CN110569800B - Detection method of handwriting signal - Google Patents
Detection method of handwriting signal Download PDFInfo
- Publication number
- CN110569800B CN110569800B CN201910852764.6A CN201910852764A CN110569800B CN 110569800 B CN110569800 B CN 110569800B CN 201910852764 A CN201910852764 A CN 201910852764A CN 110569800 B CN110569800 B CN 110569800B
- Authority
- CN
- China
- Prior art keywords
- signal
- handwriting
- window
- wrist movement
- sub
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
- G06F2218/06—Denoising by applying a scale-space analysis, e.g. using wavelet analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
- G06F2218/14—Classification; Matching by matching peak patterns
Landscapes
- Character Discrimination (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
The invention discloses a method for detecting a handwriting signal, which is characterized by detecting the handwriting signal based on angular velocity, obtaining a de-noised signal through wavelet de-noising, increasing the difference between a wrist movement gesture and the handwriting gesture through enhancing a peak value, carrying out piecewise linear fitting, detecting the signal by using a sub-window with a smaller size, and carrying out iterative combination until the sub-window is combined into a mother window containing a complete gesture signal. The invention has low time complexity, does not influence the running speed of a computer, and effectively solves the problem that the traditional window function detection method is difficult to detect the handwritten Chinese character signal compared with the prior art for detecting the handwritten Chinese character signal, thereby greatly improving the detection precision.
Description
Technical Field
The invention relates to the technical field of signal detection, in particular to a detection method of a handwriting signal.
Background
When signal acquisition is performed on handwritten Chinese characters or letters, two different signals, namely a handwriting signal and a wrist movement signal, are generally involved. In the prior art, a conventional window function detection method is usually adopted to detect a handwriting signal, and in the process of implementing the present invention, the inventor of the present application finds that the method in the prior art has at least the following technical problems:
the traditional window function detection method cannot distinguish a handwriting signal from a wrist movement signal, so that the handwriting signal is difficult to accurately detect, namely, the method in the prior art has the technical problem of low detection precision.
Disclosure of Invention
In view of the above, the present invention provides a method for detecting a handwriting signal, so as to solve or at least partially solve the technical problem of low detection accuracy in the methods in the prior art.
In order to solve the above technical problem, the present invention provides a method for detecting a handwriting signal, including:
step S1: acquiring original signal data, and carrying out denoising processing on the original signal data to obtain a denoising signal, wherein the denoising signal comprises a wrist movement signal and a handwriting signal;
step S2: performing peak value enhancement on the de-noised signal to enhance the peak value of the wrist movement signal in the de-noised signal;
step S3: extracting a handwriting signal between two adjacent peak values of the wrist movement signal by adopting a preset peak value detection algorithm, and fitting the extracted handwriting signal by adopting a piecewise linear fitting method to obtain a boundary of the wrist movement signal;
s4: the handwritten signal is detected based on the boundaries of the wrist movement signal and the method of merging sub-windows.
In an embodiment, the denoising processing performed on the original signal data in step S1 specifically includes:
step S1.1: smoothing the acquired original signal data;
step S1.2: and converting the signal data subjected to smoothing processing through wavelet transformation, then suppressing noise through setting a coefficient threshold, and obtaining a denoising signal containing a wrist movement signal and a handwriting signal through wavelet reconstruction.
In one embodiment, step S2 specifically includes:
step S2.1: obtaining a radial component g of a wrist movement signalard[i]And a tangential component gtan[i]，
gard[i]＝yi (1)
Wherein x isi，yi，ziFiltered signal data for three angular velocity axes;
step S2.2: the peak enhancement is designed from the tangential component of the wrist movement signal:
PAF [ i ] represents the peak generated by the peak enhancement.
In one embodiment, in step S3, fitting the extracted handwriting signal by using a piecewise linear fitting method to obtain a boundary of the wrist movement signal includes:
step S3.1: re-determining the interval between two adjacent peaks of the wrist movement signal to cover the handwriting signal;
step S3.2: dividing signals in the detection interval according to piecewise linear fitting, and performing linear fitting on every two sampling points from the peak position, wherein the peak value of the left boundary is positioned on the right side, and the peak value of the right boundary is positioned on the left side;
step S3.3: judging whether the fitting error is lower than a threshold value, if so, adding a sampling point and carrying out linear fitting to obtain the slope of a straight line, otherwise, taking the current segment as a complete segment and starting a new segment; and (3) iteratively executing the steps S3.2-S3.3 until the slope of the linear fitting is lower than a predetermined parameter, resetting the interval end point, and finally obtaining the boundary of the wrist movement signal.
In one embodiment, step S4 specifically includes:
step S4.1: obtaining a sub-window of N samples by using a window function, wherein N is a positive integer,
step S4.2: calculating the energy value E of the combined acceleration signal of three axes in each sub-window by adopting an angular velocity-based methodI，
Where I denotes the index of the current sub-window, xi、yi、ziValues representing three axes of acceleration, FiA combined acceleration signal representing three axes;
step S4.3: judging whether the acceleration signal energy value in each sub-window is larger than a noise threshold value, if so, taking the acceleration signal energy value as a part of the signal and recording the acceleration signal energy value until all sub-windows contained in the signal are detected;
step S4.4: and combining all the recorded sub-windows to obtain a mother window, and detecting the handwriting signal by combining the boundary of the wrist movement signal and the mother window obtained by combination.
One or more technical solutions in the embodiments of the present application have at least one or more of the following technical effects:
the invention provides a detection method of a handwriting signal, which comprises the steps of firstly, collecting original signal data, and carrying out denoising processing on the original signal data to obtain a denoising signal; secondly, performing peak value enhancement on the de-noised signal to enhance the peak value of the wrist movement signal in the de-noised signal; secondly, extracting a handwriting signal between two adjacent peak values of the wrist movement signal by adopting a preset peak value detection algorithm, and fitting the extracted handwriting signal by adopting a piecewise linear fitting method to obtain a boundary of the wrist movement signal; finally, the handwritten signal is detected based on the boundaries of the wrist movement signal and the method of merging sub-windows.
The method provided by the invention is used for enhancing the peak value of the wrist movement signal in the de-noising signal by enhancing the peak value of the de-noising signal, thereby increasing the difference between the wrist movement gesture and the handwriting gesture, namely increasing the difference between the handwriting signal and the wrist movement signal, thereby more accurately detecting the handwriting signal.
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 introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a schematic flowchart of a method for detecting a handwriting signal according to the present invention;
FIG. 2 is a denoised signal diagram obtained by wavelet reconstruction in the embodiment of the present invention;
FIG. 3 is a diagram illustrating signal detection based on a conventional window function method and a sub-window combining method according to an embodiment of the present invention;
FIGS. 4-7 are schematic diagrams of detection rates under different size combinations in the specific example of the present invention.
Detailed Description
The invention aims to provide a method for detecting a handwriting signal, aiming at the technical problem that the detection precision is not high in the method in the prior art, so that the aim of improving the detection precision is fulfilled.
In order to achieve the above purpose, the main concept of the invention is as follows:
detecting a handwritten signal by adopting an angular velocity-based method, obtaining a de-noising signal through wavelet de-noising, increasing the difference between a wrist movement gesture and the handwritten gesture by enhancing a peak value, performing piecewise linear fitting, detecting the signal by using a sub-window with a smaller size, and performing iterative combination until the sub-window is combined into a mother window containing a complete gesture signal. Wherein, when piecewise linear fitting is performed, a signal is extracted between two adjacent peaks according to a peak detection algorithm, and then the interval is narrowed to cover the handwritten signal. To eliminate the effect of the peak edge signal and to re-determine the interval, the signal in the detection interval is segmented according to a piecewise linear fit. Starting from the peak position, a linear fit is made to the first two points. The peak of the left border is located on the right side and the peak of the right border is located on the left side. If the fitting error is lower than the threshold value, adding one more sampling point; otherwise, the current segment is treated as a complete segment and a new segment is started. Fitting is performed iteratively until the slope of the linear fit is below a preset parameter, resetting the interval end.
The method of the invention has low time complexity, does not influence the running speed of a computer, effectively solves the problem that the traditional window function detection method is difficult to detect the handwritten Chinese character signal and greatly improves the detection precision compared with the prior art for detecting the handwritten Chinese character signal.
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, 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 some, but not all, embodiments of the present invention. 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.
The present embodiment provides a method for detecting a handwriting signal, please refer to fig. 1, where the method includes:
step S1: acquiring original signal data, and carrying out denoising processing on the original signal data to obtain a denoising signal, wherein the denoising signal comprises a wrist movement signal and a handwriting signal.
Specifically, since the raw signal data contains high-frequency random noise due to hand shake and internal noise of the sensor, it is necessary to perform denoising processing, and an existing method can be used.
In an embodiment, the denoising processing performed on the original signal data in step S1 specifically includes:
step S1.1: smoothing the acquired original signal data;
step S1.2: and converting the signal data subjected to smoothing processing through wavelet transformation, then suppressing noise through setting a coefficient threshold, and obtaining a denoising signal containing a wrist movement signal and a handwriting signal through wavelet reconstruction.
Specifically, the wavelet denoising method is adopted in this embodiment, and first, smoothing processing is performed on the acquired original signal data, and the angular velocity signal is converted into a time-frequency domain by wavelet transform for signal denoising. Fig. 2 shows a denoised signal diagram obtained by wavelet reconstruction.
Step S2: and performing peak value enhancement on the denoised signal so as to enhance the peak value of the wrist movement signal in the denoised signal.
Specifically, after a Chinese character is completed, the wrist is lifted and then lowered, and the movement signal for continuing writing the next Chinese character can be approximated to circular movements having different radii and speeds. Their angular velocity in the y-axis direction (y)i) Along the forearm, it is the radial component of angular velocity (g)ard). Tangential component (g)tan) Angular velocity (x) that can pass in the x-axis directioni) And angular velocity (z) in the z-axis directioni) And (4) calculating. Thus, the amplitude variation of the gyroscope's x-axis and z-axis signals during wrist movement is greater than the handwritten Chinese character's signal, i.e. gtanThe variation is small when handwriting Chinese characters, but the variation is large when the wrist moves. That is, after the peak value of the de-noising signal is enhanced, the influence on the wrist movement signal is larger, and the influence on the handwritten Chinese character signal is smaller, so that the difference between the wrist movement gesture and the handwriting gesture can be increased, and the handwriting signal and the wrist movement signal can be more accurately distinguished.
In one embodiment, step S2 specifically includes:
step S2.1: obtaining a radial component g of a wrist movement signalard[i]And a tangential component gtan[i]，
gard[i]＝yi (1)
Wherein x isi，yi，ziFiltered signal data for three angular velocity axes;
step S2.2: the peak enhancement is designed from the tangential component of the wrist movement signal:
PAF [ i ] represents the peak generated by the peak enhancement.
Specifically, the PAF is designed according to the tangential component of the wrist movement gesture, in the specific implementation process, a preset number of sampling points are adopted for convolution, and finally wave crests are generated at the points affected by the movement, wherein the preset number is 12, 15, 20 and the like.
Step S3: and extracting a handwriting signal between two adjacent peak values of the wrist movement signal by adopting a preset peak value detection algorithm, and fitting the extracted handwriting signal by adopting a piecewise linear fitting method to obtain the boundary of the wrist movement signal.
In particular, after enhancing the peaks of the wrist signal (but the handwritten signal peak enhancement is not significant), the signal may be extracted between two adjacent peaks according to the peak detection algorithm proposed by Palshikar in 2009. That is, the signal between two peaks is a handwriting signal of one word, and then the boundary of the wrist movement signal is obtained by adopting a linear fitting mode.
In one embodiment, in step S3, fitting the extracted handwriting signal by using a piecewise linear fitting method to obtain a boundary of the wrist movement signal includes:
step S3.1: re-determining the interval between two adjacent peaks of the wrist movement signal to cover the handwriting signal;
step S3.2: dividing signals in the detection interval according to piecewise linear fitting, and performing linear fitting on every two sampling points from the peak position, wherein the peak value of the left boundary is positioned on the right side, and the peak value of the right boundary is positioned on the left side;
step S3.3: judging whether the fitting error is lower than a threshold value, if so, adding a sampling point and carrying out linear fitting to obtain the slope of a straight line, otherwise, taking the current segment as a complete segment and starting a new segment; and (3) iteratively executing the steps S3.2-S3.3 until the slope of the linear fitting is lower than a predetermined parameter, resetting the interval end point, and finally obtaining the boundary of the wrist movement signal.
Specifically, when extracting a signal between two adjacent peaks based on a peak detection algorithm, it is necessary to eliminate the influence of a peak edge signal and to re-determine an interval so as to cover a handwriting signal.
In the linear fitting process, if the fitting effect of two points is not good (namely the fitting error is lower than a threshold value), one sampling point is added, three points are fitted, if the fitting effect of the three points is not good, four points are fitted, and the process is executed iteratively until the effect is good. The threshold value may be set according to practical situations, for example, set to 0.1, 0.2, etc., and the predetermined parameter may be 10, 12, etc. The boundary of the wrist movement signal can be finally obtained through the steps S3.1 to S3.3, the wrist movement signal is a signal before the next Chinese character is written after one Chinese character is written, the signal can be used for Chinese character segmentation, and the boundary of the handwriting signal can be found through the sub-window combination detection in the step S4.
Step S4: the handwritten signal is detected based on the boundaries of the wrist movement signal and the method of merging sub-windows.
In one embodiment, step S4 specifically includes:
step S4.1: obtaining a sub-window of N samples by using a window function, wherein N is a positive integer,
step S4.2: calculating the energy value E of the combined acceleration signal of three axes in each sub-window by adopting an angular velocity-based methodI，
Where I denotes the index of the current sub-window, xi、yi、ziValues representing three axes of acceleration, FiA combined acceleration signal representing three axes;
step S4.3: judging whether the acceleration signal energy value in each sub-window is larger than a noise threshold value, if so, taking the acceleration signal energy value as a part of the signal and recording the acceleration signal energy value until all sub-windows contained in the signal are detected;
step S4.4: and combining all the recorded sub-windows to obtain a mother window, and detecting the handwriting signal by combining the boundary of the wrist movement signal and the mother window obtained by combination.
Specifically, considering that signals in the interval contain a lot of redundant information and noise points, only handwritten Chinese characters are reserved for eliminating noise interference, and a window function is used for further extracting Chinese character signals. Meanwhile, considering that the writing behavior has more significant influence on the angular velocity signal, the sub-window merging method is implemented based on the angular velocity.
In the specific implementation process, if the signal energy in the sub-window is greater than the noise threshold, the signal energy is regarded as a part of the signal and recorded, then the next sub-window is continuously detected, if the signal energy is still greater than the threshold, the next sub-window is still reserved and recorded, and all sub-windows included in the signal are sequentially detected. The division of the sub-window may be implemented according to actual situations, that is, the number N of samples included in the sub-window may be 20, 30, 40, and so on. When the time interval containing the action is detected to exceed the preset time next time, the signal is considered to be detected completely.
All the sub-windows recorded in the front are combined, the combined window is called a mother window, and the signal in the mother window is used as a human motion signal to be detected, namely a handwriting signal.
Fig. 3 is a schematic diagram of signal detection based on the conventional window function method and the sub-window combining method according to an embodiment of the present invention, in which the left side is a signal detection diagram based on the conventional window function method, and the right side is a signal detection diagram based on the sub-window combining method.
The invention provides a signal detection method of handwritten Chinese characters, which has low complexity and is easy to understand and use, and solves the problem that the traditional window function detection method is difficult to detect the handwritten Chinese character signals. The method comprises the steps of firstly converting an angular velocity signal into a time-frequency domain by using wavelet transform, then suppressing noise by setting a coefficient threshold, and then obtaining a de-noised signal by wavelet reconstruction. And increases the difference between wrist motion gestures and hand writing gestures with peak enhancement functions. After enhancing the peaks of the wrist signal, the signal is extracted between two adjacent peaks according to a peak detection algorithm, and then the space is narrowed and the handwritten signal is overlaid. And finally, segmenting the signals in the detection interval according to piecewise linear fitting, and further extracting the Chinese character signals by using a window function.
In order to more clearly illustrate the implementation and beneficial effects of the method provided by the invention, the following detailed description is given by specific examples.
In the specific implementation process, the specific method flow of the invention is as follows:
input xi,yi,zi// values of three axes of acceleration
A threshold th and a preset parameter K.
And (3) outputting: determined interval (PAFNewstart, PAFNewend)// data after peak enhancement
First parameters are determined, including the size of the sub-window, the threshold th for configuration error and the slope K. The invention evaluates the detection rates under different window sizes, th and K. Sd represents the number of detected characters, and St represents the actual number of characters. The detection rate is defined as
In specific implementation, 50 Chinese characters are randomly selected from each person, and the average detection rate of the method is solved. FIGS. 4-7 show the detection rates with different size combinations (4,16,32,64,128), th (0.01,0.1,1,10) and K (0.1,1,10, 100). As can be seen from fig. 4 to 7, as the size increases, the relative error first rises and then falls. This is because large size windows will cover redundant signals, while small size windows will lose signals, which results in lower recognition performance. At the same time, too large or too small results in reduced accuracy. When th is 10, the Chinese character is hardly perceived. And as shown in fig. 6, when K is 10, the detection rate is highest (94.8%), the standard deviation is small (1.5%), and the sizes are 0.1 and 32. Therefore, the window size may be set to 32, K to 10, and th to 0.1 in the sub-window merging method.
In summary, the method provided by the invention has the following advantages: the time complexity is low, the understanding is easy, the signal detection precision is high, and the effect is good. The handwriting signal detection method based on the angular velocity is used for detecting the signal of the writing, and compared with a common signal detection method, the method effectively avoids the incomplete signal detection and greatly improves the detection precision under the condition of not influencing the running speed of a computer.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various modifications and variations can be made in the embodiments of the present invention without departing from the spirit or scope of the embodiments of the invention. Thus, if such modifications and variations of the embodiments of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to encompass such modifications and variations.
Claims (4)
1. A method for detecting a handwriting signal, comprising:
step S1: acquiring original signal data, and carrying out denoising processing on the original signal data to obtain a denoising signal, wherein the denoising signal comprises a wrist movement signal and a handwriting signal;
step S2: performing peak value enhancement on the de-noised signal to enhance the peak value of the wrist movement signal in the de-noised signal;
step S3: extracting a handwriting signal between two adjacent peak values of the wrist movement signal by adopting a preset peak value detection algorithm, and fitting the extracted handwriting signal by adopting a piecewise linear fitting method to obtain a boundary of the wrist movement signal;
s4: detecting a handwriting signal based on the boundary of the wrist movement signal and a method of combining sub-windows;
wherein, step S2 specifically includes:
step S2.1: obtaining a radial component g of a wrist movement signalard[i]And a tangential component gtan[i]，
gard[i]＝yi (1)
Wherein x isi，yi，ziFiltered signal data for three angular velocity axes;
step S2.2: the peak enhancement is designed from the tangential component of the wrist movement signal:
PAF [ i ] represents the peak generated by the peak enhancement.
2. The method as claimed in claim 1, wherein the denoising processing on the raw signal data in step S1 specifically includes:
step S1.1: smoothing the acquired original signal data;
step S1.2: and converting the signal data subjected to smoothing processing through wavelet transformation, then suppressing noise through setting a coefficient threshold, and obtaining a denoising signal containing a wrist movement signal and a handwriting signal through wavelet reconstruction.
3. The method according to claim 1, wherein in step S3, fitting the extracted handwriting signal by using a piecewise linear fitting method to obtain the boundary of the wrist movement signal includes:
step S3.1: re-determining the interval between two adjacent peaks of the wrist movement signal to cover the handwriting signal;
step S3.2: dividing signals in the detection interval according to piecewise linear fitting, and performing linear fitting on every two sampling points from the peak position, wherein the peak value of the left boundary is positioned on the right side, and the peak value of the right boundary is positioned on the left side;
step S3.3: judging whether the fitting error is lower than a threshold value, if so, adding a sampling point and carrying out linear fitting to obtain the slope of a straight line, otherwise, taking the current segment as a complete segment and starting a new segment; and (3) iteratively executing the steps S3.2-S3.3 until the slope of the linear fitting is lower than a predetermined parameter, resetting the interval end point, and finally obtaining the boundary of the wrist movement signal.
4. The method according to claim 1, wherein step S4 specifically comprises:
step S4.1: obtaining a sub-window of N samples by using a window function, wherein N is a positive integer,
step S4.2: calculating the energy value E of the combined acceleration signal of three axes in each sub-window by adopting an angular velocity-based methodI，
Where I denotes the index of the current sub-window, xi、yi、ziValues representing three axes of acceleration, FiRepresents threeA resultant acceleration signal of the shaft;
step S4.3: judging whether the acceleration signal energy value in each sub-window is larger than a noise threshold value, if so, taking the acceleration signal energy value as a part of the signal and recording the acceleration signal energy value until all sub-windows contained in the signal are detected;
step S4.4: and combining all the recorded sub-windows to obtain a mother window, and detecting the handwriting signal by combining the boundary of the wrist movement signal and the mother window obtained by combination.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910852764.6A CN110569800B (en) | 2019-09-10 | 2019-09-10 | Detection method of handwriting signal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910852764.6A CN110569800B (en) | 2019-09-10 | 2019-09-10 | Detection method of handwriting signal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110569800A CN110569800A (en) | 2019-12-13 |
| CN110569800B true CN110569800B (en) | 2022-04-12 |
Family
ID=68778691
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910852764.6A Active CN110569800B (en) | 2019-09-10 | 2019-09-10 | Detection method of handwriting signal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110569800B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101853380A (en) * | 2010-06-18 | 2010-10-06 | 华南理工大学 | Rotary characteristic extraction method based on gesture identification of acceleration sensor |
| CN103315744A (en) * | 2013-07-01 | 2013-09-25 | 中国科学院合肥物质科学研究院 | Hand tremor detection method |
| CN103902100A (en) * | 2014-04-01 | 2014-07-02 | 西北工业大学 | Speed characteristics based stroke partitioning method for intelligent freehand sketching |
| CN105303201A (en) * | 2015-11-03 | 2016-02-03 | 广东小天才科技有限公司 | Method and system for performing handwriting identification based on action sensing |
| CN105809102A (en) * | 2015-01-16 | 2016-07-27 | 新普科技股份有限公司 | Handwriting Track Recognition Method and System |
| CN105844216A (en) * | 2016-03-11 | 2016-08-10 | 南京航空航天大学 | Detection and matching mechanism for recognition of handwritten letters using WiFi signals |
| CN105988567A (en) * | 2015-02-12 | 2016-10-05 | 北京三星通信技术研究有限公司 | Handwritten information recognition method and device |
| CN106127188A (en) * | 2016-07-05 | 2016-11-16 | 武汉理工大学 | A kind of Handwritten Digit Recognition method based on gyroscope |
| CN107145844A (en) * | 2017-04-20 | 2017-09-08 | 武汉大学 | A kind of signal detecting method merged based on sub- window |
| CN109002803A (en) * | 2018-07-24 | 2018-12-14 | 武汉大学 | A kind of pen holding posture detection and Chinese characters' order of strokes recognition method based on smartwatch |
| CN110196635A (en) * | 2019-04-28 | 2019-09-03 | 浙江大学 | A kind of gesture input method based on wearable device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10048768B2 (en) * | 2016-01-22 | 2018-08-14 | Sharp Laboratories Of America, Inc. | Systems and methods for determining input movement |
-
2019
- 2019-09-10 CN CN201910852764.6A patent/CN110569800B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101853380A (en) * | 2010-06-18 | 2010-10-06 | 华南理工大学 | Rotary characteristic extraction method based on gesture identification of acceleration sensor |
| CN103315744A (en) * | 2013-07-01 | 2013-09-25 | 中国科学院合肥物质科学研究院 | Hand tremor detection method |
| CN103902100A (en) * | 2014-04-01 | 2014-07-02 | 西北工业大学 | Speed characteristics based stroke partitioning method for intelligent freehand sketching |
| CN105809102A (en) * | 2015-01-16 | 2016-07-27 | 新普科技股份有限公司 | Handwriting Track Recognition Method and System |
| CN105988567A (en) * | 2015-02-12 | 2016-10-05 | 北京三星通信技术研究有限公司 | Handwritten information recognition method and device |
| CN105303201A (en) * | 2015-11-03 | 2016-02-03 | 广东小天才科技有限公司 | Method and system for performing handwriting identification based on action sensing |
| CN105844216A (en) * | 2016-03-11 | 2016-08-10 | 南京航空航天大学 | Detection and matching mechanism for recognition of handwritten letters using WiFi signals |
| CN106127188A (en) * | 2016-07-05 | 2016-11-16 | 武汉理工大学 | A kind of Handwritten Digit Recognition method based on gyroscope |
| CN107145844A (en) * | 2017-04-20 | 2017-09-08 | 武汉大学 | A kind of signal detecting method merged based on sub- window |
| CN109002803A (en) * | 2018-07-24 | 2018-12-14 | 武汉大学 | A kind of pen holding posture detection and Chinese characters' order of strokes recognition method based on smartwatch |
| CN110196635A (en) * | 2019-04-28 | 2019-09-03 | 浙江大学 | A kind of gesture input method based on wearable device |
Non-Patent Citations (1)
| Title |
|---|
| 基于惯导信号的手写数字识别研究;张桂花;《中国优秀硕士学位论文全文数据库 信息科技辑》;20180115;1-62 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110569800A (en) | 2019-12-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108805840B (en) | Image denoising method, device, terminal and computer readable storage medium | |
| CN111833366A (en) | Edge detection method based on Canny algorithm | |
| WO2018041250A1 (en) | Touch control operation identification method and apparatus | |
| CN108416740B (en) | Iterative adaptive median filtering method for eliminating salt and pepper noise | |
| Liu | A new finger vein feature extraction algorithm | |
| US20190332858A1 (en) | Method and device for identifying wrist, method for identifying gesture, electronic equipment and computer-readable storage medium | |
| CN110619265A (en) | Ball mill cylinder vibration signal combined denoising method and device and storage medium | |
| CN112883824A (en) | Finger vein feature recognition device for intelligent blood sampling and recognition method thereof | |
| CN109190517B (en) | Finger vein image edge extraction and direction correction method | |
| CN112115853A (en) | Gesture recognition method and device, computer storage medium and electronic equipment | |
| CN107239739B (en) | Signal envelope extraction method with adjustable scale parameter control | |
| CN110569800B (en) | Detection method of handwriting signal | |
| CN107463916A (en) | Fingerprint minutiae feature acquisition methods based on spiral operator | |
| CN111639555B (en) | Finger vein image noise accurate extraction and adaptive filtering denoising method and device | |
| Xiong et al. | Revisiting shortstraw: improving corner finding in sketch-based interfaces | |
| CN115994870B (en) | Image processing method for enhancing denoising | |
| EP3125194B1 (en) | Biometric authentication device, biometric authentication method, and program | |
| CN115712822A (en) | Bearing fault diagnosis method, system and medium based on CEEMDAN-OMEDA | |
| Li et al. | Algorithm of Canny Operator Edge Pre-processing Based on Mathematical Morphology | |
| CN110502996B (en) | Dynamic identification method for fuzzy finger vein image | |
| CN110399783B (en) | Traffic action triggering platform, method and storage medium based on image analysis | |
| CN110070874B (en) | Voice noise reduction method and device for voiceprint recognition | |
| JP2005010842A (en) | Banded pattern extraction system and line pattern extraction method | |
| CN116304777B (en) | Self-adaptive electrocardiosignal denoising method and system based on reference signal during rest | |
| CN105787444A (en) | Signal denoising method based on V system |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |