unit 130 possible.

  The key technology of the present invention is the definition of an incomplete stroke template. For characters such as Chinese, Japanese, Korean, etc. (ie, kanji, Japanese characters, etc.), each character is assembled with a stroke and / or radical. Different characters may have the same stroke and radical. However, each character always includes a unique stroke or radical order different from other characters. According to the embodiment of the present invention, an incomplete stroke template for each character is defined using the unique stroke and the radical. That is, the corresponding incomplete stroke character has the unique stroke / group order so that it does not constitute part or all of any other character, that is, can be distinguished from the other character. Secondly, an incomplete stroke template of characters is constructed.

  Take the Chinese character of GB2312 character set as an example. FIG. 2 shows a flowchart of an incomplete stroke template generation method 200 according to an embodiment of the present invention.

  First, in step S210, a basic stroke and a radical are defined as shown in FIG.

Based on the definition of Unicode, there are a total of 36 basic strokes. Some strokes are discarded and 26 basic strokes are finally left as shown in FIG. The first five lines of the radical in FIG. 3 are user-defined, and the last two lines of the radical in FIG. 3 are basic radicals in GB 2312. FIG. 3 shows a total of 152 radicals, meaning that these radicals are the basic strokes / radicals defined by the Chinese character GB2312 in the embodiment of the present invention. Of course, the radicals applicable to the present application are not limited to those shown in FIG. 3, and other radical sets can be used as needed. The required radicals are not the same for kanji, Korean or Japanese.

  In order to perform indexing, in step S220, a “stroke index table” is defined for the defined basic strokes and radicals, and one index number is assigned to each stroke / radical in the basic strokes and radicals. . FIG. 4 shows a part of the “stroke index table”. When each basic stroke and radical are indexed, each character can be displayed by a series of index numbers obtained according to the index number / stroke order combination corresponding to each stroke / radical of the character.

  Thereafter, in step S230, each character is displayed using an index number corresponding to each stroke / radical of the character and an index number sequence obtained according to the stroke order combination of the character, and the index of each character. Numbers are arranged in order, and character stroke information is obtained. Therefore, a table in which the stroke information for each character is arranged in order is defined. FIG. 5 shows a part of the table in an explanatory manner.

  Thereafter, in step S240, the unique stroke / radical of each character is found, and “unique index number order” is acquired. That is, first, an index number different from other characters is identified for each character. FIG. 6 shows the same character and stroke information arrangement table as that in FIG.

  As can be seen from the table in FIG. 6, the broken line on the left side includes a common stroke of the six characters shown. The middle broken line frame includes a common stroke of the previous five characters. Further, the broken line frame on the right side displays “晋” and the common radical of the character A. Finally, the last stroke / radical of the second, third, fifth, and sixth characters (whose index numbers are 211, 226, 233, and 201, respectively) display the unique stroke / radical of the corresponding character. ing. Therefore, the “unique index table” is acquired. That is, with respect to the letter B, the index number sequence is “1-3-7-7-1-211”, and “211” therein represents the unique stroke / radical. For “晋”, the index number sequence is “1-3-7-7-6-1-236”, but there is no unique stroke / radical. For the letter A, the index number sequence is “1-3-7-7-6-1236-233”, and “233” in the index number indicates the unique stroke / radical. For the letter C, the index number sequence is “1-3-7-7-201”, and “201” in the index C represents the unique stroke / radical.

  Thereafter, the unique index number order of the character is determined, and the unique index number order is the index number order of the part from the start index number to the different index number in the index number order of one character. As can be seen from FIG. 6, among the six characters, only the character B, the character D, the character A, and the character C have unique index numbers different from those of other characters.

  As another example, if the index number sequence of “Whiz” is “302-104-1-3-10-1-1-1-5-8”, for example, the third index number “1” in the index number “1” Represents the unique stroke / radical. The unique index number sequence of “Whip” is “302-104-1”, “302-104-1-3”, “302-104-1-3-10”, “302-104-1-3-10-”. 1 "," 302-104-1-3-10-1-1 "," 302-104-1-3-10-1-1-5 "," 302-104-1-3-10-1- " 1-5-8 ". In other words, in addition to “302-104-1”, the unique index number sequence further has an index number after the unique index number.

  In step S250, an incomplete stroke character for each standard character is generated based on the acquired “unique index number order”.

  In one embodiment, the stroke / radical corresponding to the unique index number may have at least two strokes. In the embodiment, the incomplete stroke character corresponding to the unique index number sequence includes the stroke / radical corresponding to the index number sequence of the portion from the start index number to the index number one before the unique index number, and In addition, it has incomplete stroke characters including strokes / parts of each stroke corresponding to the unique index number.

  In step S260, based on the incomplete stroke character for each character, the incomplete stroke sample is obtained from the complete stroke sample of the standard character, thereby obtaining the incomplete stroke template of the standard character.

FIG. 7 shows an exercise sample of a complete stroke character of “Whip”. Based on step S240, an index number sequence of “Whip”, for example, “302-104-1-3-10-1-1-5-8”, can be obtained, and the third index number among them is obtained. “1” represents the unique stroke / radical. Therefore, based on the step S250, the index number sequence of the incomplete stroke character of “Whip” is set to “302-104-1”, “302-104-1-3”, “302-104-1-3-10”, “ "302-104-1-3-10-1", "302-104-1-3-10-1-1", "302-104-1-3-10-1-1-5" Can do. FIG. 8 shows an incomplete stroke sample of “whipping” obtained from the complete stroke sample of “whipping” shown in FIG.

  FIG. 9 shows a flowchart of a handwritten character identification method 900 according to an embodiment of the present invention.

  As shown in FIG. 9, in step S910, a handwritten locus input by the user is received. In this step, the user inputs a handwriting locus using the handwriting input unit.

  Thereafter, in step 920, preliminary processing is performed on the handwritten trajectory. The purpose of the preliminary processing is to make the handwriting trajectory smoother and use subsequent operations.

  Thereafter, in step 930, the input handwriting trajectory and the stroke template are matched to determine the degree of matching of the stroke template. Finally, in step 940, standard characters corresponding to the matched stroke template are output. That is, a standard character corresponding to the handwritten locus is identified.

  According to an embodiment of the present invention, the system stores and stores incomplete stroke templates. Therefore, every time an input of one stroke is received, steps S920 to S940 can be executed to adjust the identification result. For example, when the user thinks to input the character D, when the character E has been input, in step S930, for example, the characters E, “ “巫”, the character D, “丑”, and “correct” are determined, and in step S940, the character E, “巫”, the character D, “丑”, and “correct” are sequentially output. However, when the user continues to input the subsequent two strokes, step S930 at this time sequentially changes the characters corresponding to the stroke template in accordance with the degree of matching, for example, the characters D, “悉”, “fear”, “love”. ”And“ 晋 ”are determined, and the characters D,“ 悉 ”,“ fear ”,“ love ”, and“ 晋 ”are sequentially output in step S940.

  In one embodiment, when a user inputs a stroke, the previous ten candidates among the input stroke identification results at the present time are output, and the most likely identification result on the background of the input stroke, that is, the degree of consistency is output. The highest character can be displayed. For example, the user inputs an incomplete stroke as shown in FIG. The handwritten character identification system according to the embodiment of the present invention matches the current input stroke with all incomplete stroke templates of all characters, and determines that the incomplete stroke template “whip” is the most likely candidate. Discover. As shown in FIG. Finally, the first identification result, “Whip” shown in FIG. 10C, is displayed in the background of the handwritten locus.

In some embodiments, to avoid incomplete strokes characters are conflation fully stroke character, discarding a number of incomplete strokes character similar to full stroke character, i.e. distinguished from full stroke character The incomplete stroke character whose specified portion is a specific stroke / radical is discarded. For example, the character D, which is a character, has a unique radical of 226 (heart), which means that it has three incomplete stroke characters, that is, the character G. However, the character H, which is the first incomplete stroke character, may be confused with the character E. This is because the character N, which is a stroke to be distinguished, may be erroneously identified due to noise or dirt on the input screen when the character E is input. Therefore, for the character D, which is a character, the first incomplete stroke character is discarded and only the other two incomplete stroke characters are left. FIG. 11 shows two incomplete stroke samples and complete stroke samples extracted from the character D. According to an embodiment of the present invention, the specific stroke / radical may be one of the character N and the character O.

  According to some embodiments of the present invention, in order to effectively predict a character, several intermediate characters are constructed to be the intersection of similar characters. For example, for “Awake”, the character F, and “醍”, the character I is the common part of these three characters. This means that the common part can be an intermediate character of the three characters, and an incomplete stroke template is generated for the intermediate character. For example, as in the template extracted from the sample shown in FIG. 13, the character K, the character L, and the character M are assumed to be incomplete stroke characters of “Awake”, the character F, and “醍”. In this way, when the user inputs a handwritten character, once the character J is input, “Awake”, the character F, and “醍” can be identified. FIG. 12 shows one example of intermediate character construction. As shown in FIG. 12, the common part in the index number sequence of “Awake”, the character F, and “、” is “293-236” and corresponds to the character M. An incomplete stroke template is generated for the common portion, and the incomplete stroke template generated in this way and the stroke template of the common portion are set as the incomplete stroke templates of “Awake” and the characters F and “醍”.

  According to the embodiment of the present invention, a table of the maximum number of strokes for identifying each standard character is further defined. When calculating the degree of matching, a different weight is given depending on the number of strokes so that the identification result can be adjusted based on the difference between the number of input strokes and the number of strokes of the matching template.

  The handwritten character identification method and system proposed by the embodiment of the present invention can identify characters such as Chinese, Japanese, and Korean.

Next, the performance of the handwritten character identification prediction system according to the present invention is evaluated using the identification rate and the number of input strokes. A system that has a high identification rate and that requires input of relatively few strokes is a high-performance system. Otherwise, walk has a low recognition rate is system that requires input of large amounts of stroke low performance system.

In order to calculate the prediction performance, the prediction ratio of each character is defined as follows:
PR = Number of strokes omitted / Total number of strokes (1)
In this, the “number of strokes omitted” is the difference between the “total number of strokes” and the number of strokes actually input by the user, and the “total number of strokes” is the total number of strokes of the standard character.

Based on Table 2, it can be seen that for characters having a large number of strokes, the handwritten character identification system and method according to the present invention can save a large amount of input strokes and input time. In this, “1 candidate correct” represents the identification result that the first candidate character was correct, and “previous 2 candidate correct” represents the identification result that the first or second candidate character was correct. Yes. The same applies to the following.

  The handwritten character identification method and system according to embodiments of the present invention can be applied to electronic whiteboards, tablet PCs, desktop PCs, mobile phones, PDAs, and other electronic devices that support handwriting input. When a user inputs to these devices using a finger or a stylus pen on the screen, the handwritten character identification method and system according to the embodiment of the present invention can output the identification result on the screen accordingly.

  Other settings of the embodiment of the invention disclosed herein include software programs that perform the steps and operations of the embodiment according to the method outlined above. More specifically, a computer program product comprises a computer readable medium on which computer program logic is encoded and when the computer program logic is executed on a computing device, the computer program logic provides related operations. In this way, the above technical means are provided. When executed on at least one processor of a computing system, computer program logic is capable of causing the processor to perform the operations (methods) described in the embodiments of the present invention. This type of setting of the present invention is typically software, code and / or other data structures on a computer readable medium such as an optical medium (eg, CD-ROM), floppy disk, hard disk, or the like, or Firmware or other medium of microcode, such as on one or more ROM or RAM or PROM chips, or application specific integrated circuits (ASICs), or downloadable software images in one or more modules, shared databases, etc. Provided as a setting or code in Software or firmware or such an arrangement can be mounted on the computing device to cause one or more processors in the computing device to perform the techniques described in the embodiments of the present invention. A software process coupled to a computing device, such as in a set of data communication devices or other entities, can also provide a device according to the present invention. The apparatus according to the present invention can be used for a plurality of software processes on a plurality of data communication devices, or for all software processes operating on a set of special purpose small computers, or for all software processes operating on a single computer. It can also be applied between.

  Strictly speaking, it should be understood that embodiments of the present invention can be implemented as a software program on software, software and hardware, or a single software and / or a single circuit.

  In the above description, the technical means of the present invention is shown only by way of example, but it should be noted that the present invention is in no way meant to be limited to the above steps and unit structures. It is. Where possible, steps and unit structures can be adjusted and discarded as needed. Thus, a step or unit is by no means a necessary element of the overall inventive concept implementing the present invention. Accordingly, the technical features necessary for the present invention are limited only to the minimum requirements of the entire inventive concept capable of realizing the present invention, and not limited to the specific examples described above. .

  Thus far, the present invention has been described in connection with the preferred embodiment. It should be understood by those skilled in the art that various other modifications, substitutions and additions can be made without departing from the spirit and scope of the present invention. Accordingly, the scope of the invention should not be limited to the specific embodiments described above, but should be limited by the appended claims.

Claims (11)

Receiving a handwritten trajectory input by the user;
Matching the handwriting trajectory to at least one stroke template to determine the consistency of the stroke template;
Outputting a standard character corresponding to the matched stroke template according to the degree of matching, comprising:
The stroke template is a standard character matching template, and
At least one standard character matching template comprises the at least one standard character complete stroke template and the at least one standard character incomplete stroke standard character incomplete stroke template, wherein the at least one standard The incomplete stroke standard characters supported by the incomplete stroke template of characters do not constitute part or all of any other standard characters,
The incomplete stroke template is
Defining basic strokes and radicals;
Assigning one index number for each basic stroke and each stroke / radical in the radical;
Displaying each standard character using an index number corresponding to each stroke / radical of the standard character and an index number sequence obtained according to the stroke order combination of the standard character;
Ordering the standard character index number sequence,
Identifying index numbers that differ between the at least one standard character and other standard characters;
Determining the unique index number sequence of the at least one standard character that is the index number sequence of the portion of the at least one standard character from the starting index number to the different index number in the index number sequence;
Generating incomplete stroke standard characters corresponding to the unique index number sequence to be incomplete stroke standard characters of the at least one standard character;
Obtaining an incomplete stroke sample of the standard character corresponding to the incomplete stroke standard character based on the complete stroke sample of the at least one standard character;
Obtaining an incomplete stroke template of the at least one standard character based on the incomplete stroke sample of the at least one standard character.   The method of claim 1, wherein the matching and outputting steps are performed each time a stroke is received.   The method of claim 1, wherein the at least one stroke template that matches the handwriting trajectory is all matching templates for all standard characters. Matching the handwriting trajectory with at least one stroke template to determine the consistency of the stroke template;
The method according to claim 1, further comprising the step of weighting the degree of matching based on a difference between the number of strokes already input and the number of strokes of a standard character corresponding to a stroke template to be matched.   The method according to claim 1, further comprising displaying a standard character corresponding to the stroke template having the highest degree of matching on the background of the handwriting trajectory.   The method of claim 1, wherein the unique index number order further comprises an index number after the different index number. When the stroke / radical corresponding to the different index number has at least two strokes, the incomplete stroke standard character corresponding to the unique index number sequence further includes:
A stroke / radical corresponding to the index number order of the portion from the start index number to the index number one before the different index number, and in addition, a stroke / radical stroke corresponding to the different index number The method according to claim 6, further comprising: an incomplete stroke standard character having two parts.   The method according to claim 1, wherein an incomplete stroke template whose stroke / radical corresponding to the different index number is one of the specific strokes / radicals is discarded from the acquired incomplete stroke templates. The specific stroke / radical is the following letter N and the following letter O

9. The method of claim 8, comprising: The incomplete stroke template of the at least one standard character further comprises:
Determining at least two standard characters, the same stroke portion being each incomplete stroke portion;
The same stroke portion as an intermediate standard character, and for the intermediate standard character, generating an incomplete stroke template;
The generated incomplete stroke template and the stroke template corresponding to the intermediate standard character as an incomplete stroke template for each standard character in the at least two standard characters, and a template generated by The method of claim 1, comprising: The method according to claim 1, wherein the standard character is one of Chinese, Japanese, and Korean .

Images