CN115482547A - Method, storage medium and equipment for realizing examination question correction trace identification - Google Patents
Method, storage medium and equipment for realizing examination question correction trace identification Download PDFInfo
- Publication number
- CN115482547A CN115482547A CN202211274064.1A CN202211274064A CN115482547A CN 115482547 A CN115482547 A CN 115482547A CN 202211274064 A CN202211274064 A CN 202211274064A CN 115482547 A CN115482547 A CN 115482547A
- Authority
- CN
- China
- Prior art keywords
- correction
- topic
- trace
- question
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G06V30/40—Document-oriented image-based pattern recognition
- G06V30/41—Analysis of document content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- 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
- G06V30/40—Document-oriented image-based pattern recognition
- G06V30/41—Analysis of document content
- G06V30/412—Layout analysis of documents structured with printed lines or input boxes, e.g. business forms or tables
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method for identifying examination question correcting traces, a storage medium and equipment, wherein the method comprises the following steps: s1, collecting question data of target data; s2, collecting a page image which is corrected by using a specified correction color as an image to be recognized, and then acquiring the page number of the image to be recognized; and S3, extracting correction traces of the image to be recognized. By the method and the device, the correction trace can be effectively associated with the corresponding question, so that intelligent correction with higher automation degree is facilitated, the habit of a teacher is not limited, and the user experience is facilitated to be improved.
Description
Technical Field
The invention relates to the technical field of intelligent education, in particular to a method, a storage medium and equipment for realizing examination question correction trace identification.
Background
Currently, an intelligent classroom is rapidly developed, an automatic grading function of materials such as test paper and homework appears, but a material correcting function of the test paper and the homework is still to be completed, and in the correcting function of the test paper or the homework, how to associate and bind corrected contents with topics is critical, for example, how to recognize that a correction trace is a correction trace of a second topic and not a correction trace of a third topic. At present, most of teaching and assisting materials are customized according to an identification model, a correction area needs to be appointed, and a teacher can correct in a fixed correction frame to correspond to a topic, so that the method can realize the associated binding of the topic, but the following defects exist: 1. the teaching materials need to be customized according to the recognition model, so that the teaching materials supported by the recognition model are large in limitation. 2. Teachers can only modify the teacher in the appointed frame, so that the modifying habit of teachers is limited, and many teachers are not used.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide a method, a storage medium and equipment for realizing the identification of examination question correcting traces.
In order to achieve the purpose, the invention adopts the following technical scheme:
a method for identifying examination question correcting traces comprises the following specific processes:
s1, collecting question data of target data; the question data comprises a data Id, page numbers of all pages, and question numbers, question stems, answers, analysis and coordinate data in the pages of questions contained in all the pages;
s2, collecting a page image which is corrected by using a specified correction color as an image to be recognized, and then acquiring the page number of the image to be recognized;
s3, performing correction trace extraction on the image to be recognized:
s3.1, converting the image to be identified into an image of an hsv color space, and extracting a correction trace according with a color value range according with a preset correction color value range;
s3.2, carrying out Hough line detection on the correction traces extracted in the step S3.1, and connecting lines within a set distance;
s3.3, acquiring the minimum external rectangular frame of each continuous correction trace, and taking out redundant rectangular frames through a non-maximum inhibition algorithm;
s4, identifying the correction traces in each minimum circumscribed rectangular frame obtained in the step S3 by using a correction trace identification model trained in advance to obtain a correction result corresponding to each correction trace, and matching to obtain the incidence relation between each correction trace and each question:
s4.1, collecting a large amount of correction trace sample data, and performing neural network learning training by using the sample data to obtain a correction trace identification model for identifying correction results corresponding to different correction traces;
s4.2, obtaining the coordinate data range of each minimum external rectangular frame obtained in the step S3 and the coordinate data range of each topic contained in the page of the current page number in the topic data, and matching to obtain the topic associated with each correction mark;
and S4.3, identifying by using the correction trace identification model to obtain a correction result corresponding to each correction trace.
Further, the specific process of step S1 is as follows:
s1.1, acquiring original page images of all pages of target data, performing rectangle correction and size resetting operation on all the original page images, and setting the original page images to be uniform in size;
s1.2, marking the areas of the big questions and the small questions under each big question in the page image processed in the step S1.1, and obtaining the coordinates of the corresponding areas.
Furthermore, in step S1.2, area division is performed on each topic in the page image by using a horizontal line marking method, and area division is performed on the small topics in each topic by using a rectangular frame-out method; when the big questions are divided into regions, marking the initial horizontal position and the ending horizontal position of each big question on a page image by using transverse lines; when a rectangular division mode is adopted, each subtotal under the big topic is framed by a rectangular frame; after the area division is completed, for each topic, the coordinates of the upper left corner of the horizontal line of the starting horizontal position and the ending horizontal position of each topic are respectively collected as the coordinate data of the corresponding topic, and for each topic, the coordinates of the upper left corner and the lower right corner of the corresponding rectangular frame are collected as the coordinate data of the corresponding topic.
Further, the specific process of step S4.2 is: calculating the area of the minimum circumscribed rectangle corresponding to each correction trace; for a certain correcting trace, calculating the area of an intersection region of the coordinate data range of the minimum circumscribed rectangle corresponding to the correcting trace and each track, and then calculating the ratio of the area of the intersection region to the area of the minimum circumscribed rectangle frame corresponding to the correcting trace to obtain the intersection ratio of the minimum circumscribed rectangle frame corresponding to the correcting trace and each track; and when the intersection ratio of the minimum circumscribed rectangle frame corresponding to the correction trace and a certain road question exceeds a preset ratio threshold value, the correction trace is considered to be associated with the road question.
Further, the method further comprises step S5: and counting the number of correction results in the coordinate data range of each big question and the number of correct correction results, thereby calculating the accuracy of each big question.
The invention also provides a computer-readable storage medium having stored therein a computer program which, when executed by a processor, implements the method of any one of claims 1 to 5.
The invention also provides a computer apparatus comprising a processor and a memory for storing a computer program; the processor, when executing the computer program, is configured to perform the method of any of claims 1-5.
The invention has the beneficial effects that: by the method and the device, the correction trace can be effectively associated with the corresponding question, so that intelligent correction with higher automation degree is facilitated, the habit of a teacher is not limited, and the user experience is facilitated to be improved.
Detailed Description
The present invention will be further described below, and it should be noted that the present embodiment is based on the technical solution, and a detailed implementation manner and a specific operation process are given, but the protection scope of the present invention is not limited to the present embodiment.
The embodiment provides a method for identifying examination question correcting traces, which comprises the following specific processes:
s1, collecting question data of target data; the question data comprises a data Id, the page number of each page, and the question number, the question stem, the answer, the resolution and the coordinate data in the page of the question contained in each page. Specifically, the specific process of step S1 is as follows:
s1.1, acquiring original page images of all pages of target data, performing rectangle correction and resizing operation on all the original page images, and setting the original page images to be uniform in size.
S1.2, marking the areas of the big questions and the small questions under each big question in the page image processed in the step S1.1, and obtaining the coordinates of the corresponding areas.
In this embodiment, a horizontal line marking mode is adopted to perform area division on each large topic in a page image, and a rectangular framing mode is adopted to perform area division on a small topic in each large topic; when the big questions are divided into regions, marking the initial horizontal position and the ending horizontal position of each big question on a page image by using transverse lines; when a rectangular division mode is adopted, each subtotal under the big topic is framed by a rectangular frame; after the area division is completed, for each topic, the coordinates of the upper left corner of the horizontal line of the starting horizontal position and the ending horizontal position of each topic are respectively collected as the coordinate data of the corresponding topic, and for each topic, the coordinates of the upper left corner and the lower right corner of the corresponding rectangular frame are collected as the coordinate data of the corresponding topic.
S2, collecting the page image which is corrected by the specified correction color as the image to be recognized, and then obtaining the page number of the image to be recognized. The image page number acquisition technology is mature, and is not described herein again.
S3, performing correction trace extraction on the image to be recognized:
s3.1, converting the image to be identified into an image of an hsv color space, and extracting a correction trace according with the color value range according with the preset correction color value range (in a general scene, a teacher uses red for correction).
And S3.2, carrying out Hough line detection on the correction traces extracted in the step S3.1, and connecting lines within a set distance. The color extraction process may have discontinuous correction marks, and the step is used for continuously connecting the discontinuous correction marks.
And S3.3, acquiring the minimum external rectangular frame of each continuous correction trace, and taking out redundant rectangular frames through a non-maximum inhibition algorithm (a plurality of rectangular frames may appear on one correction trace, and the redundant rectangular frames can be removed by adopting the algorithm, so that only one rectangular frame is ensured on one correction trace.
And S4, recognizing the correction traces in the minimum circumscribed rectangular frames obtained in the step S3 by using the correction trace recognition model trained in advance, and obtaining correction results corresponding to the correction traces. The specific process is as follows:
s4.1, collecting a large amount of correction trace sample data (such as correct mark representation and wrong mark representation and/or circle representation), utilizing the sample data to carry out neural network learning training to obtain a correction trace identification model for identifying correction results corresponding to different correction traces (when a correction trace is received again, the trace can be identified to represent correct or wrong)
S4.2, obtaining the coordinate data range of each minimum external rectangular frame obtained in the step S3 and the coordinate data range of each topic contained in the page of the current page number in the topic data, and matching to obtain the topic associated with each correction mark;
in this embodiment, the specific process of step S4.2 is: calculating the area of the minimum circumscribed rectangle corresponding to each correction trace; for a certain correcting trace, calculating the area of an intersection region of the coordinate data range of the minimum circumscribed rectangle corresponding to the correcting trace and each track, and then calculating the ratio of the area of the intersection region to the area of the minimum circumscribed rectangle frame corresponding to the correcting trace to obtain the intersection ratio of the minimum circumscribed rectangle frame corresponding to the correcting trace and each track; and when the intersection ratio of the minimum circumscribed rectangle frame corresponding to the correction trace and a certain road question exceeds a preset ratio threshold value, the correction trace is considered to be associated with the road question.
It should be noted that the purpose of matching the correction traces and the titles by using the intersection is that a part of teacher correction traces are larger, and the correction traces of the current title are often continued to the previous title or the next title, so that noise is generated in the correction of the previous title or the next title, and the correction striving rate is influenced.
And S4.3, identifying by using the correction trace identification model to obtain a correction result corresponding to each correction trace.
In this embodiment, the method further includes step S5: and counting the number of correction results in the coordinate data range of each big question and the number of correct correction results, thereby calculating the accuracy of each big question. For example, if a topic includes four small topics and there are three hooks and one cross correction trace, the statistical result is 3 correct topics and 1 wrong topic.
Various corresponding changes and modifications can be made by those skilled in the art based on the above technical solutions and concepts, and all such changes and modifications should be included in the protection scope of the present invention.
Claims (7)
1. A method for identifying examination question correcting traces is characterized by comprising the following specific processes:
s1, acquiring question data of target data; the question data comprises a data Id, page numbers of all pages, and question numbers, question stems, answers, analysis and coordinate data in the pages of questions contained in all the pages;
s2, collecting a page image which is corrected by using a specified correction color as an image to be recognized, and then acquiring the page number of the image to be recognized;
s3, performing correction trace extraction on the image to be recognized:
s3.1, converting the image to be identified into an image of an hsv color space, and extracting correction traces according with the color value range according with the preset color value range of correction colors;
s3.2, carrying out Hough line detection on the correction traces extracted in the step S3.1, and connecting lines within a set distance;
s3.3, acquiring the minimum external rectangular frame of each continuous correction trace, and taking out redundant rectangular frames through a non-maximum inhibition algorithm;
s4, identifying the correction traces in each minimum circumscribed rectangular frame obtained in the step S3 by using a correction trace identification model trained in advance to obtain a correction result corresponding to each correction trace, and matching to obtain the incidence relation between each correction trace and each question:
s4.1, collecting a large amount of sample data of the correction marks, and performing neural network learning training by using the sample data to obtain a correction mark identification model for identifying correction results corresponding to different correction marks;
s4.2, obtaining the coordinate data range of each minimum circumscribed rectangular frame obtained in the step S3 and the coordinate data range of each topic contained in the current page of the current page number in the topic data, and matching to obtain the topic associated with each correction mark;
and S4.3, identifying by using an identifying model of the correction marks to obtain correction results corresponding to the correction marks.
2. The method according to claim 1, wherein the specific process of step S1 is as follows:
s1.1, acquiring original page images of all pages of target data, performing rectangle correction and size resetting operation on all the original page images, and setting the original page images to be uniform in size;
s1.2, marking the areas of the big questions and the small questions under each big question in the page image processed in the step S1.1, and obtaining the coordinates of the corresponding areas.
3. The method according to claim 2, characterized in that in step S1.2, area division is performed on each large topic in the page image by means of horizontal line marking, and area division is performed on the small topics in each large topic by means of rectangular framing; when dividing the area of the big topics, marking the initial horizontal position and the ending horizontal position of each big topic on the page image by using a horizontal line; when a rectangular division mode is adopted, each subtotal under the big topic is framed by a rectangular frame; after the area division is completed, for each topic, the coordinates of the upper left corner of the horizontal line of the starting horizontal position and the ending horizontal position of each topic are respectively collected as the coordinate data of the corresponding topic, and for each topic, the coordinates of the upper left corner and the lower right corner of the corresponding rectangular frame are collected as the coordinate data of the corresponding topic.
4. The method according to claim 1, wherein the specific process of step S4.2 is: calculating the area of the minimum circumscribed rectangle corresponding to each correction trace; for a certain correcting trace, calculating the area of an intersection region of the coordinate data range of the minimum circumscribed rectangle corresponding to the correcting trace and each track, and then calculating the ratio of the area of the intersection region to the area of the minimum circumscribed rectangle frame corresponding to the correcting trace to obtain the intersection ratio of the minimum circumscribed rectangle frame corresponding to the correcting trace and each track; and when the intersection ratio of the minimum circumscribed rectangle frame corresponding to the correction trace and a certain road question exceeds a preset ratio threshold value, the correction trace is considered to be associated with the road question.
5. The method according to claim 1, further comprising step S5: and counting the number of correction results in the coordinate data range of each big question and the number of correct correction results, thereby calculating the accuracy of each big question.
6. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method of any one of claims 1 to 5.
7. A computer device comprising a processor and a memory, the memory for storing a computer program; the processor is adapted to carry out the method of any one of claims 1 to 5 when executing the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211274064.1A CN115482547A (en) | 2022-10-18 | 2022-10-18 | Method, storage medium and equipment for realizing examination question correction trace identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211274064.1A CN115482547A (en) | 2022-10-18 | 2022-10-18 | Method, storage medium and equipment for realizing examination question correction trace identification |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115482547A true CN115482547A (en) | 2022-12-16 |
Family
ID=84395126
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211274064.1A Pending CN115482547A (en) | 2022-10-18 | 2022-10-18 | Method, storage medium and equipment for realizing examination question correction trace identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115482547A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI821081B (en) * | 2022-12-22 | 2023-11-01 | 倍利科技股份有限公司 | Medical image paging system |
-
2022
- 2022-10-18 CN CN202211274064.1A patent/CN115482547A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI821081B (en) * | 2022-12-22 | 2023-11-01 | 倍利科技股份有限公司 | Medical image paging system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109815932B (en) | Test paper correcting method and device, electronic equipment and storage medium | |
| CN108171297B (en) | Answer sheet identification method | |
| CN110956138B (en) | Auxiliary learning method based on home education equipment and home education equipment | |
| CN110085068A (en) | Learning tutoring method and device based on image recognition | |
| CN113159014B (en) | Objective question reading method, device, equipment and storage medium based on handwritten question number | |
| CN111104883B (en) | Job answer extraction method, apparatus, device and computer readable storage medium | |
| CN113793538B (en) | Method and system for collecting teaching assistance operation wrong questions | |
| CN107067399A (en) | A kind of paper image segmentation processing method | |
| CN115482547A (en) | Method, storage medium and equipment for realizing examination question correction trace identification | |
| CN111008594A (en) | Error correction evaluation method, related equipment and readable storage medium | |
| CN109284702B (en) | Answer sheet scoring and marking system based on image mode | |
| CN115757702A (en) | Choice correcting method, storage medium and equipment | |
| CN115620332B (en) | Automatic reading and amending method and equipment based on paper operation | |
| CN115564267A (en) | Knowledge point mastering degree identification method, storage medium and equipment | |
| CN116704606A (en) | Physicochemical experiment operation behavior identification method, system, device and storage medium | |
| CN114140282B (en) | Method and device for quickly reviewing answers of general teaching classroom based on deep learning | |
| CN110956174A (en) | Device number identification method | |
| CN115601768A (en) | Method, device and equipment for judging written characters and storage medium | |
| CN113486786B (en) | Automatic operation modifying system | |
| CN115482535A (en) | Test paper automatic correction method, storage medium and equipment | |
| CN114299523A (en) | Auxiliary operation identification and correction analysis method and analysis system | |
| CN113033480A (en) | Answer sheet-based objective question reading method, device, equipment and storage medium | |
| CN112613500A (en) | Campus dynamic scoring system based on deep learning | |
| CN114241503B (en) | Method and system for acquiring error cause, readable storage medium and device | |
| CN114937275A (en) | Correction trace identification method based on target detection |
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 |