CN111554149B - System and method for copybook scoring - Google Patents

System and method for copybook scoring Download PDF

Info

Publication number
CN111554149B
CN111554149B CN202010414909.7A CN202010414909A CN111554149B CN 111554149 B CN111554149 B CN 111554149B CN 202010414909 A CN202010414909 A CN 202010414909A CN 111554149 B CN111554149 B CN 111554149B
Authority
CN
China
Prior art keywords
chinese character
written
image
ith
copybook
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
Application number
CN202010414909.7A
Other languages
Chinese (zh)
Other versions
CN111554149A (en
Inventor
秦宇恒
秦毅
张家赫
才亨坤
臧文斌
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Heilongjiang Deya Culture Media Co ltd
Original Assignee
Heilongjiang Deya Culture Media Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Heilongjiang Deya Culture Media Co ltd filed Critical Heilongjiang Deya Culture Media Co ltd
Priority to CN202010414909.7A priority Critical patent/CN111554149B/en
Publication of CN111554149A publication Critical patent/CN111554149A/en
Application granted granted Critical
Publication of CN111554149B publication Critical patent/CN111554149B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B11/00—Teaching hand-writing, shorthand, drawing, or painting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
    • G06V20/63—Scene text, e.g. street names
    • 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
    • G06V30/24—Character recognition characterised by the processing or recognition method
    • G06V30/242—Division of the character sequences into groups prior to recognition; Selection of dictionaries
    • G06V30/244—Division of the character sequences into groups prior to recognition; Selection of dictionaries using graphical properties, e.g. alphabet type or font
    • G—PHYSICS
    • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B7/00—Electrically-operated teaching apparatus or devices working with questions and answers
    • G09B7/02—Electrically-operated teaching apparatus or devices working with questions and answers of the type wherein the student is expected to construct an answer to the question which is presented or wherein the machine gives an answer to the question presented by a student

Abstract

A system and a method for copybook scoring belong to the technical field of Chinese character writing evaluation. The invention solves the problems that the evaluation result of the copybook written well by the existing method is not accurate and the calligraphy practicing effect cannot be measured. According to the method, the interesting region in the image is extracted, the position of the interesting region is corrected, the characteristics of each character in the written copybook are extracted, the extracted characteristics are compared with the characteristics of the copybook template, and the grading result of the written copybook is obtained. The method does not need to manually score, avoids the influence of human subjective factors, and can effectively improve the accuracy of evaluation. And the evaluation results of each time are stored and can be used for grading comparison of each writing so as to measure the calligraphy practicing effect. The invention can be applied to writing evaluation.

Description

System and method for copybook scoring
Technical Field
The invention belongs to the technical field of Chinese character writing evaluation, and particularly relates to a system and a method for copybook scoring.
Background
The characters are tools for recording languages, expressing and communicating ideas, the expression effect can be enhanced by writing the characters correctly, rightly and beautifully, the Chinese character writing level is improved, and the method is also an important aspect for improving the ethnic character quality.
Writing is the continuation and deepening of literacy, is an important component of primary school Chinese teaching, and is also the most essential function that primary school students should have. For elementary school students, writing is the most important basis, the most important learning habit and learning ability.
In the process of examination reading of various examinations, good writing can also give a good impression, so that it is important for elementary students to develop good writing habits from practice of calligraphy. At present, most copybooks are used for calligraphy practice, then the written copybooks are evaluated manually, and the manual evaluation mode has certain subjectivity, so that the evaluation result is inaccurate, the calligraphy practice effect cannot be compared, and the calligraphy practice effect cannot be measured.
Disclosure of Invention
The invention aims to solve the problems that the evaluation result of a written copybook is inaccurate and the calligraphy practicing effect cannot be measured by adopting the conventional method, and provides a system and a method for copybook scoring.
The technical scheme adopted by the invention for solving the technical problems is as follows:
based on one aspect of the invention, the system for signature scoring comprises an image acquisition module, a neural network module, a position correction module, a feature extraction module, a comparison scoring module and a data storage module, wherein:
the image acquisition module is used for acquiring images of copybooks written by a practicer, and the Chinese characters in the copybooks are respectively positioned in each field character lattice; inputting the collected image into a neural network module;
the neural network module is used for identifying the input image and extracting an interested area image;
the position correction module is used for correcting the position of the image of the region of interest to obtain an image after position correction; inputting the image after the position correction into a characteristic extraction module;
the characteristic extraction module is used for extracting the characteristics of the input image so as to extract the characteristics of the input image; inputting the extracted features into a contrast scoring module;
the comparison and scoring module is used for comparing the input characteristics with the copybook template characteristics and outputting a scoring result of the written copybook;
the data storage module is used for storing the scoring result.
According to another aspect of the present invention, a scoring method for a system for scoring a copybook specifically includes the following steps:
after a practicer writes Chinese characters in each field character lattice of the copybook, the image acquisition module is used for acquiring images of the written copybook and inputting the acquired images into the neural network module;
the neural network module identifies the input image and extracts an interested area image;
then, the position correction module is used for carrying out position correction on the image of the region of interest to obtain an image after the position correction; performing feature extraction on each character in the image after the position correction by using a feature extraction module, and extracting features corresponding to each character;
the comparison and scoring module compares the extracted features with the features of each Chinese character in the copybook template to obtain a scoring result of the written copybook; and then the obtained scoring result is stored in a data storage module.
The invention has the beneficial effects that: the invention provides a system and a method for signature grading, which firstly extract an interested region in an image, then extract the characteristics of each character in a written signature after correcting the position of the interested region, and obtain the grading result of the written signature after comparing the extracted characteristics with the characteristics of a signature template.
The method does not need to manually score, avoids the influence of human subjective factors, and can effectively improve the accuracy of evaluation. And the evaluation results of each time are stored and can be used for grading comparison of each writing so as to measure the calligraphy practicing effect.
Drawings
FIG. 1 is a block diagram of a system for signature scoring of the present invention.
Detailed Description
First embodiment this embodiment will be described with reference to fig. 1. The system for grading the copybook in the embodiment comprises an image acquisition module, a neural network module, a position correction module, a feature extraction module, a comparison grading module and a data storage module, wherein:
the image acquisition module is used for acquiring images of copybooks written by a practicer, and the Chinese characters in the copybooks are respectively positioned in each field character lattice; inputting the collected image into a neural network module;
the neural network module is used for identifying the input image and extracting an interested area image;
the position correction module is used for correcting the position of the image of the region of interest to obtain an image after position correction; inputting the image after the position correction into a characteristic extraction module;
the characteristic extraction module is used for extracting the characteristics of the input image so as to extract the characteristics of the input image; inputting the extracted features into a contrast scoring module;
the comparison and scoring module is used for comparing the input characteristics with the copybook template characteristics and outputting a scoring result of the written copybook;
the data storage module is used for storing the scoring result.
The copybook comprises a plurality of rows and columns of field lattices, a practicer writes Chinese characters in each field lattice of the copybook, the image acquisition module uses a high-resolution camera to shoot the written copybook, the neural network module utilizes a neural network model to extract images of regions of interest, the images of the regions of interest comprise all field lattice regions, and because the shooting angle of the camera possibly has certain errors, the angles of the images of the regions of interest need to be corrected.
After the interested region image is extracted, position correction is carried out on the interested region image according to the vertex coordinate of the uppermost left corner grid, the vertex coordinate of the uppermost right corner grid, the vertex coordinate of the lowermost left corner grid and the vertex coordinate of the lowermost right corner grid. After correction, the features of each word can be extracted. And scoring each character according to the extracted characteristics to finally obtain the score of the whole copybook, and storing the score of the copybook written this time for comparison with writing in the future, so that the calligraphy practicing effect is convenient to measure, and meanwhile, the method can also be used for comparison among different practicers, for example, in a calligraphy match, the method can be used for scoring of a player, the influence of human factors is reduced, and the evaluation is more objective, fair and accurate.
The second embodiment is as follows: the difference between this embodiment and the first embodiment is that the position correction module is configured to perform position correction on the image of the region of interest to obtain a position-corrected image, and specifically includes:
in the interested region image, judging the offset angle of the interested region image according to the coordinate of the uppermost left square grid and the coordinate of the uppermost right square grid of the copybook, and then carrying out position correction on the interested region image according to the offset angle to obtain the interested region image after position correction, thereby ensuring that the region of the square grid in the interested region image after position correction is correct.
The third concrete implementation mode: the second difference between this embodiment and the second embodiment is that the feature extraction module is configured to perform feature extraction on an input image to extract features of the input image, where the features include an upper left vertex coordinate, a lower left vertex coordinate, an upper right vertex coordinate, and a lower right vertex coordinate of each stroke of a Chinese character in the input image.
The coordinates of the top left corner vertex, the bottom left corner vertex, the top right corner vertex and the bottom right corner vertex of each Chinese character stroke refer to: and coordinates in a coordinate system which takes the vertex of the leftmost lower corner grid in the region of interest as the origin, the horizontal right direction as the positive direction of the x axis and the vertical upward direction as the positive direction of the y axis.
The fourth concrete implementation mode: the third difference between this embodiment and the specific embodiment is that the comparison and scoring module is configured to compare the input features with the feature of the copybook template, and output a scoring result of the written copybook, and specifically includes:
for the region of interest after position correctionThe vertex coordinates of the ith Chinese character in the domain image and the top left corner of the stroke of the ith Chinese character are
Figure BDA0002494637270000038
The vertex coordinate of the lower left corner is
Figure BDA0002494637270000031
Vertex coordinates at the upper right corner of
Figure BDA0002494637270000032
The vertex coordinate of the lower right corner is
Figure BDA0002494637270000033
Then calculate
Figure BDA0002494637270000034
The value of (a), wherein,
Figure BDA0002494637270000035
is composed of
Figure BDA0002494637270000036
And
Figure BDA0002494637270000037
the maximum value of (a) is,
Figure BDA0002494637270000041
is composed of
Figure BDA0002494637270000042
And
Figure BDA0002494637270000043
the minimum value of (a) to (b),
Figure BDA0002494637270000044
is composed of
Figure BDA0002494637270000045
And
Figure BDA0002494637270000046
the maximum value of (a) is,
Figure BDA0002494637270000047
is composed of
Figure BDA0002494637270000048
Figure BDA0002494637270000049
And
Figure BDA00024946372700000410
minimum value of (1), to be calculated
Figure BDA00024946372700000411
The value of (b) is taken as the size of the ith Chinese character;
calculating the sizes of other Chinese characters in the region-of-interest image after position correction, and solving the mean square deviation D of the size values of all the Chinese characters;
then will be
Figure BDA00024946372700000412
As the central position of the ith Chinese character written;
in the same way, the central position of each Chinese character in the copybook template is calculated;
calculating the distance between the center position of the ith Chinese character in the region-of-interest image after position correction and the center position of the ith Chinese character in the copybook template; then, averaging the distances corresponding to the Chinese characters to obtain an average value M of the distances;
subtracting the mean square error D and the mean value M from the basic score Q to obtain an initial score Q1 of the well-written copybook;
translating the written ith Chinese character according to the calculated central position of the written ith Chinese character to ensure that the center of the written ith Chinese character is superposed with the center of the ith Chinese character in the copybook template;
after translating the written ith Chinese character, extracting a feature vector of the written ith Chinese character by using a feature extraction module; calculating the Euclidean distance between the ith written Chinese character and the ith Chinese character of the template;
introducing a ranking mechanism, adding the Euclidean distance corresponding to the ith written Chinese character into a database of the system, and performing descending order arrangement on all the Euclidean distances corresponding to the character in the database to obtain the ordering serial number L of the Euclidean distance corresponding to the ith written Chinese characteri；
Simultaneously respectively calculating the sorting sequence number L of the Euclidean distance corresponding to each Chinese character written at the timeiI is 1,2, …, I represents the total number of the written copybooks; then averaging the sequence numbers corresponding to each Chinese character to obtain
Figure BDA00024946372700000413
The initial score Q1 is compared with
Figure BDA00024946372700000414
And (5) making a sum to obtain the final score of the copybook written at the time.
The copybook template is characterized in that: and establishing a copybook template coordinate system by taking the vertex of the matts at the leftmost lower corner of the copybook template as the origin, taking the horizontal right direction as the positive direction of an x 'axis and the vertical upward direction as the positive direction of a y' axis, and respectively obtaining the vertex coordinates of the upper left corner, the lower left corner, the upper right corner and the lower right corner of each Chinese character in the copybook template under the copybook template coordinate system.
According to the embodiment, the size of each written Chinese character is calculated firstly, then the mean square error of each Chinese character is calculated, the uniformity of the size of the written Chinese characters can be well measured, when the written Chinese characters are large and small, punishment of scoring can be carried out, the uniformity of the size of the written Chinese characters can be effectively evaluated, and the uniformity of writing is improved.
By comparing the difference between the central position of the written Chinese character and the central position of the template Chinese character, the written deviation degree can be measured in a balanced manner, and the writing quality is improved.
And translating the written Chinese characters according to the calculated central position, extracting the characteristic vector by using a CNN network, and calculating the Euclidean distance. By translating the process, the calculation of the Euclidean distance is not influenced by the position offset of the Chinese characters, and the evaluation result is more accurate. For example, in the case of not considering the position offset, the euclidean distance corresponding to some chinese characters is very small, but the calculated euclidean distance is too large due to the offset of the center position.
The writing evaluation method of the embodiment can ensure the evaluation accuracy, and compared with the evaluation method only performing x-axis deviation or y-axis deviation, the method is more accurate because large x-axis deviation or large y-axis deviation does not necessarily represent that the writing quality is not good, and conversely, small deviation does not necessarily represent that the writing quality is not good, and comprehensive evaluation is performed.
The fifth concrete implementation mode: the difference between the embodiment mode and the specific embodiment mode is that the system further comprises a Chinese character writing module and a writing timing module, wherein the Chinese character writing module is used for a practicer to directly write, and the writing timing module is used for timing writing.
The sixth specific implementation mode: the scoring method for the system for scoring a copybook according to the first embodiment of the present invention specifically includes the following steps:
after a practicer writes Chinese characters in each field character lattice of the copybook, the image acquisition module is used for acquiring images of the written copybook and inputting the acquired images into the neural network module;
the neural network module identifies the input image and extracts an interested area image;
then, the position correction module is used for carrying out position correction on the image of the region of interest to obtain an image after the position correction; performing feature extraction on each character in the image after the position correction by using a feature extraction module, and extracting features corresponding to each character;
the comparison and scoring module compares the extracted features with the features of each Chinese character in the copybook template to obtain a scoring result of the written copybook; and then the obtained scoring result is stored in a data storage module.
The scoring of the copybook written at this time is stored for comparison with writing in the future, so that the calligraphy practicing effect is convenient to measure, and meanwhile, the scoring method can also be used for comparison among different practicers, for example, in a calligraphy match, the scoring method can be used for scoring of the practicers, the influence of human factors is reduced, and the evaluation is more objective, fair and accurate.
A seventh specific embodiment, which is different from the sixth specific embodiment, in that the position correction module is used to perform position correction on the image of the region of interest to obtain a position-corrected image, and a specific process of the position-corrected image is as follows:
in the interested region image, judging the offset angle of the interested region image according to the coordinate of the uppermost left square grid and the coordinate of the uppermost right square grid of the copybook, and then carrying out position correction on the interested region image according to the offset angle to obtain the interested region image after position correction, thereby ensuring that the region of the square grid in the interested region image after position correction is correct.
The eighth specific implementation manner and the seventh specific implementation manner are different in that the feature extraction module is used for extracting features of each Chinese character in the position-corrected image, where the features include an upper left corner vertex coordinate, a lower left corner vertex coordinate, an upper right corner vertex coordinate, and a lower right corner vertex coordinate of each Chinese character stroke in the position-corrected image.
The ninth specific embodiment is different from the eighth specific embodiment in that the comparison and scoring module compares the extracted features with the feature of the copybook template to obtain a scoring result of the written copybook; and then the obtained scoring result is stored in a data storage module, and the specific process is as follows:
for the ith Chinese character in the region-of-interest image after the position correction, the vertex coordinate of the top left corner of the stroke of the ith Chinese character is
Figure BDA0002494637270000061
The vertex coordinate of the lower left corner is
Figure BDA0002494637270000062
Vertex coordinates at the upper right corner of
Figure BDA0002494637270000063
The vertex coordinate of the lower right corner is
Figure BDA0002494637270000064
Then calculate
Figure BDA0002494637270000065
The value of (a), wherein,
Figure BDA0002494637270000066
is composed of
Figure BDA0002494637270000067
And
Figure BDA0002494637270000068
the maximum value of (a) is,
Figure BDA0002494637270000069
is composed of
Figure BDA00024946372700000610
And
Figure BDA00024946372700000611
the minimum value of (a) to (b),
Figure BDA00024946372700000612
is composed of
Figure BDA00024946372700000613
And
Figure BDA00024946372700000614
the maximum value of (a) is,
Figure BDA00024946372700000615
is composed of
Figure BDA00024946372700000616
Figure BDA00024946372700000617
And
Figure BDA00024946372700000618
minimum value of (1), to be calculated
Figure BDA00024946372700000619
The value of (b) is taken as the size of the ith Chinese character;
calculating the sizes of other Chinese characters in the region-of-interest image after position correction, and solving the mean square deviation D of the size values of all the Chinese characters;
then will be
Figure BDA00024946372700000620
As the central position of the ith Chinese character written;
in the same way, the central position of each Chinese character in the copybook template is calculated;
calculating the distance between the center position of the ith Chinese character in the region-of-interest image after position correction and the center position of the ith Chinese character in the copybook template; then, averaging the distances corresponding to the Chinese characters to obtain an average value M of the distances;
subtracting the mean square error D and the mean value M from the basic score Q to obtain an initial score Q1 of the well-written copybook;
translating the written ith Chinese character according to the calculated central position of the written ith Chinese character to ensure that the center of the written ith Chinese character is superposed with the center of the ith Chinese character in the copybook template;
after translating the written ith Chinese character, extracting a feature vector of the written ith Chinese character by using a feature extraction module; calculating the Euclidean distance between the ith written Chinese character and the ith Chinese character of the template;
introducing a ranking mechanism, adding the Euclidean distance corresponding to the ith written Chinese character into a database of the system, and performing descending order arrangement on all the Euclidean distances corresponding to the character in the database to obtain the ordering serial number L of the Euclidean distance corresponding to the ith written Chinese characteri；
Simultaneously respectively calculating the sorting sequence number L of the Euclidean distance corresponding to each Chinese character written at the timeiI is 1,2, …, I represents the total number of the written copybooks; then averaging the sequence numbers corresponding to each Chinese character to obtain
Figure BDA0002494637270000071
The initial score Q1 is compared with
Figure BDA0002494637270000072
Performing a sum to obtain a final score of the copybook which is written at the time;
and the scoring result of the copybook written at the current time is stored in the data storage module, so that the scoring comparison of each writing is convenient.
And measuring the calligraphy practicing effect according to the grading comparison of the copybook which is written each time.
The above-described calculation examples of the present invention are merely to explain the calculation model and the calculation flow of the present invention in detail, and are not intended to limit the embodiments of the present invention. It will be apparent to those skilled in the art that other variations and modifications of the present invention can be made based on the above description, and it is not intended to be exhaustive or to limit the invention to the precise form disclosed, and all such modifications and variations are possible and contemplated as falling within the scope of the invention.

Claims (7)

1. A system for signature scoring, the system comprising an image acquisition module, a neural network module, a location correction module, a feature extraction module, a contrast scoring module, and a data storage module, wherein:
the image acquisition module is used for acquiring images of copybooks written by a practicer, and the Chinese characters in the copybooks are respectively positioned in each field character lattice; inputting the collected image into a neural network module;
the neural network module is used for identifying the input image and extracting an interested area image;
the position correction module is used for correcting the position of the image of the region of interest to obtain an image after position correction; inputting the image after the position correction into a characteristic extraction module;
the characteristic extraction module is used for extracting the characteristics of the input image so as to extract the characteristics of the input image; inputting the extracted features into a contrast scoring module;
the comparison and scoring module is used for comparing the input characteristics with the copybook template characteristics and outputting a scoring result of the written copybook; the method comprises the following steps:
for the ith Chinese character in the region-of-interest image after the position correction, the vertex coordinate of the top left corner of the stroke of the ith Chinese character is
Figure FDA0003367388700000011
The vertex coordinate of the lower left corner is
Figure FDA0003367388700000012
Vertex coordinates at the upper right corner of
Figure FDA0003367388700000013
The vertex coordinate of the lower right corner is
Figure FDA0003367388700000014
Then calculate
Figure FDA0003367388700000015
The value of (a), wherein,
Figure FDA0003367388700000016
is composed of
Figure FDA0003367388700000017
And
Figure FDA0003367388700000018
the maximum value of (a) is,
Figure FDA0003367388700000019
is composed of
Figure FDA00033673887000000110
And
Figure FDA00033673887000000111
the minimum value of (a) to (b),
Figure FDA00033673887000000112
is composed of
Figure FDA00033673887000000113
And
Figure FDA00033673887000000114
the maximum value of (a) is,
Figure FDA00033673887000000115
is composed of
Figure FDA00033673887000000116
Figure FDA00033673887000000117
And
Figure FDA00033673887000000118
minimum value of (1), to be calculated
Figure FDA00033673887000000119
The value of (b) is taken as the size of the ith Chinese character;
calculating the sizes of other Chinese characters in the region-of-interest image after position correction, and solving the mean square deviation D of the size values of all the Chinese characters;
then will be
Figure FDA00033673887000000120
As the central position of the ith Chinese character written;
in the same way, the central position of each Chinese character in the copybook template is calculated;
calculating the distance between the center position of the ith Chinese character in the region-of-interest image after position correction and the center position of the ith Chinese character in the copybook template; then, averaging the distances corresponding to the Chinese characters to obtain an average value M of the distances;
subtracting the mean square error D and the mean value M from the basic score Q to obtain an initial score Q1 of the well-written copybook;
translating the written ith Chinese character according to the calculated central position of the written ith Chinese character to ensure that the center of the written ith Chinese character is superposed with the center of the ith Chinese character in the copybook template;
after translating the written ith Chinese character, extracting a feature vector of the written ith Chinese character by using a feature extraction module; calculating the Euclidean distance between the ith written Chinese character and the ith Chinese character of the template;
introducing a ranking mechanism, adding the Euclidean distance corresponding to the ith written Chinese character into a database of the system, and performing descending order arrangement on all the Euclidean distances corresponding to the character in the database to obtain the ordering serial number L of the Euclidean distance corresponding to the ith written Chinese characteri；
Simultaneously respectively calculating the sorting sequence number L of the Euclidean distance corresponding to each Chinese character written at the timeiI is 1,2, …, I represents the total number of the written copybooks; then averaging the sequence numbers corresponding to each Chinese character to obtain
Figure FDA0003367388700000021
The initial score Q1 is compared with
Figure FDA0003367388700000022
Performing a sum to obtain a final score of the copybook which is written at the time;
the data storage module is used for storing the scoring result.
2. The system for signature scoring as recited in claim 1, wherein the position correction module is configured to perform position correction on the image of the region of interest to obtain a position-corrected image, which is specifically:
in the interested region image, judging the offset angle of the interested region image according to the coordinate of the uppermost left square grid and the coordinate of the uppermost right square grid of the copybook, and then carrying out position correction on the interested region image according to the offset angle to obtain the interested region image after position correction, thereby ensuring that the region of the square grid in the interested region image after position correction is correct.
3. The system of claim 2, wherein the feature extraction module is configured to perform feature extraction on the input image to extract features of the input image, and the features include top left vertex coordinates, bottom left vertex coordinates, top right vertex coordinates, and bottom right vertex coordinates of each stroke of a Chinese character in the input image.
4. The system for signature scoring as recited in claim 3, further comprising a chinese character writing module for direct writing by a learner and a writing timing module for timing writing.
5. A scoring method for a system for signature scoring as claimed in claim 1, said method comprising in particular the steps of:
after a practicer writes Chinese characters in each field character lattice of the copybook, the image acquisition module is used for acquiring images of the written copybook and inputting the acquired images into the neural network module;
the neural network module identifies the input image and extracts an interested area image;
then, the position correction module is used for carrying out position correction on the image of the region of interest to obtain an image after the position correction; performing feature extraction on each character in the image after the position correction by using a feature extraction module, and extracting features corresponding to each character;
the comparison and scoring module compares the extracted features with the features of each Chinese character in the copybook template to obtain a scoring result of the written copybook; then, storing the obtained scoring result in a data storage module; the specific process comprises the following steps:
for the ith Chinese character in the region-of-interest image after the position correction, the vertex coordinate of the top left corner of the stroke of the ith Chinese character is
Figure FDA0003367388700000023
The vertex coordinate of the lower left corner is
Figure FDA0003367388700000024
Vertex coordinates at the upper right corner of
Figure FDA0003367388700000025
The vertex coordinate of the lower right corner is
Figure FDA0003367388700000031
Then calculate
Figure FDA0003367388700000032
The value of (a), wherein,
Figure FDA0003367388700000033
is composed of
Figure FDA0003367388700000034
And
Figure FDA0003367388700000035
the maximum value of (a) is,
Figure FDA0003367388700000036
is composed of
Figure FDA0003367388700000037
And
Figure FDA0003367388700000038
the minimum value of (a) to (b),
Figure FDA0003367388700000039
is composed of
Figure FDA00033673887000000310
And
Figure FDA00033673887000000311
the maximum value of (a) is,
Figure FDA00033673887000000312
is composed of
Figure FDA00033673887000000313
Figure FDA00033673887000000314
And
Figure FDA00033673887000000315
minimum value of (1), to be calculated
Figure FDA00033673887000000316
The value of (b) is taken as the size of the ith Chinese character;
calculating the sizes of other Chinese characters in the region-of-interest image after position correction, and solving the mean square deviation D of the size values of all the Chinese characters;
then will be
Figure FDA00033673887000000317
As the central position of the ith Chinese character written;
in the same way, the central position of each Chinese character in the copybook template is calculated;
calculating the distance between the center position of the ith Chinese character in the region-of-interest image after position correction and the center position of the ith Chinese character in the copybook template; then, averaging the distances corresponding to the Chinese characters to obtain an average value M of the distances;
subtracting the mean square error D and the mean value M from the basic score Q to obtain an initial score Q1 of the well-written copybook;
translating the written ith Chinese character according to the calculated central position of the written ith Chinese character to ensure that the center of the written ith Chinese character is superposed with the center of the ith Chinese character in the copybook template;
after translating the written ith Chinese character, extracting a feature vector of the written ith Chinese character by using a feature extraction module; calculating the Euclidean distance between the ith written Chinese character and the ith Chinese character of the template;
introducing a ranking mechanism, adding the Euclidean distance corresponding to the ith written Chinese character into a database of the system, and performing descending order arrangement on all the Euclidean distances corresponding to the character in the database to obtain the ordering serial number L of the Euclidean distance corresponding to the ith written Chinese characteri；
Simultaneously respectively calculating the sorting sequence number L of the Euclidean distance corresponding to each Chinese character written at the timeiI is 1,2, …, I represents the total number of the written copybooks; then averaging the sequence numbers corresponding to each Chinese character to obtain
Figure FDA00033673887000000318
The initial score Q1 is compared with
Figure FDA00033673887000000319
Performing a sum to obtain a final score of the copybook which is written at the time;
and the scoring result of the copybook written at the current time is stored in the data storage module, so that the scoring comparison of each writing is convenient.
6. The scoring method of claim 5, wherein the position correction module performs position correction on the image of the region of interest to obtain a position-corrected image, and the specific process is as follows:
in the interested region image, judging the offset angle of the interested region image according to the coordinate of the uppermost left square grid and the coordinate of the uppermost right square grid of the copybook, and then carrying out position correction on the interested region image according to the offset angle to obtain the interested region image after position correction, thereby ensuring that the region of the square grid in the interested region image after position correction is correct.
7. The method of claim 6, wherein the feature extraction module is used to extract features of each Chinese character in the position-corrected image, wherein the features include top-left vertex coordinates, bottom-left vertex coordinates, top-right vertex coordinates, and bottom-right vertex coordinates of each Chinese character stroke in the position-corrected image.
CN202010414909.7A 2020-05-15 2020-05-15 System and method for copybook scoring Active CN111554149B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010414909.7A CN111554149B (en) 2020-05-15 2020-05-15 System and method for copybook scoring

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010414909.7A CN111554149B (en) 2020-05-15 2020-05-15 System and method for copybook scoring

Publications (2)

Publication Number Publication Date
CN111554149A CN111554149A (en) 2020-08-18
CN111554149B true CN111554149B (en) 2022-04-29

Family

ID=72004738

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010414909.7A Active CN111554149B (en) 2020-05-15 2020-05-15 System and method for copybook scoring

Country Status (1)

Country Link
CN (1) CN111554149B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112396032B (en) * 2020-12-03 2023-04-07 北京有竹居网络技术有限公司 Writing detection method and device, storage medium and electronic equipment
CN112686212A (en) * 2021-01-26 2021-04-20 吉林省华粹文化传播有限公司 Five-point coordinate positioning grid, positioning method and online writing evaluation system
CN113326009B (en) * 2021-03-05 2022-05-31 临沂大学 Paper calligraphy work copying method and device
CN113033539B (en) * 2021-03-30 2022-12-06 北京有竹居网络技术有限公司 Calligraphy practicing lattice detection method and device, readable medium and electronic equipment
CN116189209B (en) * 2023-04-14 2023-07-04 浙江太美医疗科技股份有限公司 Medical document image classification method and device, electronic device and storage medium

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3870366B2 (en) * 2001-01-25 2007-01-17 研文社印刷株式会社 Work rating test management system
JP2017194612A (en) * 2016-04-21 2017-10-26 京セラドキュメントソリューションズ株式会社 Calligraphy learning support device, calligraphy learning support method, and calligraphy learning program
CN109214471A (en) * 2018-10-10 2019-01-15 北京米蓝科技有限公司 Evaluate the method and system of the written word in copybook of practising handwriting
CN109543777B (en) * 2018-11-28 2020-10-27 中国科学院自动化研究所 Method and system for evaluating writing quality of handwritten Chinese characters
CN110232377A (en) * 2019-01-07 2019-09-13 广东爱贝佳科技有限公司 A kind of artificial intelligence points-scoring system that copybook practices calligraphy and method
CN110503101A (en) * 2019-08-23 2019-11-26 北大方正集团有限公司 Font evaluation method, device, equipment and computer readable storage medium
CN111144270B (en) * 2019-12-23 2023-05-05 智慧神州(北京)科技有限公司 Neural network-based handwritten text integrity evaluation method and evaluation device

Also Published As

Publication number Publication date
CN111554149A (en) 2020-08-18

Similar Documents

Publication Publication Date Title
CN111554149B (en) System and method for copybook scoring
CN108764257B (en) Multi-view pointer instrument identification method
CN202939608U (en) Test paper inspecting system
CN104463101B (en) Answer recognition methods and system for character property examination question
CN109543777B (en) Method and system for evaluating writing quality of handwritten Chinese characters
CN109918523B (en) Circuit board component detection method based on YOLO9000 algorithm
CN109460762B (en) Answer sheet scoring method based on image recognition
CN110298250A (en) A kind of writing scoring and error correction method and interactive system
CN110414563A (en) Total marks of the examination statistical method, system and computer readable storage medium
CN105224921A (en) A kind of facial image preferentially system and disposal route
CN102081742B (en) Method for automatically evaluating writing ability
CN108537218A (en) A kind of identifying processing method and device of answering card
CN109508661A (en) A kind of person's of raising one's hand detection method based on object detection and Attitude estimation
CN103455826B (en) Efficient matching kernel body detection method based on rapid robustness characteristics
CN113191309A (en) Method and system for recognizing, scoring and correcting handwritten Chinese characters
CN115082266A (en) Student education subject comprehensive development analysis and evaluation system based on deep learning
CN107341824A (en) A kind of comprehensive evaluation index generation method of image registration
CN110059705A (en) A kind of OCR recognition result decision method and equipment based on modeling
CN108230322A (en) A kind of eyeground feature detection device based on weak sample labeling
CN210038810U (en) Intelligent evaluation equipment and system
CN111814606A (en) Automatic marking system for technical image processing and pattern recognition and implementation method
CN116959008A (en) Text evaluation model construction method and device, electronic equipment and storage medium
Murch et al. Establishing the quantitative basis for sufficiency thresholds and metrics for friction ridge pattern detail and the foundation for a standard
CN111612045B (en) Universal method for acquiring target detection data set
CN114254605A (en) Answer sheet template generation method, answer sheet identification method and device and electronic equipment

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