CN114677767A - Formula generation method and device, computing equipment and computer storage medium - Google Patents
Formula generation method and device, computing equipment and computer storage medium Download PDFInfo
- Publication number
- CN114677767A CN114677767A CN202210129247.8A CN202210129247A CN114677767A CN 114677767 A CN114677767 A CN 114677767A CN 202210129247 A CN202210129247 A CN 202210129247A CN 114677767 A CN114677767 A CN 114677767A
- Authority
- CN
- China
- Prior art keywords
- formula
- track point
- track
- matched
- points
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/90335—Query processing
- G06F16/90344—Query processing by using string matching techniques
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Character Discrimination (AREA)
Abstract
The embodiment of the invention provides a formula generation method, a formula generation device, a calculation device and a computer storage medium. The formula generation method comprises the following steps: identifying track points generated in response to the handwriting operation, and determining a formula to be matched corresponding to the track points; according to the formula to be matched, determining a complete formula corresponding to the formula to be matched from a pre-constructed formula library; and outputting the complete formula. The technical scheme provided by the embodiment of the invention can realize the technical effect of conveniently and quickly generating the complete formula for the user.
Description
Technical Field
The embodiment of the invention relates to the technical field of artificial intelligence, in particular to a formula generation method, a formula generation device, a formula generation calculating device and a computer storage medium.
Background
At present, more and more classrooms are equipped with large screens or intelligent blackboards and other devices, and in an actual teaching scene, users often need to perform various teaching demonstration on the large screens or the intelligent blackboards.
When a scene needing to write and call a formula, such as a chemical equation, is encountered, the input of the keyboard is complicated and inefficient, and part of the expression modes of the upper and lower corner marks are not familiar to users and are easy to miss.
Disclosure of Invention
The embodiment of the invention provides a formula generation method, a formula generation device, a calculation device and a computer storage medium.
In a first aspect, an embodiment of the present invention provides a formula generation method, including:
identifying track points generated in response to handwriting operation, and determining a formula to be matched corresponding to the track points;
according to the formula to be matched, determining a complete formula corresponding to the formula to be matched from a formula library constructed in advance;
and outputting the complete formula.
In a second aspect, an embodiment of the present invention provides a formula generating apparatus, including:
the recognition module is used for recognizing track points generated in response to handwriting operation and determining a formula to be matched corresponding to the track points;
the matching module is used for determining a complete formula corresponding to the formula to be matched from a pre-constructed formula library according to the formula to be matched;
and the output module is used for outputting the complete formula.
In a third aspect, an embodiment of the present invention provides a computing device, including a processing component and a storage component;
the storage component stores one or more computer instructions; the one or more computer instructions are called and executed by the processing component to realize the formula generation method provided by the embodiment of the invention.
In a third aspect, an embodiment of the present invention provides a computer storage medium, which stores a computer program, and when the computer program is executed by a computer, the computer program implements a formula generation method according to the embodiment of the present invention.
In the embodiment of the invention, the track points generated in response to the handwriting operation are firstly identified, the complete formula corresponding to the handwritten formula is determined from the formula library by utilizing the identified formula to be matched, the track identification and the formula matching are connected, and the complete formula generated by matching can be determined by writing partial formulas by a user, so that the technical effect of improving the convenience of formula calling is realized.
These and other aspects of the invention are apparent from and will be elucidated with reference to the embodiments described hereinafter.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a flowchart of formula generation according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a formula generation apparatus according to an embodiment of the present invention;
FIG. 3 is a schematic structural diagram of a computing device according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of another computing device according to an embodiment of the present invention.
Detailed Description
In order to make those skilled in the art better understand the technical solution of the present invention, the technical solution in the embodiment of the present invention will be clearly and completely described below with reference to the drawings in the embodiment of the present invention.
In some of the flows described in the present specification and claims and in the above figures, a number of operations are included that occur in a particular order, but it should be clearly understood that these operations may be performed out of order or in parallel as they occur herein, with the order of the operations being indicated as 101, 102, etc. merely to distinguish between the various operations, and the order of the operations by themselves does not represent any order of performance. Additionally, the flows may include more or fewer operations, and the operations may be performed sequentially or in parallel. It should be noted that, the descriptions of "first", "second", etc. in this document are used for distinguishing different messages, devices, modules, etc., and do not represent a sequential order, nor limit the types of "first" and "second" to be different.
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
At present, more and more classrooms are equipped with large screens or intelligent blackboards and other devices, and in an actual teaching scene, users often need to perform various teaching demonstration on the large screens or the intelligent blackboards.
When a scene needing to write and call a formula, such as a chemical equation, is encountered, the input of the keyboard is complicated and inefficient, and part of the expression modes of the upper and lower corner marks are not familiar to users and are easy to miss.
In the related art, a trajectory-based online modality recognition method and an image-based offline modality recognition method are generally adopted to recognize and generate formulas.
The track-based online modal identification method is characterized in that a track point sequence collected in the handwriting process of a user is subjected to feature extraction, and then handwriting identification is carried out on a time sequence, so that accurate identification is extremely difficult in the case of repeated drawing of the same stroke.
In the related art, a handwriting recognition framework for recognizing handwriting data is usually based on a three-stage framework of segmentation and combined re-recognition, and when continuous handwriting such as continuous cursive handwriting and continuous running handwriting is recognized, because no pen break exists in the continuous handwriting, the probability of incorrect segmentation of a segmentation module during segmentation of the continuous handwriting is high, and the recognition accuracy is low.
The offline modal recognition method based on the image is to perform feature extraction on the image obtained by handwriting of a user, and then perform handwriting recognition from space, and errors are easily recognized even under the condition that space is overlapped.
An embodiment of the present invention provides a formula generation method, where the formula generation provided in the embodiment of the present invention may be performed by an electronic device, and the electronic device may be a terminal device with a display screen and data processing capabilities, such as a PC, a notebook computer, a smart phone, and an electronic whiteboard
Fig. 1 is a flowchart of a formula generation method according to an embodiment of the present invention, and as shown in fig. 1, the formula generation method may include the following steps:
and 101, identifying track points generated in response to the handwriting operation, and determining a formula to be matched corresponding to the track points.
According to the embodiment of the invention, the handwriting operation can be triggered by the user touching the display screen of the electronic device by hand or sliding on the display screen of the electronic device by using a mouse, a touch pen or other tools.
According to the embodiment of the invention, the display screen of the electronic equipment can display the sliding track of the user on the display screen in real time.
According to the embodiment of the invention, after the handwriting operation is executed, the user can trigger the track point identification operation in a pause mode.
According to the embodiment of the invention, the hand-drawn track generated by the hand-drawing operation can be divided into at least one group of strokes by starting and dropping the strokes, the position of the initial point of the track and the position of the terminal point of the track after pause are taken as the initial position and the terminal position of one stroke, and the track point written between the two moments is the track of the stroke. The data of each track point comprises the state and the position information of the track point at the moment, namely whether the track point is a starting point or not and absolute coordinates on a screen.
According to the embodiment of the invention, after all track points included in the track generated by the handwriting operation are divided according to the strokes, the track points generated in response to the handwriting operation can be identified based on the strokes, and the formula to be matched corresponding to the track points is determined.
According to the embodiment of the invention, the track points are identified by using an algorithm such as an artificial neural network and machine learning.
According to embodiments of the invention, the trajectory points may be determined from the handwritten trajectory at time intervals or distance intervals.
102, according to the formula to be matched, determining a complete formula corresponding to the formula to be matched from a formula library constructed in advance.
And 103, outputting the complete formula.
According to the embodiment of the invention, the common formulas of all disciplines or specific disciplines can be collected in advance according to the prior knowledge, and the generation formula library is constructed. For example, for the mathematical discipline, all formulas related to each learning stage of junior high school mathematics, high school mathematics and the like can be acquired, and a mathematical discipline formula library can be constructed and generated. For another example, for a chemical discipline, all chemical equations involved in learning stages of junior high school chemistry, high school chemistry and the like can be obtained, and a chemical discipline formula library can be constructed and generated.
According to the embodiment of the invention, after the formula to be matched corresponding to the handwriting operation is identified, the complete formula corresponding to the formula to be matched can be determined from the pre-constructed formula library in a manner of conveniently matching the formula to be matched with the formula in the formula library.
According to the embodiment of the present invention, after the to-be-matched formula is identified and generated, confirmation information of the matching operation may be output, and in response to a confirmation operation initiated by a user for the confirmation information, the matching operation shown in step 102 may be performed to determine a complete formula corresponding to the to-be-matched formula from a formula library constructed in advance, and output the complete formula.
According to an embodiment of the present invention, the formula to be matched may be at least part of formula elements of a complete formula, that is, the complete formula contains all formula elements included in the formula to be matched, and the formula to be matched is only a component of the complete formula.
In some possible application scenarios, for a more complex formula, a user may write only a part of a complete formula by handwriting, for example, for a chemical equation, only a part or all of compounds before an equal sign may be written, in this case, in order to simplify the complexity of writing, an incomplete formula written by the user may be directly identified, a formula to be matched is determined, and the complete formula is determined from a formula library by using the incomplete formula to be matched.
According to the embodiment of the invention, as the formula to be matched is an incomplete formula, a plurality of formulas can be matched with the formula to be matched, and therefore, after the matching operation is finished, a plurality of complete formulas can be output, so that a user can select from the plurality of complete formulas. Wherein the plurality of complete formulas may be ordered based on relevance to the formula to be matched.
In the embodiment of the invention, the track points generated in response to the handwriting operation are firstly identified, the complete formula corresponding to the handwritten formula is determined from the formula library by utilizing the identified formula to be matched, the track identification and the formula matching are connected, and the complete formula generated by matching can be determined by writing partial formulas by a user, so that the technical effect of improving the convenience of formula calling is realized.
According to an embodiment of the present invention, step 102 may be specifically implemented as:
determining the number of characters of a formula to be matched;
intercepting standard formulas in a formula library according to a preset rule to determine a plurality of candidate formulas with lengths corresponding to the number of characters;
respectively calculating the similarity between the formula to be matched and a plurality of candidate formulas;
determining a target formula from the plurality of candidate formulas according to the similarity;
and taking a standard formula corresponding to the target formula as a complete formula.
According to the embodiment of the invention, in order to improve the matching efficiency, the identified formula may be filtered first, for example, chinese characters and latex format characters in the formula are filtered, and a formula to be matched, which only retains numbers, letters and special symbols, is generated.
According to an embodiment of the present invention, the preset rule may be, for example, a cut from left to right of the standard formula. For example, when the number of characters of the formula to be matched is 5, for the standard formula, 1 to 5 formula elements from the left may be retained, and all formula elements on the right side of the 5 th formula element may be deleted to generate the candidate formula.
According to an embodiment of the present invention, in a case that the standard formula is an equation, the preset rule may be, for example, to intercept formula elements before the equal sign of the standard formula from left to right.
According to the embodiment of the present invention, the number of characters of the candidate formula may not be completely equal to the number of characters of the formula to be matched, and at this time, the preset rule may be, for example, randomly generating an integer, adding the integer and the number of characters to generate an intercept number, and intercepting the standard formula by using the intercept number.
According to the embodiment of the invention, the matching result is more accurate under the condition that the number of characters of the candidate formula and the formula to be matched is more approximate, so that the randomly generated integer can be taken as a value between [ -2,2], for example, that is, the difference value between the number of characters of the formula to be matched and the number of characters of the candidate formula is two characters at most.
According to an embodiment of the present invention, the similarity may be calculated using, for example, an edit distance algorithm, a euclidean distance algorithm, a manhattan distance algorithm, a chebyshev distance algorithm, or the like.
According to the embodiment of the invention, after the plurality of similarity results are calculated and generated, the plurality of candidate formulas can be sorted in a descending order based on the similarity results, and the top N candidate formulas in the sorting result are taken as the target formulas. Wherein N is a positive integer, and a value of N can be flexibly selected by a person skilled in the art based on actual application requirements, for example, N can be 2, 3, 4, 5, 10, and the like.
According to the embodiment of the invention, after the plurality of similarity results are calculated and generated, only the plurality of similarity results of which the similarity results are greater than the preset threshold value can be sorted in a descending order.
According to the embodiment of the invention, under the condition that the user only writes a few formula elements, the matching is directly performed according to the identified formula to be matched at the moment, more complete formulas may be matched, and in this case, the user selects the expected formula from the more complete formulas more complicatedly. Thus, the formula to be matched may be augmented where the user has written only a few formula elements.
According to the embodiment of the present invention, in a case that the number of characters of the formula to be matched is smaller than a preset character threshold, before intercepting the standard formula in the formula library according to a preset rule, the formula generating method further includes:
and expanding the formula to be matched so that the number of characters of the expanded formula to be matched is greater than or equal to a preset character threshold value.
According to the embodiment of the invention, the formula to be matched can be expanded in a mode of randomly acquiring characters from the dictionary and adding the characters to the formula to be matched.
According to the embodiment of the invention, if the number of characters of the formula to be matched is less than the preset character threshold value, it indicates that the formula elements written by the user are fewer, for example, the user only writes C, O two formula elements, if the formula to be matched is directly used for matching with the standard formula in the formula library, a large number of standard formulas are extracted, so that not only is the time for matching the formula increased, but also the time cost for the user is correspondingly increased when the user searches for the desired formula from the large number of standard formulas.
According to the embodiment of the invention, by expanding the formula to be matched, the time length for matching the subsequent formula can be reduced, and the time cost for selecting the expected formula from the matching result by the user is correspondingly reduced.
According to the embodiment of the invention, because the formula is usually written based on a certain rule, although the formula to be matched is expanded by randomly adding characters, some expanded formulas to be matched which do not conform to the formula writing rule can be filtered in the formula matching process, namely, the expanded formulas to be matched which do not conform to the formula writing rule cannot be matched with the corresponding standard formula from the formula library, so that the rest expanded formulas to be matched which conform to the formula writing rule can be matched with the corresponding standard formula from the formula library and output to a user, and the probability that the user finds the expected formula from the matching result is increased.
According to other embodiments of the present invention, for the extended formula to be matched, the standard formula including all elements of the extended formula to be matched may be first obtained from the formula library, and then the obtained standard formulas are sorted. Specifically, the sorting process may be: firstly, determining the character interval number of the expanded formula to be matched from the first character to the last character, and then sequencing the obtained standard formula based on the character interval number.
For example, the augmented formula to be matched may be, for example, an FCO, that is, a user writes only F, C, O formula elements, so that all standard formulas containing F, C, O can be obtained from the formula library, and on this basis, in order to improve the accuracy of matching the standard formulas, the appearance order of the formula elements in the standard formulas obtained from the formula library may be the same as that of the formula to be matched, specifically, in this example, the appearance order of the formula elements in the formula to be matched is F, C, O in sequence, so that all standard formulas containing F, C, O and having an appearance order of F, C, O in sequence can be obtained from the formula library.
After a plurality of standard formulas corresponding to the formulas to be matched are obtained, the obtained standard formulas can be sequenced. Specifically, the number of character intervals of the formula to be matched from the first character to the last character, in this example, from the first character F to the last character O, may be determined first. Therefore, the number of the character intervals to be sorted from the character F to the character O in the obtained standard formulas can be determined, and then the standard formulas are sorted according to the difference value between the number of the character intervals to be sorted and 1. For example, the standard formula may be 3FCO 2Then the character F is separated from the character O by one character, and for example, the standard formula may be 2FClO, then the character F is separated from the character O by two characters, in this case 3FCO2May be ordered before 2 FClO.
In the standard formula, if the same formula element appears multiple times, the minimum interval number can be used as the interval number to be sorted. For example, the standard formula may be FCO + SO, since the same formula element O appears twice, where the interval between the formula element F and the formula element O appearing for the first time is 1, and the interval between the formula element O appearing for the second time is 4, the number of intervals to be sorted of the standard formula may be determined to be 1.
After the number of the intervals of the characters to be sorted is determined, if two or more standard formulas with the same number of the intervals of the characters to be sorted exist, the total number of the characters of the standard formulas can be determined, and the standard formulas are sorted according to the total length of the characters. For example, the plurality of standard formulas may be ordered in ascending order of total number of characters. For example, standard equation 1 may be FCO + H2O, Standard equation 2 may be FCO + H2CO3Since the total number of characters of the standard formula 1 is 7, which is smaller than the total number of characters 9 of the standard formula 2, the standard formula 1 can be sorted earlier than the standard formula 2. But not limited to, the plurality of standard formulas may also be sorted in descending order of the total number of characters.
According to an embodiment of the invention, the formulas in the formula library are typically standard formulas found in textbooks, however, the user may not write in the manner in which the standard formulas are written. For example, for a chemical equation, the standard formula is C + O2＝CO2However, the user may not write in this format, but as O2+C＝CO2. At this time, since the order of appearance of the compounds in the equation written by the user is different from that of the standard equation, the calculation time and the calculation amount of the matching operation are increased.
Therefore, the formula library can be correspondingly expanded so as to improve the speed, the efficiency and the accuracy of formula matching.
According to an embodiment of the invention, the formulas in the formula library comprise equations;
the formula library is constructed and generated by the following operations:
rearranging formula elements before the equal sign in the original equation obtained from the original formula library; and/or
Rearranging formula elements after the equal sign in the original equation obtained from the original formula library;
and generating a formula library by using the formulas generated after rearrangement.
According to the embodiment of the invention, a plurality of conversion formulas which are substantially the same as the standard formula but have different forms can be generated by arranging and combining the formula elements in the standard formula.
According to the embodiment of the invention, after the plurality of conversion formulas are generated based on the standard formula, the mapping relation between the plurality of conversion formulas and the standard formula can be established.
According to the embodiment of the invention, after the matching operation is completed, only one of the multiple complete formulas can be reserved to simplify the selection items of the user under the condition that the multiple complete formulas have the mapping relation with the same standard formula.
According to the embodiment of the present invention, step 101 may be specifically implemented as:
preprocessing the obtained track points to generate a track point matrix;
and inputting the track point matrix into a formula recognition model, and outputting a formula to be matched.
According to the embodiment of the invention, the formula recognition model can adopt an end-to-end track recognition model, and can be composed of a six-layer Time Delay Neural Network (TDNN), a four-layer bidirectional long-short term memory artificial neural network (Bi-LSTM) and a full connection layer (linear). The trace point matrix generated by preprocessing the trace points does not need to be subjected to additional segmentation and combination processing, and can be directly input into a formula recognition model to return a prediction result in an end-to-end mode.
According to the embodiment of the invention, after the acquired track points are preprocessed, a plurality of strokes can be reserved, and the total number of the track points is M, so that the hand-drawn track is characterized by a 4 × M track point matrix, 4 elements in each row of the track point matrix respectively represent an x coordinate and a y coordinate of the track point, whether the track point is a starting point or not and a time interval from the previous track point to the track point.
In the embodiment of the present invention, M may be 800, and when the number M of the trace points retained after the obtained trace points are preprocessed is less than 800, 0 may be used to complement to 800.
According to the embodiment of the invention, the preprocessing the acquired track points and the generating of the track point matrix comprises the following steps:
sequencing the track points by using the spatial characteristics of the track points to generate a first track point sequence;
according to the embodiment of the invention, in some application scenarios, for example, in the writing scenario of the chemical equation, the writing of the chemical equation is not strictly written in the order from left to right, but the chemical elements and compounds are written first, and then the formula is performed, and the process of the formula may also break the writing order from left to right. Therefore, in the embodiment of the present invention, after the writing operation of the user is completed, all the track points are reordered by using the spatial features of the track points, and the first track point sequence is generated.
According to the embodiment of the invention, when the track points are sequenced, the abscissa of the track points is taken as a first standard, and the ordinate of the track points is taken as a second standard for sequencing.
Specifically, an abscissa mean value may be first calculated for each stroke of the formula to be matched generated by the handwriting operation, and then the plurality of strokes of the formula to be matched are sorted by using a bubble sorting algorithm according to the abscissa mean value. When the mean values of the abscissas of at least two strokes are the same, the mean values of the ordinates of the two strokes can be respectively calculated, and the two strokes are sorted according to the mean values of the ordinates. The bubble sorting algorithm is the prior art, and is not described in detail in the embodiments of the present invention. According to the embodiment of the invention, the mean abscissa value can be calculated by dividing the sum of the abscissas of all the track points in a stroke by the number of the track points contained in the stroke.
According to the embodiment of the invention, the mean value of the vertical coordinates can be calculated by dividing the sum of the vertical coordinates of all track points in a stroke by the number of the track points contained in the stroke.
Carrying out weighted normalization on the first track point sequence to generate a second track point sequence;
and generating a track point matrix based on the second track point sequence.
According to the embodiment of the invention, the track point comprises state information, and the state information is used for representing whether the track point is a stroke termination point.
According to an embodiment of the present invention, the tracks generated by the handwriting operation may be saved as a track set G, G ═ S1，S2，…，SnIn which S isiTrace of ith pen representing handwriting, Si＝{Pi1，Pi２，…，PimIn which P isijIs the jth trace point coordinate of the ith pen, Pij＝{Ｘij，YijAnd f, flag and t, wherein x and y are click positions of the current user, the flag is used for representing whether the track point is a starting point, in the track points of the same stroke, t is the time interval between the current track point and the previous track point, and the starting point, namely the pen-down point t of the stroke, can be 0.
According to the embodiment of the invention, for example, when it is monitored that the user does not perform handwriting operation any more after the preset time, the completion of handwriting input of the user can be determined, all track points generated by the handwriting operation at this time can be obtained, and then all track points are divided according to strokes by using a flag value (for example, when the flag takes 1, the track point is represented as a starting point, and when the flag takes 0, the track point is represented as a non-starting point).
According to an embodiment of the present invention, the status information may include a flag value, for example.
The formula generation method further comprises:
dividing a plurality of track points in the second track point sequence into at least one track point set according to the state information, wherein the track points in each track point set belong to the same stroke;
according to the embodiment of the present invention, for example, the second track point sequence includes 10 track points, and the state information set of the second track point sequence may be {0, 0, 1, 0, 0, 0, 1, 0, 0, 1}, that is, the third, seventh, and tenth track points in the second track point sequence are starting points, respectively, so that the second track point sequence may be divided into three track point sets, where the first track point set is {0, 0, 1}, the second track point set is {0, 0, 0, 1}, the third track point set is {0, 0, 1}, and the first track point set, the second track point set, and the third track point set may represent one stroke, respectively.
Based on the second track point sequence, generating a track point matrix comprises:
determining whether a track point group meeting sparse sampling conditions exists in the track point set;
if the third track point sequence exists, sparse sampling is carried out on the track point group, and a third track point sequence is generated;
According to the embodiment of the invention, the sparse sampling condition can be determined by using the distance and the angle between the track points in the track point set. For example, for two adjacent track points in the same stroke, if the euclidean distance between the next track point and the previous track point is smaller than a preset distance threshold (for example, 0.25), and the next track point is not the termination track point of the stroke, the next track point is deleted. For adjacent three track points in the same stroke, if the included angle degree between the front two track point connecting line and the rear two track point connecting line is greater than a preset angle threshold (for example, 175 °), and the last track point is not the end point of the stroke, the middle track point can be deleted.
According to the embodiment of the invention, the track points input by handwriting of the user often comprise a plurality of redundant track points, so that the redundant track points can be removed by sparse sampling of the track point set, and the standard and easily-recognized track points are obtained.
Determining whether a track point set with the track point quantity smaller than a preset threshold exists in the third track point sequence;
and if the third track point sequence exists, inserting at least one track point into the track point set according to a preset rule until the number of the track points in the track point set is greater than or equal to a preset threshold value, and generating a track point matrix, otherwise, taking the third track point sequence as the track point matrix.
According to the embodiment of the invention, after sparse sampling, some simpler strokes may only have a few track points, if the generated track point matrix is directly input into a neural network, the strokes are easily filtered as noise, and distance-based resampling can be performed for the situation. For a stroke, if the number of the track points contained in the stroke is less than a threshold (for example, 8 points can be contained), connecting two adjacent points in the stroke, taking the middle point of the longest connecting line as a virtual track point to be inserted into the stroke, and repeating the operation until the number of the track points is greater than the threshold.
According to the embodiment of the invention, the weighted normalization of the first track point sequence is represented by the following formula (1):
wherein x isnewAbscissa, x, representing trace points generated after weighted normalizationoldRepresents the abscissa of the trace point before weighted normalization,
represents the abscissa weighted mean, sd represents the variance, ynewRepresenting the ordinate, y, of the trace points generated after weighted normalizationoldRepresents the ordinate of the trace point before weighted normalization,
which represents a weighted mean of the ordinate and,
xijthe abscissa representing the jth trace point in the ith stroke,
yijdenotes the ordinate, l, of the jth trace point in the ith stroke ijRepresents the weight of the jth track point in the ith stroke,
according to the embodiment of the invention, since the track points may be generated by taking points at equal time intervals in the handwritten track or may be generated by taking points at equal distance intervals in the handwritten track, an adaptive normalization algorithm as shown in formula (1) is adopted.
In the embodiment of the invention, in the formula (1), sd represents the variance in the y direction, and due to the written characteristics of the formula (usually represented by a single line), the heights of the formulas to be matched generated after scaling by equal proportion are kept nearly consistent after normalization by using sd.
According to the embodiment of the invention, when the track points are preprocessed, a weighting normalization mode is adopted, so that the prediction accuracy of the formula recognition model can be increased.
According to an embodiment of the present invention, in the case of modifying the handwritten formula, the formula generating method further includes:
acquiring a newly added track point set of the newly added strokes;
determining the overlapping rate of the newly added track point set and the existing track point set under the condition that the newly added track point set is overlapped with at least one existing track point set, wherein the track point set is used for expressing at least one stroke in a handwriting formula;
Deleting the existing track point set or the newly added track point set under the condition that the overlapping rate is greater than a preset overlapping threshold value; otherwise, combining the newly added track point set into the track point set.
According to the embodiment of the present invention, merging the newly added trace point sets into the trace point sets includes:
determining the overlapping track points of the newly added track point set and the existing at least one track point set;
deleting overlapped track points from the newly added track point set to generate a first track point set;
and merging the first track point set into the track point set.
According to the embodiment of the invention, when a user is unsatisfied with a stroke after writing the stroke, or unsatisfied with the matching result and intends to correct a handwriting track, the written track is not erased and rewritten, but the existing writing track is corrected.
In an embodiment of the present invention, it may be determined whether the user desires to overlay an existing track with a new track or desires to modify an existing track with a new track based on an overlap ratio of the new track and the existing track.
The difference between the two modification modes of covering and correcting is that under the condition that a user expects to cover the existing track by the newly added track, the track points corresponding to the existing track can be replaced by the track points corresponding to the newly added track, and under the condition that the user expects to modify the existing track by the newly added track, the track points corresponding to the newly added track can be merged into the track point set representing the same stroke of the existing track.
According to the embodiment of the invention, whether the newly added track point set is overlapped with the existing at least one track point set can be determined by calculating the distance between the track point in the newly added track point set and the track point in the existing at least one track point set. For example, when the distances from the trace points in the newly added trace point set to the trace points in the at least one existing trace point set by the preset percentage are all smaller than the preset distance threshold, the newly added trace point set may be determined to overlap with the trace point set, and the ratio of the number of the trace points in the newly added trace point set, which are smaller than the preset distance threshold, to the total number of the trace points in the newly added trace point set to the number of the trace points in the trace point set, which is smaller than the preset distance threshold, may be used as the overlap ratio.
According to the embodiment of the invention, when the overlapping rate is greater than the preset overlapping threshold, it can be determined that the user expects to cover the existing track with the newly added track, and the existing track point set or the newly added track point set can be deleted.
According to the embodiment of the present invention, in the case that the overlapping rate is smaller than the preset overlapping threshold, it may be further determined whether the user desires to modify the existing trajectory by using the newly added trajectory by:
For newly added track SK＝{PK1，PK2，…，PKnAnd existing trajectories Si＝{Pi1，Pi2，…，PimIf SKTo the existing trajectory SiIs smaller than a threshold value (e.g. may be 0.03), the trace point may be set as the existing trace SiThe a-th track point of (1), then the existing track SiPoint of medium minimum distance to existing track SiThe set of starting points can be recorded as Si0＝{Pia，Pia+1，…，PimIf point set Si0Middle 90% track point to stroke SKAre all less than the threshold and the point set Si0If the total number of the new tracks S is more than 5, determining that the user desires to utilize the new track SKFor the existing track SiModified, S can beKIs combined to SiIn, i.e. Si＝{Pi1，Pi2，…，Pia-1，PK1，PK2，…，PKnAnd will be SiAnd setting the flag of the corresponding track point of the middle starting point to be 0. Similarly, if new trace S is addedKTail point set and existing track SjThe head point sets of (1) are highly coincident, then SjRegarded as stroke SKIs continued.
Fig. 2 is a schematic structural diagram of a formula generation apparatus according to an embodiment of the present invention, and as shown in fig. 2, the apparatus includes: an identification module 201, a matching module 202 and an output module 203.
The recognition module 201 is configured to recognize trace points generated in response to a handwriting operation, and determine a formula to be matched corresponding to the trace points;
the matching module 202 is used for determining a complete formula corresponding to the formula to be matched from a pre-constructed formula library according to the formula to be matched;
And the output module 203 is used for outputting the complete formula.
According to an embodiment of the present invention, the matching module 202 includes a character number determination unit, a truncation unit, a similarity calculation unit, a target determination unit, and a complete formula determination unit.
The character number determining unit is used for determining the number of characters of the formula to be matched;
the intercepting unit is used for intercepting the standard formulas in the formula library according to a preset rule so as to determine a plurality of candidate formulas with the lengths corresponding to the number of characters;
the similarity calculation unit is used for calculating the similarity between the formula to be matched and the candidate formulas respectively;
the target determining unit is used for determining a target formula from the candidate formulas according to the similarity;
and the complete formula determining unit is used for taking the standard formula corresponding to the target formula as the complete formula.
According to an embodiment of the present invention, the matching module 202 further includes an expansion unit in case the number of characters is smaller than the preset character threshold.
And the expansion unit is used for expanding the formula to be matched so that the number of characters of the expanded formula to be matched is greater than or equal to a preset character threshold value.
According to an embodiment of the invention, the formulas in the formula library comprise equations.
According to an embodiment of the present invention, the formula generation apparatus 200 further comprises a formula library generation module.
A formula library generation module for performing the following operations:
rearranging formula elements before the equal sign in the original equation obtained from the original formula library; and/or
Rearranging formula elements after the equal sign in the original equation obtained from the original formula library;
and generating a formula library by using the formulas generated after rearrangement.
According to an embodiment of the present invention, the recognition module 201 includes a preprocessing sub-module and a recognition sub-module.
The preprocessing submodule is used for preprocessing the acquired track points to generate a track point matrix;
and the identification submodule is used for inputting the trace point matrix into the formula identification model and outputting the formula to be matched.
According to the embodiment of the invention, the preprocessing submodule comprises a sorting unit, a normalizing unit and a matrix generating unit.
The sequencing unit is used for sequencing the track points by utilizing the spatial characteristics of the track points to generate a first track point sequence;
the normalization unit is used for carrying out weighted normalization on the first track point sequence to generate a second track point sequence;
and the matrix generating unit is used for generating a track point matrix based on the second track point sequence.
According to the embodiment of the invention, the track point comprises state information, and the state information is used for representing whether the track point is a stroke termination point.
According to an embodiment of the present invention, the preprocessing sub-module further includes a dividing unit.
And the dividing unit is used for dividing the plurality of track points in the second track point sequence into at least one track point set according to the state information, wherein the track points in each track point set belong to the same stroke.
According to the embodiment of the invention, the matrix generation submodule comprises a judgment unit, a sparse sampling unit, a set determination unit and an interpolation unit.
The judging unit is used for determining whether a track point group meeting the sparse sampling condition exists in the track point set or not;
the sparse sampling unit is used for carrying out sparse sampling on the track point group to generate a third track point sequence under the condition that the track point group meeting the sparse sampling condition exists in the track point set;
the set determining unit is used for determining whether a track point set with the track point quantity smaller than a preset threshold exists in the third track point sequence;
and the interpolation unit is used for inserting at least one track point to the track point set according to a preset rule under the condition that the track point set with the track point quantity smaller than a preset threshold exists in the third track point sequence until the track point quantity in the track point set is larger than or equal to the preset threshold, and generating a track point matrix, otherwise, taking the third track point sequence as the track point matrix.
According to an embodiment of the present invention, the normalization unit is configured to perform weighted normalization on the first track point sequence by using the following formula (1) to generate a second track point sequence.
Wherein x isnewAbscissa, x, representing the locus points generated after weighted normalizationoldThe abscissa representing the trace point before weighted normalization,
represents the abscissa weighted mean, sd represents the variance, ynewOrdinate, y, representing the locus points generated after weighted normalizationoldRepresenting the ordinate of the trace point before weighted normalization,
a weighted mean of the ordinate is represented,
xijthe abscissa representing the jth track point in the ith stroke,
yijthe ordinate, l, representing the jth track point in the ith strokeijRepresents the weight of the jth track point in the ith stroke,
according to the embodiment of the present invention, the formula generating apparatus 200 further includes a newly added trace point obtaining module, an overlap ratio determining module, and an adjusting module.
The newly added track point acquisition module is used for acquiring a newly added track point set of the newly added stroke;
the overlap rate determining module is used for determining the overlap rate of the newly added track point set and the existing track point set under the condition that the newly added track point set is overlapped with at least one existing track point set, wherein the track point set is used for representing at least one stroke in the handwriting formula;
The adjusting module is used for deleting the existing track point set or adding a new track point set under the condition that the overlapping rate is greater than the preset overlapping threshold value; otherwise, combining the newly added track point set into the track point set.
According to an embodiment of the present invention, the adjustment module includes an overlapped trace point determination unit, a deletion unit, and a merging unit.
The overlapping track point determining unit is used for determining the overlapping track points of the newly added track point set and the existing at least one track point set;
the deleting unit is used for deleting the overlapped track points from the newly added track point set to generate a first track point set;
and the merging unit is used for merging the first track point set into the track point set.
The formula generating apparatus in fig. 2 may execute the formula generating method in the embodiment shown in fig. 1, and the implementation principle and the technical effect are not described again. The specific manner in which each module, sub-module, unit, and sub-unit of the formula generation apparatus in the above embodiments performs operations has been described in detail in the embodiments related to the method, and will not be described in detail here.
In one possible design, the formula generating apparatus of the embodiment shown in fig. 2 may be implemented as a computing device, as shown in fig. 3, which may include: a processor 301, a memory 302, and a communication interface 303. Wherein the memory 302 has stored thereon executable code, which when executed by the processor 301, causes the processor 301 to at least implement the formula generation method as provided in the previous embodiments.
Fig. 4 is a schematic structural diagram of another computing device provided in this embodiment, and as shown in fig. 4, the computing device 400 may include one or more of the following components: processing component 402, memory 404, power component 406, multimedia component 408, audio component 410, input/output (I/O) interface 412, sensor component 414, and communication component 416.
The processing component 402 generally controls overall operation of the computing device 400, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing component 402 may include one or more processors 420 to execute instructions to perform all or a portion of the method steps 101-105 described above. Further, the processing component 402 can include one or more modules that facilitate interaction between the processing component 402 and other components. For example, the processing component 402 can include a multimedia module to facilitate interaction between the multimedia component 408 and the processing component 402.
The memory 404 is configured to store various types of data to support operation at the computing device 400. Examples of such data include instructions for any application or method operating on computing device 400, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 404 may be implemented by any type or combination of volatile or non-volatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.
The power components 406 provide power to the various components of the computing device 400. The power components 406 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the computing device 400.
The multimedia component 408 includes a screen that provides an output interface between the computing device 400 and the user. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 408 includes a front facing camera and/or a rear facing camera.
The audio component 410 is configured to output and/or input audio signals. For example, audio component 410 includes a Microphone (MIC) configured to receive external audio signals when computing device 400 is in operating modes, such as call mode, record mode, and voice recognition mode. The received audio signal may further be stored in the memory 404 or transmitted via the communication component 416. In some embodiments, audio component 410 also includes a speaker for outputting audio signals.
The input/output interface 412 provides an interface between the processing component 402 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.
The sensor component 414 includes one or more sensors for providing various aspects of state assessment for the computing device 400. For example, the sensor component 414 can detect an open/closed state of the computing device 400, a relative positioning of components, the sensor component 414 can also detect a change in position of the computing device 400 or a component of the computing device 400, a presence or absence of user contact with the computing device 400, an orientation or acceleration/deceleration of the computing device 400, and a change in temperature of the computing device 400. The sensor assembly 414 may include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 414 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 414 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
The communication component 416 is configured to facilitate communications between the computing device 400 and other devices in a wired or wireless manner. The computing device 400 may access a wireless network based on a communication standard, such as WiFi, 2G or 3G or 4G or a combination thereof. In an exemplary embodiment, the communication component 416 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 416 further includes a Near Field Communication (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.
An embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a computer, the method for generating a formula according to the embodiment shown in fig. 1 may be implemented.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (13)
1. A method for generating a formula, comprising:
identifying track points generated in response to handwriting operation, and determining a formula to be matched corresponding to the track points;
according to the formula to be matched, determining a complete formula corresponding to the formula to be matched from a pre-constructed formula library;
and outputting the complete formula.
2. The method according to claim 1, wherein the determining, according to the formula to be matched, a complete formula corresponding to the formula to be matched from a pre-constructed formula library comprises:
determining the number of characters of the formula to be matched;
intercepting the standard formulas in the formula library according to a preset rule to determine a plurality of candidate formulas with lengths corresponding to the number of the characters;
Respectively calculating the similarity between the formula to be matched and the candidate formulas;
determining a target formula from the plurality of candidate formulas according to the similarity;
and taking a standard formula corresponding to the target formula as the complete formula.
3. The method of claim 2, wherein in the case that the number of characters is less than a preset character threshold, before the intercepting the standard formula in the formula library according to a preset rule, the method further comprises:
and expanding the formula to be matched so that the number of characters of the expanded formula to be matched is greater than or equal to the preset character threshold value.
4. The method of any one of claims 1-3, wherein the formulas in the formula library comprise equations;
the formula library is constructed and generated through the following operations:
rearranging formula elements before the equal sign in the original equation obtained from the original formula library; and/or
Rearranging formula elements after the equal sign in the original equation obtained from the original formula library;
and generating the formula library by using the rearranged formulas.
5. The method of claim 1, wherein the recognition of the trace points generated in response to the handwriting operation, and the determining of the formula to be matched corresponding to the trace points comprises:
Preprocessing the acquired track points to generate a track point matrix;
and inputting the track point matrix into a formula identification model, and outputting the formula to be matched.
6. The method according to claim 5, wherein the preprocessing the obtained trace points and generating a trace point matrix comprises:
sequencing the track points by using the spatial characteristics of the track points to generate a first track point sequence;
carrying out weighted normalization on the first track point sequence to generate a second track point sequence;
and generating the track point matrix based on the second track point sequence.
7. The method according to claim 6, wherein the track point comprises state information, and the state information is used for representing whether the track point is a stroke termination point;
the method further comprises the following steps:
dividing a plurality of track points in the second track point sequence into at least one track point set according to the state information, wherein the track points in each track point set belong to the same stroke;
the generating the track point matrix based on the second track point sequence comprises:
determining whether a track point group meeting sparse sampling conditions exists in the track point set;
If the third track point sequence exists, sparse sampling is carried out on the track point group, and a third track point sequence is generated;
determining whether a track point set with the track point quantity smaller than a preset threshold exists in the third track point sequence;
and if the third track point sequence exists, inserting at least one track point into the track point set according to a preset rule until the number of the track points in the track point set is greater than or equal to the preset threshold value, and generating the track point matrix, otherwise, taking the third track point sequence as the track point matrix.
8. The method of claim 6, wherein the weighted normalization of the first sequence of trajectory points is represented by the following equation:
wherein x isnewAbscissa, x, representing trace points generated after weighted normalizationoldRepresents the abscissa of the trace point before weighted normalization,
represents the abscissa weighted mean, sd represents the variance, ynewRepresenting the ordinate, y, of the trace points generated after weighted normalizationoldRepresents the ordinate of the trace point before weighted normalization,
which represents a weighted mean of the ordinate and,
xijthe abscissa representing the jth trace point in the ith stroke,
yijdenotes the ordinate, l, of the jth trace point in the ith stroke ijRepresents the weight of the jth track point in the ith stroke,
9. the method according to claim 1, wherein in case of modifying the formula to be matched, the method further comprises:
acquiring a newly added track point set of the newly added strokes;
under the condition that the newly added track point set is overlapped with at least one existing track point set, determining the overlapping rate of the newly added track point set and the existing track point set, wherein the track point set is used for representing at least one stroke in a handwriting formula;
deleting the existing track point set or the newly added track point set under the condition that the overlapping rate is greater than a preset overlapping threshold value; and otherwise, combining the newly added track point set into the track point set.
10. The method of claim 9, wherein merging the set of newly added trace points into the set of trace points comprises:
determining the overlapping track points of the newly added track point set and the existing at least one track point set;
deleting the overlapped track points from the newly added track point set to generate a first track point set;
and merging the first track point set into the track point set.
11. An apparatus for generating a formula, comprising:
the identification module is used for identifying track points generated in response to handwriting operation and determining a formula to be matched corresponding to the track points;
the matching module is used for determining a complete formula corresponding to the formula to be matched from a pre-constructed formula library according to the formula to be matched;
and the output module is used for outputting the complete formula.
12. A computing device comprising a processing component and a storage component;
the storage component stores one or more computer instructions; the one or more computer instructions are for execution by the processing component to implement the formula generation method of any of claims 1-10.
13. A computer storage medium storing a computer program which, when executed by a computer, implements the formula generation method according to any one of claims 1 to 10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210129247.8A CN114677767A (en) | 2022-02-11 | 2022-02-11 | Formula generation method and device, computing equipment and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210129247.8A CN114677767A (en) | 2022-02-11 | 2022-02-11 | Formula generation method and device, computing equipment and computer storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114677767A true CN114677767A (en) | 2022-06-28 |
Family
ID=82072517
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210129247.8A Pending CN114677767A (en) | 2022-02-11 | 2022-02-11 | Formula generation method and device, computing equipment and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114677767A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117315665A (en) * | 2023-11-30 | 2023-12-29 | 上海又寸科技有限公司 | Automatic question reading method and system based on original handwriting recognition |
-
2022
- 2022-02-11 CN CN202210129247.8A patent/CN114677767A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117315665A (en) * | 2023-11-30 | 2023-12-29 | 上海又寸科技有限公司 | Automatic question reading method and system based on original handwriting recognition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Raheja et al. | Robust gesture recognition using Kinect: A comparison between DTW and HMM | |
| JP2002203208A (en) | Device, method and program for recognizing on-line character and computer readable storage medium | |
| WO2010006087A9 (en) | Process for providing and editing instructions, data, data structures, and algorithms in a computer system | |
| CN108805224B (en) | Multi-symbol hand-drawn sketch recognition method and device with sustainable learning ability | |
| Premaratne et al. | Centroid tracking based dynamic hand gesture recognition using discrete Hidden Markov Models | |
| CN107273032A (en) | Information composition method, device, equipment and computer-readable storage medium | |
| CN114365075B (en) | Method for selecting a graphical object and corresponding device | |
| US20110149115A1 (en) | Electronic device and method for operating a presentation application file | |
| EP4273742A1 (en) | Handwriting recognition method and apparatus, electronic device, and medium | |
| EP4307096A1 (en) | Key function execution method, apparatus and device, and storage medium | |
| US20150088799A1 (en) | Input assistance device, input assisting method, and recording medium storing input assistance program | |
| CN114730241A (en) | Gesture stroke recognition in touch user interface input | |
| CN114677767A (en) | Formula generation method and device, computing equipment and computer storage medium | |
| Gharasuie et al. | Real-time dynamic hand gesture recognition using hidden Markov models | |
| US9465534B2 (en) | Method for automatic computerized process control, and computerized system implementing the same | |
| CN107422921B (en) | Input method, input device, electronic equipment and storage medium | |
| CN113918030A (en) | Handwriting input method and device and handwriting input device | |
| CN113687724A (en) | Candidate character display method and device and electronic equipment | |
| JP6051991B2 (en) | Signal processing apparatus, signal processing method, and signal processing program | |
| JP4686502B2 (en) | Online character recognition device | |
| CN111596810A (en) | Scribble identification method, device, equipment and storage medium | |
| CN111382598A (en) | Identification method and device and electronic equipment | |
| CN106201078B (en) | Track completion method and terminal | |
| CN115016641A (en) | Conference control method, device, conference system and medium based on gesture recognition | |
| KR20190132885A (en) | Apparatus, method and computer program for detecting hand from video |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |