CN108492870B - Method and system for testing and detecting picture clock based on digital pen - Google Patents
Method and system for testing and detecting picture clock based on digital pen Download PDFInfo
- Publication number
- CN108492870B CN108492870B CN201710100530.7A CN201710100530A CN108492870B CN 108492870 B CN108492870 B CN 108492870B CN 201710100530 A CN201710100530 A CN 201710100530A CN 108492870 B CN108492870 B CN 108492870B
- Authority
- CN
- China
- Prior art keywords
- strokes
- stroke
- digital
- outline
- pointer
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/32—Digital ink
- G06V30/333—Preprocessing; Feature extraction
- G06V30/347—Sampling; Contour coding; Stroke extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/32—Digital ink
- G06V30/36—Matching; Classification
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Character Discrimination (AREA)
Abstract
The invention provides a test and detection method for a painting clock based on a digital pen, which comprises the following steps: 1) acquiring handwriting data of a user for clock drawing test; 2) recognizing outline strokes, pointer strokes and digital strokes of the handwriting data, and extracting stroke characteristics; 3) and detecting the outline strokes, the pointer strokes and the digital strokes according to the stroke characteristics to realize the test of the drawing clock. The invention also provides a system for testing and detecting the painting clock based on the digital pen, which comprises: a digital pen for drawing graphics; and the information acquisition equipment is connected with the digital pen and comprises a liquid crystal digital screen, a data acquisition application and a data analysis application, handwriting data of the digital pen is input and recorded through the liquid crystal display screen, the handwriting data is acquired through the data acquisition application, and stroke characteristics of the handwriting data are identified and scored through the data analysis application.
Description
Technical Field
The invention belongs to the field of digital medical treatment, and particularly relates to a detection method and a detection system for a drawing clock test (CDT) based on a digital pen track.
Background
Technological innovations have played a major role in the advancement of the field of telemedicine and care. In the home care sector in particular, there is already technology available for monitoring and controlling some chronic diseases, which effectively reduces the cost of medical care (reference: Kayyali, Basel, Zeb Kimmel, and Steve van Kuiken. "spring the marker for high-tech home health." McKinsey Quarry separator (2011)). Home medical care technology may make people more actively concerned about their health (ref: Horwitz, c.m., et al, "Is home health technology estimate for reactive self-care. For example, people can use the electronic sphygmomanometer to detect own blood pressure at home to prevent stroke. A person with diabetes can also perform a simple finger prick test at home to detect his or her blood glucose level.
With the rapid growth of aging population, cognitive dysfunction such as dementia and the like are becoming common diseases (reference: Esther Landhuis "Presence of Dementia, AD, in China Eclipses Predictions" Alzheimer, 14Jun 2013, URL: http:// www.alzforum.org/news/research-news/prediction-severity-AD-China-experiments-pr syndromes), and unlike other diseases with physiologically obvious characteristics, early detection of cognitive dysfunction is difficult. Older people and their families will consider some of the early signs of dementia, such as Memory decline, as a normal effect of increasing age (ref: r.e. powers, j.w. ashford and s.pessin, Memory materials, Alzheimer's Foundation of America, 2008.). The deviation and lack of patient cognition for the disease is a major obstacle to the identification of early stage dementia. And healthcare practitioners are largely reluctant to implement means of dementia detection in their offices (references: p. iracleous, j. x. nie, c.s. tracy, r. moineidin, z. island, k.i. shulman and r.e. upsshur, Primary care physics' approaches to scientific screening: International business of geographic Psychiatry 25(1) (2010), 23-29).
There are several medical approaches to detecting cognitive function, where mapping is a fundamental information source used by neuropsychologists to understand their patients (ref: SMTHAD. on the Use of drawing Tasks in Neuropsychological Association [ J ]. American psychological Association,2009, 23). Lezake (lexak.m.d.) considers that the mapping task is central in neuropsychology. The cognitive function and the hand function are required to be drawn, so that the cognitive function and the hand function are combined to save the examination time. In clinical application, the evaluation of cognitive function and motor function (especially the evaluation of hand function) often affect each other, and the single analysis of either function is not good for the objective understanding of the clinical general symptoms.
The clock-drawing test is a common and easy-to-operate tool for testing the cranial nerve system, and requires a person to be tested to draw a clock independently on paper and mark a designated time as required. The bell test is simple to implement, but it requires the involvement of a number of cognitive functions to accomplish it. Meanwhile, the test has small relevance to the culture, and a person receiving the test does not need necessary culture level, and can draw a clock which is considered to be correct by himself to finish the test only by listening to simple instruction words.
However, the current practice of nervous system examination is still to be operated ten years ago, and the patient is required to conduct a clock-drawing test on a given piece of paper with a pen, and the test needs manual monitoring and analysis, which consumes labor and effort, and has slow judgment speed, and the judgment has strong subjectivity, and results of different cases may have great discrepancy and great errors.
The rapid development of information technology has prompted the advancement of nervous system prevention and diagnostic techniques. The novel interactive technology helps people to obtain information which cannot be obtained by the traditional method, and fuzzy qualitative data are converted into digital quantitative data. The digital input device provides an irreplaceable drawing input mode, can automatically collect accurate handwriting data in a two-dimensional X-Y plane, can effectively quantify cognitive and hand movement function data, and improves the comprehensiveness and objectivity of disease information judgment. The present invention provides a solution for this.
Disclosure of Invention
The invention aims to provide a digital pen-based painting clock test detection method and system, which can collect handwriting data of a digital pen during painting clock test, effectively quantize cognitive and hand movement function data and improve the comprehensiveness and objectivity of disease information judgment.
In order to solve the technical problems, the technical scheme of the invention is as follows:
a test and detection method for a painting clock based on a digital pen comprises the following steps:
1) acquiring handwriting data of a user for clock drawing test;
2) recognizing outline strokes, pointer strokes and digital strokes of the handwriting data, and extracting stroke characteristics;
3) and detecting the outline strokes, the pointer strokes and the digital strokes according to the stroke characteristics to realize the test of the drawing clock.
Further, a RandomForest classifier is adopted for stroke recognition.
Further, preprocessing handwriting data before recognition, comprising:
1) resampling the handwriting data to ensure that the distances between sampling points are equal to T;
2) and the handwriting data is smoothed, so that the stroke jitter degree is reduced.
Further, the stroke recognition method comprises the following steps:
1) recognizing outline strokes in a total stroke set S1 to obtain an outline stroke set O, searching a redundant stroke set RS1 of the outline strokes, and removing the stroke set O and RS1 in S1 to obtain a stroke set S2, namely S2 is S1- (O + RS 1);
2) identifying pointer strokes in the stroke set S2 to obtain a pointer stroke set H, searching a redundant stroke set RS2 of the pointer strokes, and removing the stroke set H and the RS2 in S2 to obtain a stroke set S3, namely S3 is S2- (H + RS 2);
3) Digital strokes are identified in stroke set S3.
Further, contour strokes, pointer strokes and digital strokes are detected by a drawing and clock scoring mechanism of Beijing collaborating with hospitals, and the scoring mechanism is totally divided into 3 points, wherein the contour strokes, the pointer strokes and the digital strokes respectively account for 1 point.
Further, the stroke characteristics of the outline stroke include: 1) whether outline strokes exist; 2) whether the outline strokes are closed; 3) the number of strokes N1; 4) the ratio R1 of the height H to the width W of the outline stroke closure; 5) the ratio R2 of the outline stroke length L to the inscribed circle perimeter P of the outline stroke closure.
Further, the stroke characteristics of the pointer stroke include: 1) whether the hour hand stroke and the minute hand stroke are intersected or not; 2) a ratio R3 of the hour hand stroke length HL to the minute hand stroke length ML; 3) quadrant quadrint 1 where the hour hand strokes are located; 4) quadrant quadrirange 2 in which the split-pin strokes are located.
Further, the stroke characteristics of the digital strokes include: the correct number of digits N2.
Further, pointer strokes are segmented by adopting a ShortStraw method to obtain hour pointer strokes and minute pointer strokes.
Further, the digital strokes are aggregated according to different numbers by the K-means algorithm.
A digital pen-based paint clock test detection system, comprising:
A digital pen for drawing graphics;
and the information acquisition equipment is connected with the digital pen and comprises a liquid crystal digital screen, a data acquisition application and a data analysis application, handwriting data of the digital pen is input and recorded through the liquid crystal display screen, the handwriting data is acquired through the data acquisition application, and stroke characteristics of the handwriting data are identified and scored through the data analysis application.
Further, the information acquisition device is also used for displaying the drawing task, and the task comprises the following steps: drawing the outline of the clock, drawing the time number and drawing the pointer.
Further, Microsoft Tablet PC Platform SDK 1.7 is used as the stroke recognition engine for numbers.
The invention has the beneficial effects that: compared with the prior art, the invention uses the digital pen and the information acquisition equipment to obtain accurate handwriting data, and can improve the data detection accuracy of the picture clock test. By data analysis application, recognition algorithm and scoring mechanism, the automation of handwriting data analysis and diagnosis of the drawing clock test can be realized, and the time and labor required by the traditional drawing clock test are reduced.
Drawings
FIG. 1 is a schematic diagram of a handwriting data collection method for drawing clock testing according to the present invention.
FIG. 2 is a flow chart of a method for testing a picture clock according to the present invention.
FIG. 3 is a schematic diagram of a point closure.
FIG. 4 is a handwriting chart of a user's clock test.
Fig. 5 is a diagram illustrating the effect of the ShortStraw algorithm.
Fig. 6 is a diagram illustrating the effect of the number recognition algorithm.
Detailed Description
The present invention will be described in detail with reference to the following embodiments, which are intended to explain the purpose, features, and advantages of the invention.
The embodiment provides a digital pen-based painting clock test detection method and system, handwriting data of the digital pen is collected during the painting clock test, cognitive and hand movement function data can be effectively quantized, and comprehensiveness and objectivity of disease information judgment are improved. Handwriting data of the digital pen is collected, stroke feature extraction is carried out by combining a handwriting recognition algorithm, a CDT data model based on content understanding is established, and a basis is established for the cognitive-motor function evaluation method, as shown in figure 1.
The operation flow of this embodiment is shown in fig. 2, and specifically includes the following steps:
firstly, drawing a clock on a liquid crystal digital screen by using a digital pen, and acquiring original data of handwriting.
1. A user opens a data acquisition application;
2. after the user selects to start, displaying the drawing task of the user on the information acquisition equipment: specifically, "draw the outline of the clock", "mark a numeral indicating time on the outline", "set a pointer so that the time of the clock is displayed as 11: 10 ";
3. The user draws a clock on the liquid crystal digital screen, and the handwriting data of the user is recorded during the clock drawing, wherein the handwriting data is a two-dimensional vector (X, Y) of an X-Y coordinate system, and is shown in figure 4.
In the present embodiment, the digital pen uses Wacom KP-701E; the liquid crystal digital screen uses Wacom 13HD, the screen size of which is 13.0 inches, the resolution is 1280 × 800, and the operating system is winods 8, but the liquid crystal digital screen is not limited to this device, and other devices capable of collecting equivalent information may be used.
Secondly, preprocessing the original data:
one) resamples the raw data, the algorithm is as follows:
1. first find the closed rectangle of the original sample point and calculate the diagonal length BbDiagnonal of the rectangle.
Wherein, let XmaxIs the maximum value of X in the original sampling point set Points, XminIs a minimum value of YmaxIs the maximum value of Y in the original sampling point set Points, YminAt the minimum, the closed rectangle of the original sample point is the starting point (X)min,Ymin) And (X)max,Ymax) The rectangle formed by the diagonal vertices is shown in fig. 3.
Wherein the diagonal length BbDiagnonal can be given by the following formula:
2. initializing the sampling Point distance D to be 0, and setting the first original sampling Point to be Point0Added to the set of resample points rppoints.
3. Calculating two adjacent Points in the original sampling Point set Points by adopting the formulai，Pointi-1The euclidean distance D between, D is added to D (i.e., D + D).
4. If D is less than the threshold T, then i is incremented by 1 and the process jumps back to step 3, where T is BbDiagnonal 1/40.
5. Otherwise, a new Point is generatedjThe PointjThe distance from the last resampling Point is approximately equal to T, and Point isjAdded to the set of resample points RPoids and inserted at PointiBefore, jump back to step 3.
II) smoothing the original data:
three continuous Points are taken out from the original sampling Point set Pointsi-1，Pointi，Pointi+1Calculating the average value of three points X and Y, which are Xj，yj. Generating a new PointjThe X and Y are originally set as Xj，yjPoint ofjAdded to the set of smoothed sample points, spiints.
Thirdly, handwriting recognition:
stroke recognition is carried out on the handwriting, and the process is as follows: identifying outline strokes in a total stroke set S1 to obtain an outline stroke set O, searching a redundant stroke set RS1 of the outline strokes related to O in 3 strokes before and after O, and removing the stroke set O and RS1 in S1 to obtain a stroke set S2, namely S2 is S1- (O + RS 1); identifying pointer strokes in a stroke set S2 to obtain a pointer stroke set H, wherein the H comprises an hour hand set H1 and a minute hand set H2, searching redundant strokes related to H1 and H2 in 3 strokes before and after H1 and H2 respectively and classifying the redundant strokes into a pointer stroke redundant set RS2, and removing the stroke set H and RS2 in S2 to obtain a stroke set S3, namely S3 is S2- (H + RS 2); digital strokes are identified in stroke set S3. The method comprises the following specific steps:
One) outline stroke recognition:
through analysis of each drawing clock file, the longest strokes are generally found in the outline strokes. The specific algorithm is as follows:
1. first, the longest Stroke Stroke is found from the Stroke set S1lIt is added to the set of outline strokes O. Judging whether O is closed, if not, continuing to execute the step 2, otherwise, jumping to the step 6;
2. by StrokelRespectively searching N strokes (N takes the experience value of 3) forwards and backwards for the reference as a candidate stroke set C;
3. selecting and Stroke from ClThe Stroke string with the two ends closest to each otheri；
4. If StrokeiAnd StrokelThe distance between the two is less than the threshold value T, then the Stroke is startediAdding O, deleting Stroke from CiAnd continuing to execute the step 5, otherwise jumping to the step 6;
5. judging whether O is closed, if not, executing the step 3, otherwise, continuing to execute the step 6;
6. and saving the Outline and exiting.
Two) pointer stroke recognition:
the main problem with pointer stroke recognition is that the user may use one stroke to draw an hour hand and a minute hand, and to solve this problem, the stroke needs to be divided first. Strokes were segmented using the ShortStraw method (ref: Wolin, Aaron, Brian Eoff, and Tracy hammond. "ShortStraw: a Simple and efficient Corner Finder for polylines." SBM.2008.).
The ShortStraw method is briefly described here: the Stroke strokes are first resampled, and as indicated above, a set of resample points RPoids is obtained. Taking a set of points from RPoids using a sliding window of length W
If it is not good
And point
The ratio R of the distance D between the points and the length L of the Point set is less than the threshold T1, then Point is considered to be PointiAs a division point of the Stroke, fig. 5 is an effect diagram using the shortdraw method, in which (b) is the effect diagram after division.
Where T1 takes on a value that is 0.85 times the mean of the distances between all neighboring points in the set RPoids.
The main algorithm of pointer stroke recognition is as follows: the Center point of the outline stroke closure is Center, strokes with the distance from the Center smaller than a threshold value T2 are searched for to form a candidate stroke set C1 of pointer strokes, all strokes in C1 are segmented by using a ShortStraw method to obtain a new stroke set C2, the longest stroke in C2 is searched for to be a minute hand stroke, a second longest stroke is searched for to be an hour hand stroke, and the minute hand stroke and the hour hand stroke form a pointer stroke set H.
Wherein the value of T2 is 1/10 of the minimum side length of the point set RPoids bounding rectangle.
Three) digital stroke recognition:
the first major problem of digital stroke recognition is the need to group strokes according to their numbers, see fig. 6, where (a) is the original data and (b) is the result of grouping the strokes.
The polymerization method is as follows: the spatial distance D between strokes and the temporal order S of the strokes are used as two measurement variables, and the K-means algorithm is used to obtain the aggregation result of the strokes. The reason for using D and S as the metric variables is: 1) from the structure of the numbers, the strokes in the same number are closely connected, and the strokes between different numbers have larger space distance; 2) in the habit of writing numbers, because the number of strokes in a number is limited (generally 1 to 2 strokes), people often draw one stroke and then draw the next stroke, so that the strokes in the same number are also closely connected in time sequence.
After the strokes are aggregated, a stroke aggregation set Cluster can be obtained, and next, each stroke aggregation in the Cluster needs to be identified, which belongs to symbol identification. There are many methods for symbol Recognition (see Chhabra, Atul K. "Graphic symbol Recognition: An overview." Graphics Recognition Algorithms and systems. Springer Berlin Heidelberg, 1998.68-79.). In the system, from the perspective of implementation convenience, the TabletPC Platform SDK 1.7 is selected as a recognition engine of a single number, but not limited to.
Four) redundant stroke discrimination:
the target Stroke Stroke is judged by the following methodtWhether it is a redundant stroke of stroke set U:
firstly, set of strokes U ═ Stroke1,Stroke2,...StrokenThe strokes in the Chinese character are linked into a backbone drawing Stroke according to the sequenceuI.e. the last Stroke Stroke in U is excludednExterior random Stroke StrokeiThe end point of (2) and the next Strokei+1Is linked to the starting point of (a).
For StrokeuUniformly distributing 10 sampling points and adding the sampling points into a point set UPoids, and calculating all points in the UPoids and strokestMean value D of the distances ofmeanIf D ismeanIf the value is less than the threshold value T3, the Stroke is consideredtAre redundant strokes of stroke set U.
Wherein any point of UPoids and StroketIs defined as the distance from the point to the StroketThe minimum value of the distance of any point in the graph. The value of the threshold T3 is Stroke t1/20 in length.
Respective stroke characteristics of the outline stroke, the pointer stroke, and the number stroke are extracted from the above four steps (see the following steps), and evaluation and scoring are performed as follows based on these stroke characteristics.
Fourthly, scoring the drawing condition
One) outline stroke scoring:
the determination is made using the features mentioned above, wherein the method of determining whether to close is as follows: for Outline stroke Outline, the sampling point set contained in the Outline stroke Outline is Points outlineFrom the first Point0Taking the point L3 back,make up a set of PointsstartFrom the last sampling Point|points|-1Taking L3 Points forward to form a sampling point set PointsendCalculating two sampling point sets Pointsstart，PointsendD1, the distance calculation formula is as follows:
wherein the Distance (Point1, Point2) function represents the euclidean Distance between points Point1 and Point 2. The calculated D1 is compared to T4 and if D1 is not greater than T4, the outline stroke is considered closed.
The threshold T4 takes the value of BbDiagonal 1/20.
Calculating the ratio R2 of the outline stroke length L to the inscribed circle perimeter P of the outline stroke closure by the following method: the length L of the Outline stroke Outline is the Euclidean distance of two adjacent points in the sampling point set which are added in sequence. The inscribed circumference P is pi × min (width, height), where width and height are the width and height of the Outline stroke Outline closure, respectively, and R2 is L/P.
When the following condition (stroke characteristic) is satisfied: 1) there are outline strokes; 2) closing the outline strokes; 3) n1< T5; 4) the ratio of the height H to the width W of the outline stroke closure, R1, 1/T6< ═ R1< ═ T6; 5) the ratio of the outline stroke length L to the inscribed circle perimeter P of the outline stroke closure, R2, 1/T7 [ -R2 [ -T7. The empirical value T5-5 and T6-T7-1.5 are taken from experimental data. At this point, the outline is scored 1, otherwise, the outline is scored 0.
Two) pointer stroke scoring:
the above-mentioned characteristics are used for judging, wherein the method for judging whether the hour hand stroke HourHandd and the minute hand stroke MinuteHandd are intersected is as follows: find the first sampling Point of the Hourward stroke0The last sampling Point1(ii) a First sample Point of Minutehand2The last sampling Point3。
Two strokes are considered to intersect if D2< T4.
The calculation method of the quadrant where the hour hand is located is as follows: obtained using equation (3)
And
rotate clockwise in the same direction, rotate in the same direction, rotada wing 1, the quadrant of the hourward
The calculation method of the Quadrant in which the minute hand is located is the same as that of the hour hand, and the Quadrant in which the minute hand is located is Quadrant 2.
When the following condition (stroke characteristic) is satisfied: 1) the hour hand strokes are intersected with the minute hand strokes; 2) the ratio R3 of the hour hand stroke length HL to the minute hand stroke length ML is less than 1; 3) quadrant quadrirange 1 where the hour hand strokes are located, quadrirange 1 is 4; 4) the Quadrant quadrnt 2 in which the split strokes are located, quadrnt 2 is 1. When the above conditions are met, 1 point is given to the pointer stroke, otherwise, 0 point is given.
Three) scoring the digital strokes:
the determination is made using the above-mentioned features, wherein the method of determining the correct number of digits is as follows: the correctness of the digital stroke Digit is mainly determined by two conditions: 1) the digital value represented by the digital stroke Digit is i, 1 ═ i ═ 12, and i is one and only one; 2) the position of the digital stroke Digit is correct.
Wherein the detection method under the condition 2) comprises the following steps: the Center point of the outline stroke closure, Center1, and the Center point of the digital stroke Digit closure, Center2, are obtained. The Center2-Center1 is used for obtaining a vector
Selecting (0, -1) as a reference vector
The (0, -1) is chosen as the reference because the coordinate system of the liquid crystal digitizer is downward as the positive direction of the Y-axis. Using the following formula
And
rotation angle in the clockwise direction:
after obtaining the rotation angle RotatedAngle, when 1<＝i<When it is equal to 11, if
The digital stroke Digit is correctly positioned. When i is 12, if 0<＝RotatedAngle<L8 or 2 x pi-L8<＝RotatedAngle<2 x pi, the Digit stroke Digit is correctly positioned. Wherein
When the condition is satisfied: if the number N2> of correct digits is T8, the digits are given a score of 1, otherwise the digits are given a score of 0. T8 takes the empirical value of 6.
The detection method and system of the present invention have been described in detail by way of examples, but the specific implementation form of the present invention is not limited thereto, and various obvious changes and modifications can be made by those skilled in the art without departing from the spirit and principles of the present invention. The protection scope of the invention is subject to the claims.
Claims (9)
1. A test and detection method for a painting clock based on a digital pen comprises the following steps:
1) Acquiring handwriting data of a user for clock drawing test;
2) recognizing outline strokes, pointer strokes and digital strokes of the handwriting data, and extracting stroke characteristics; the stroke recognition method comprises the following steps: recognizing outline strokes in a total stroke set S1 to obtain an outline stroke set O, searching a redundant stroke set RS1 of the outline strokes, and removing the stroke set O and RS1 in S1 to obtain a stroke set S2, namely S2 is S1- (O + RS 1); identifying pointer strokes in the stroke set S2 to obtain a pointer stroke set H, searching a redundant stroke set RS2 of the pointer strokes, and removing the stroke set H and the RS2 in S2 to obtain a stroke set S3, namely S3 is S2- (H + RS 2); identifying digital strokes in stroke set S3; the method for searching the redundant strokes comprises the following steps: firstly, linking the strokes into a backbone drawing according to the sequence, uniformly distributing a plurality of sampling points on the backbone drawing, adding the sampling points into a point set, calculating the distance mean value of all points in the point set and the target strokes, and if the distance mean value is less than a threshold value, considering the target strokes as redundant strokes;
3) and detecting the outline strokes, the pointer strokes and the digital strokes according to the stroke characteristics to realize the test of the drawing clock.
2. The method of claim 1, wherein stroke recognition is performed using a RandomForest classifier.
3. A method as claimed in claim 1, wherein the pre-processing of the handwriting data prior to recognition comprises:
1) resampling the handwriting data to ensure that the distances between sampling points are equal to T;
2) and the handwriting data is smoothed, so that the stroke jitter degree is reduced.
4. The method according to claim 1, wherein the outline strokes, the pointer strokes and the digital strokes are detected by a bell scoring mechanism of Beijing collaborating with Hospital, which scores 3 points in total, wherein the outline strokes, the pointer strokes and the digital strokes respectively account for 1 point.
5. The method of claim 1,
the stroke characteristics of the outline stroke include: 1) whether outline strokes exist; 2) whether the outline strokes are closed; 3) the number of strokes N1; 4) the ratio R1 of the height H to the width W of the outline stroke closure; 5) a ratio R2 of contour stroke length L to inscribed circle perimeter P of the contour stroke closure;
the stroke characteristics of the pointer stroke include: 1) whether the hour hand stroke and the minute hand stroke are intersected or not; 2) a ratio R3 of the hour hand stroke length HL to the minute hand stroke length ML; 3) quadrant quadrint 1 where the hour hand strokes are located; 4) quadrant quadrirange 2 in which the split-pin strokes are located;
The stroke characteristics of the digital strokes include: the correct number of digits N2.
6. The method of claim 1, wherein the pointer strokes are segmented using the ShortStraw method to obtain hour hand strokes and minute hand strokes; and aggregating the digital strokes according to different numbers by a K-means algorithm.
7. A digital pen based test system for testing a painting clock for implementing the method of any one of claims 1 to 6, the system comprising:
a digital pen for drawing graphics;
and the information acquisition equipment is connected with the digital pen and comprises a liquid crystal digital screen, a data acquisition application and a data analysis application, handwriting data of the digital pen is input and recorded through the liquid crystal display screen, the handwriting data is acquired through the data acquisition application, and stroke characteristics of the handwriting data are identified and scored through the data analysis application.
8. The system of claim 7, wherein the information gathering device is further configured to display a drawing task, the task comprising:
drawing the outline of the clock, drawing the time number and drawing the pointer.
9. The system of claim 7, wherein Microsoft Tablet PC Platform SDK 1.7 is used as the digital stroke recognition engine.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710100530.7A CN108492870B (en) | 2017-02-23 | 2017-02-23 | Method and system for testing and detecting picture clock based on digital pen |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710100530.7A CN108492870B (en) | 2017-02-23 | 2017-02-23 | Method and system for testing and detecting picture clock based on digital pen |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108492870A CN108492870A (en) | 2018-09-04 |
| CN108492870B true CN108492870B (en) | 2022-01-18 |
Family
ID=63343885
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710100530.7A Active CN108492870B (en) | 2017-02-23 | 2017-02-23 | Method and system for testing and detecting picture clock based on digital pen |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108492870B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109409329B (en) * | 2018-11-14 | 2022-04-22 | 科大讯飞股份有限公司 | Method, device, equipment and storage medium for distinguishing and evaluating graphic subjects |
| CN110384484A (en) * | 2019-07-25 | 2019-10-29 | 科大讯飞股份有限公司 | A kind of pattern evaluation method, device, equipment and storage medium |
| CN110859599B (en) * | 2019-11-26 | 2022-05-20 | 复旦大学 | Automatic cognitive function screening system for cerebrovascular disease nerve injury patients |
| CN111091910B (en) * | 2019-12-17 | 2023-11-03 | 中国科学院自动化研究所 | Intelligent evaluation system based on painting clock test |
| CN113535042B (en) * | 2020-09-18 | 2023-09-22 | 厦门市和家健脑智能科技有限公司 | Method and device for generating cognitive recognition image based on old people |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008115555A1 (en) * | 2007-03-21 | 2008-09-25 | Massachusetts Institute Of Technology | Method and apparatus for measuring representational motions in a medical context |
| CN105046730A (en) * | 2015-07-09 | 2015-11-11 | 北京盛世宣合信息科技有限公司 | Writing trace demonstration method and apparatus applied to writing brush |
| CN105807957A (en) * | 2016-03-01 | 2016-07-27 | 上海卓易科技股份有限公司 | Input method and intelligent pen |
| CN106446566A (en) * | 2016-09-29 | 2017-02-22 | 北京理工大学 | Elderly cognitive function classification method based on random forest |
-
2017
- 2017-02-23 CN CN201710100530.7A patent/CN108492870B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008115555A1 (en) * | 2007-03-21 | 2008-09-25 | Massachusetts Institute Of Technology | Method and apparatus for measuring representational motions in a medical context |
| CN105046730A (en) * | 2015-07-09 | 2015-11-11 | 北京盛世宣合信息科技有限公司 | Writing trace demonstration method and apparatus applied to writing brush |
| CN105807957A (en) * | 2016-03-01 | 2016-07-27 | 上海卓易科技股份有限公司 | Input method and intelligent pen |
| CN106446566A (en) * | 2016-09-29 | 2017-02-22 | 北京理工大学 | Elderly cognitive function classification method based on random forest |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108492870A (en) | 2018-09-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108492870B (en) | Method and system for testing and detecting picture clock based on digital pen | |
| Saikumar et al. | A novel implementation heart diagnosis system based on random forest machine learning technique. | |
| KR101846370B1 (en) | Method and program for computing bone age by deep neural network | |
| CN108056786B (en) | Bone age detection method and device based on deep learning | |
| CN102395320B (en) | Medical apparatus and method for controlling the medical apparatus | |
| Lin et al. | AANet: Adaptive attention network for COVID-19 detection from chest X-ray images | |
| JP2022020620A (en) | Relevance feedback to improve performance of clustering model for clustering patients with similar profiles together | |
| CN112184617B (en) | Spine MRI image key point detection method based on deep learning | |
| CN113420826B (en) | Liver focus image processing system and image processing method | |
| US9256777B2 (en) | Method for recognizing gesture and electronic device | |
| JP2009087038A (en) | Image processor and image processing method | |
| Corradi et al. | Lung ultrasound and B-lines quantification inaccuracy: B sure to have the right solution | |
| JP2014029644A (en) | Similar case retrieval device and similar case retrieval method | |
| CN104382570A (en) | Digitized full-automatic health condition detection device | |
| CN112508884A (en) | Comprehensive detection device and method for cancerous region | |
| CN107970027B (en) | Radial artery detection and human body constitution identification system and method | |
| CN111180026A (en) | Special diagnosis and treatment view system and method | |
| CN117425431A (en) | Electrocardiogram analysis support device, program, electrocardiogram analysis support method, and electrocardiogram analysis support system | |
| CN113674859A (en) | Heart birth defect diagnosis method and system | |
| AU2021102129A4 (en) | Automatic labeling method of emphysema in CT image based on image report | |
| WO2020099941A1 (en) | Application of deep learning for medical imaging evaluation | |
| CN114359194A (en) | Multi-mode stroke infarct area image processing method based on improved U-Net network | |
| Kanna et al. | Cardiac arrhythmia detector using cnn application | |
| CN113178254A (en) | Intelligent medical data analysis method and device based on 5G and computer equipment | |
| Shi et al. | Integrating digital pen devices with traditional medical screening tool for cognitive assessment |
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 |