CN110689013A - Automatic marking method and system based on feature recognition - Google Patents
Automatic marking method and system based on feature recognition Download PDFInfo
- Publication number
- CN110689013A CN110689013A CN201910959432.8A CN201910959432A CN110689013A CN 110689013 A CN110689013 A CN 110689013A CN 201910959432 A CN201910959432 A CN 201910959432A CN 110689013 A CN110689013 A CN 110689013A
- Authority
- CN
- China
- Prior art keywords
- answer sheet
- scanning
- area
- descriptor
- answer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/24—Aligning, centring, orientation detection or correction of the image
- G06V10/242—Aligning, centring, orientation detection or correction of the image by image rotation, e.g. by 90 degrees
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Electrically Operated Instructional Devices (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an automatic paper marking method and system based on feature recognition, which adopts an ORB algorithm and a RANSAC algorithm to analyze test papers, improves the existing paper marking system, provides a complete answer processing system, comprises a webpage end template manufacturing tool, an answer scanning and recognizing client and a subjective question marking webpage end, almost realizes compatibility on arbitrarily typeset answer papers, provides a solution for the problem of the absence of an answer card without a positioning point or the positioning point caused by printing and scanning, can also be well corrected under the condition that an invigilator receives the answer paper and receives the answer paper, and furthest ensures the reliability of the paper marking.
Description
Technical Field
The invention relates to the technical field of big data analysis application, in particular to an automatic examination paper marking method and system based on feature recognition, which are suitable for examination paper marking.
Background
In the existing paper marking system, two methods are generally used, one method is to use a cursor machine for scanning and marking, and due to the characteristics of rapidness and accuracy, the method is very common in standard examinations at home and abroad, but the method has obvious defects, can only process selection and judgment questions, cannot process subjective questions, has certain requirements on pen materials for filling, filling weight, answer sheet paper quality, printing quality and answer sheet structure, and requires a special cursor reader; the other method is that a general scanner is used for scanning student test paper into an electronic answer paper and then identifying the electronic answer paper, and then positioning points around the test paper and a standard answer sheet made in advance are used for correcting, comparing and cutting according to question blocks.
Disclosure of Invention
In view of the problems described in the background art, the present invention provides an automatic paper marking method and system based on feature recognition, which solves the problems in the prior art and ensures the reliability of paper marking to the maximum extent.
In order to achieve the purpose, the invention provides the following technical scheme:
an automatic scoring method and system based on feature recognition comprises the following steps:
q1, defining the structure of the test paper, counting the information of the test paper, defining the question blocks of each area in the test paper, wherein the question blocks are divided into: a test paper title area, an examination admission card number filling area, an objective question filling area and a subjective question answering area, and points out the scores of all question blocks, a scanner is used for scanning an answer sheet gray level picture of a blank test paper, and the picture is uploaded to a server to serve as an answer sheet template for standard comparison;
q2, storing information, storing the examination numbers and the corresponding names of the students in the server;
q3, scanning and identifying the answer sheet, connecting the answer sheet after the student answers with a local client and controlling a scanner, scanning the answer sheet after the student answers to obtain an answer sheet scanning picture, correcting the front-back sequence of the answer sheet scanning picture, the paper feeding direction of the answer sheet, checking whether the information of the student examination number on the answer sheet is repeated, filled by mistake or not, and then correcting the rotation, stretching, translation, scaling and small distortion of the answer sheet scanning picture;
firstly, extracting key points of an answer sheet template, adopting ORB algorithm for scanning when scanning and correcting the answer sheet template, positioning a group of characteristic points by using FAST (FAST open search algorithm), calculating haar corner measurement of FAST characteristic points,
calculating a feature vector of the FAST feature point through formula one, namely a descriptor a, through the following formula:
and obtaining the direction of the gradient direction relative to the feature center according to a formula II, namely:
θ＝atan2(m01，m10) Formula two
Calculating to obtain a descriptor and storing the descriptor in a memory;
secondly, extracting key points of the answer sheet after answering to obtain a descriptor B;
then, matching the descriptor A with the descriptor B, wherein the descriptor A is defined as a query set, the descriptor B is defined as a training set, the two sets are matched one by one, and the selected distance formula III is as follows:
after the optimal matching is selected, a round of cross validation is carried out, so that matching errors are reduced;
and finally, determining two matched corresponding set points according to the matched descriptors. Determining transformation matrixes of the two sets through a RANSAC algorithm, and correcting student test paper to obtain a final picture;
q4, segmenting the final picture obtained in Q3 through a mask according to the position information of each question block prestored on the server, identifying, positioning each subjective question area and each student information area, and uploading the corrected picture and the identification result to the server;
q5, identify the content of the objective topic filling area, and manually review the content of the subjective topic area.
In the technical scheme, for the examination card number filling area, the examination numbers filled by students are positioned to lattice points in the grid to obtain the examination numbers of the students, and the obtained examination numbers are compared with data in a server library in Q2, if the examination numbers do not exist or need manual verification and correction repeatedly.
In the technical scheme, for the objective question filling area, the unfilled options and the filled options are classified into two categories, and the category with large area is used as the option selected by the student.
The invention adopts ORB algorithm and RANSAC algorithm to analyze the test paper, improves the existing paper marking system, provides a complete answer sheet processing system, comprises a webpage end template manufacturing tool, an answer sheet scanning and identifying client and a subjective question marking webpage end, almost realizes compatibility of arbitrarily typesetted answer sheets, provides a solution for the problem of the absence of answer sheets without positioning points or the positioning points caused by printing and scanning, can be well corrected under the conditions of supervision teacher receiving and receiving the answer sheets, ensures the reliability of paper marking to the maximum extent, saves time for marking the paper, eliminates errors possibly caused by manual participation in paper marking, covers the information of students in a paper marking mode of block paper marking, and ensures the paper marking process to be more fair.
Drawings
FIG. 1 is a schematic diagram of a test paper structure used in the automated paper inspection of the present invention.
Wherein: the examination paper examination system comprises a 1 test paper title area, a 2 admission examination certificate number filling area, a 3 examination missing mark area, a 4 objective question filling area and a 5 subjective question answering area.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings and embodiments of the present invention, and it is obvious that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, an embodiment of the present invention describes an automatic scoring method and system based on feature recognition for the answer sheet shown in fig. 1, including the following steps:
the method comprises the following steps:
q1, test paper structure definition. And carrying out information statistics on the test paper information, defining question blocks of each area in the test paper, forming the question blocks into independent question blocks, indicating the scores of the question blocks, scanning a gray picture of a blank test paper by using a scanner, and uploading the gray picture to a server to be used as a template test paper for standard comparison. The test paper structure of the answer sheet is generally divided into five parts: the method comprises the steps of marking a test paper title area, an examination admission card number filling area, an objective question filling area, a subjective question answering area and an examination missing mark area with a template making tool, then storing test paper structure information into a server, framing the whole objective question filling area in the objective question filling area, indicating the distribution condition of options, namely the options are behind or under the question number, and the surrounding frame type of the options, namely square brackets or square frames, and needing to frame an additional option for identification, positioning the positions of the options and the question numbers by a component detection method, and storing the position information in the server.
Q2, store information. The examination numbers and the corresponding names of the students are stored in the server in advance.
Q3, answer sheet scanning identification. Connecting the answer sheets after the students answer with a local client and controlling a scanner, scanning the answer sheets after the students answer to obtain answer sheet scanning pictures, correcting the front and back sequence of the answer sheet scanning pictures, the paper feeding direction of the answer sheets, checking whether the information of the student examination numbers on the answer sheets is repeated, filled by mistake or not, and then correcting the rotation, stretching, translation, scaling and small distortion of the answer sheet scanning pictures;
firstly, extracting key points of an answer sheet template, adopting ORB algorithm for scanning when scanning and correcting the answer sheet template, positioning a group of characteristic points by using FAST (FAST open search algorithm), calculating haar corner measurement of FAST characteristic points,
calculating a feature vector of the FAST feature point through formula one, namely a descriptor a, through the following formula:
and obtaining the direction of the gradient direction relative to the feature center according to a formula II, namely:
θ＝atan2(m01，m10) Formula two
Calculating to obtain a descriptor and storing the descriptor in a memory;
secondly, extracting key points of the answer sheet after answering to obtain a descriptor B;
then, matching the descriptor A with the descriptor B, wherein the descriptor A is defined as a query set, the descriptor B is defined as a training set, the two sets are matched one by one, and the selected distance formula III is as follows:
after the optimal matching is selected, a round of cross validation is carried out, so that matching errors are reduced;
and finally, determining two matched corresponding set points according to the matched descriptors. Determining transformation matrixes of the two sets through a RANSAC algorithm, and correcting student test paper to obtain a final picture;
for the student test number area, the longitudinal direction is always uniformly distributed by ten numbers of 0-9, the average distribution distance of each longitudinal value can be calculated, meanwhile, the horizontal distance of each filling area is also the same, the minimum distance is easy to obtain, at the moment, the test numbers filled by students can be positioned to grid points in a grid, so that the student test numbers are obtained, the obtained test numbers are compared with the test numbers in a database, if the test numbers do not exist or need manual verification and correction repeatedly, if the test number filling area uses bar codes or two-dimensional codes, the subsequent detection and manual verification of the test numbers are not needed.
For the examination lacking marked area, when the filling area of the student is larger than the specified threshold value, the student is considered to lack the examination, otherwise, filling and answering information collection are carried out.
And for the objective question filling area, classifying the unfilled options and filled options into two categories, taking the category with large area as the options selected by the student, and comparing the item numbers in the server in the step Q1 with the positions of the options to obtain the options selected and unselected by the student.
For the main subject area, the client does not need to process when the scanner scans, and only the picture is uploaded to the server.
Q4, segmenting the final picture obtained in Q3 through a mask according to the position information of each question block prestored on the server, identifying, positioning each subjective question area and each student information area, and uploading the corrected picture and the identification result to the server;
q5, identify the content of the objective topic filling area, and review and score by manually reviewing the content of the subjective topic area.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (3)
1. An automatic scoring method and system based on feature recognition are characterized in that: the method comprises the following steps:
q1, defining the structure of the test paper, counting the information of the test paper, defining the question blocks of each area in the test paper, wherein the question blocks are divided into: a test paper title area, an examination admission card number filling area, an objective question filling area and a subjective question answering area, and points out the scores of all question blocks, a scanner is used for scanning an answer sheet gray level picture of a blank test paper, and the picture is uploaded to a server to serve as an answer sheet template for standard comparison;
q2, storing information, storing the examination numbers and the corresponding names of the students in the server;
q3, scanning and identifying the answer sheet, connecting the answer sheet after the student answers with a local client and controlling a scanner, scanning the answer sheet after the student answers to obtain an answer sheet scanning picture, correcting the front-back sequence of the answer sheet scanning picture, the paper feeding direction of the answer sheet, checking whether the information of the student examination number on the answer sheet is repeated, filled by mistake or not, and then correcting the rotation, stretching, translation, scaling and small distortion of the answer sheet scanning picture;
firstly, extracting key points of an answer sheet template, adopting ORB algorithm for scanning when scanning and correcting the answer sheet template, positioning a group of characteristic points by using FAST (FAST open search algorithm), calculating haar corner measurement of FAST characteristic points,
calculating a feature vector of the FAST feature point through formula one, namely a descriptor a, through the following formula:
and obtaining the direction of the gradient direction relative to the feature center according to a formula II, namely:
θ＝atan2(m01，m10) Formula two
Calculating to obtain a descriptor and storing the descriptor in a memory;
secondly, extracting key points of the answer sheet after answering to obtain a descriptor B;
then, matching the descriptor A with the descriptor B, wherein the descriptor A is defined as a query set, the descriptor B is defined as a training set, the two sets are matched one by one, and the selected distance formula III is as follows:
after the optimal matching is selected, a round of cross validation is carried out, so that matching errors are reduced;
and finally, determining two matched corresponding set points according to the matched descriptors. Determining transformation matrixes of the two sets through a RANSAC algorithm, and correcting student test paper to obtain a final picture;
q4, segmenting the final picture obtained in Q3 through a mask according to the position information of each question block prestored on the server, identifying, positioning each subjective question area and each student information area, and uploading the corrected picture and the identification result to the server;
q5, identify the content of the objective topic filling area, and manually review the content of the subjective topic area.
2. The automatic scoring method and system based on feature recognition as claimed in claim 1, wherein: and for the examination reference number filling area, positioning the grid points of the examination numbers filled by the students in the grid, obtaining the examination numbers of the students, and comparing the obtained examination numbers with the data in the server library in the Q2, wherein the examination numbers do not exist or need manual verification and correction repeatedly.
3. The automatic scoring method and system based on feature recognition as claimed in claim 1, wherein: and for the objective question filling area, for the unfilled options and filled options, performing two classifications, and taking the class with a large area as the option selected by the student.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910959432.8A CN110689013A (en) | 2019-10-10 | 2019-10-10 | Automatic marking method and system based on feature recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910959432.8A CN110689013A (en) | 2019-10-10 | 2019-10-10 | Automatic marking method and system based on feature recognition |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110689013A true CN110689013A (en) | 2020-01-14 |
Family
ID=69111789
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910959432.8A Pending CN110689013A (en) | 2019-10-10 | 2019-10-10 | Automatic marking method and system based on feature recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110689013A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111382721A (en) * | 2020-03-20 | 2020-07-07 | 上海乂学教育科技有限公司 | Examination room monitoring system based on artificial intelligence |
| CN112347946A (en) * | 2020-11-10 | 2021-02-09 | 成都兴唐信息技术有限公司 | Method and system for identifying multi-type answer sheet |
| CN112686143A (en) * | 2020-12-29 | 2021-04-20 | 科大讯飞股份有限公司 | Objective question filling recognition method, electronic device and storage medium |
| CN113033535A (en) * | 2021-03-11 | 2021-06-25 | 山东智多分教育科技有限公司 | Computer network marking system and marking method |
| CN113177433A (en) * | 2021-03-17 | 2021-07-27 | 北京焦耳科技有限公司 | Test paper scanning identification method, device and medium |
| CN113408521A (en) * | 2021-07-16 | 2021-09-17 | 北京南昊科技股份有限公司 | Answer picture identification method, reading and amending device and storage medium |
| CN113487701A (en) * | 2021-07-05 | 2021-10-08 | 北京鑫泰昊岳科技有限公司 | Examination input method and system |
| CN113762274A (en) * | 2021-11-10 | 2021-12-07 | 江西风向标教育科技有限公司 | Answer sheet target area detection method, system, storage medium and equipment |
| CN116798036A (en) * | 2023-06-27 | 2023-09-22 | 广州市南方人力资源评价中心有限公司 | Method and device for identifying and checking answer sheet objective question identification result |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008122788A (en) * | 2006-11-14 | 2008-05-29 | Canon Marketing Japan Inc | Image processing apparatus, image processing method, program for performing the method, and recording medium |
| CN101334814A (en) * | 2008-04-28 | 2008-12-31 | 华北电力大学 | Automatic scanning and reading system and reading method |
| CN104851094A (en) * | 2015-05-14 | 2015-08-19 | 西安电子科技大学 | Improved method of RGB-D-based SLAM algorithm |
| US20160148074A1 (en) * | 2014-11-26 | 2016-05-26 | Captricity, Inc. | Analyzing content of digital images |
| CN110110714A (en) * | 2019-04-28 | 2019-08-09 | 重庆学析优科技有限公司 | Method and system are corrected automatically on a kind of line of papery operation |
-
2019
- 2019-10-10 CN CN201910959432.8A patent/CN110689013A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008122788A (en) * | 2006-11-14 | 2008-05-29 | Canon Marketing Japan Inc | Image processing apparatus, image processing method, program for performing the method, and recording medium |
| CN101334814A (en) * | 2008-04-28 | 2008-12-31 | 华北电力大学 | Automatic scanning and reading system and reading method |
| US20160148074A1 (en) * | 2014-11-26 | 2016-05-26 | Captricity, Inc. | Analyzing content of digital images |
| CN104851094A (en) * | 2015-05-14 | 2015-08-19 | 西安电子科技大学 | Improved method of RGB-D-based SLAM algorithm |
| CN110110714A (en) * | 2019-04-28 | 2019-08-09 | 重庆学析优科技有限公司 | Method and system are corrected automatically on a kind of line of papery operation |
Non-Patent Citations (1)
| Title |
|---|
| ETHAN RUBLEE等: "《ORB: An efficient alternative to SIFT or SURF》", 《2011 INTERNATIONAL CONFERENCE ON COMPUTER VISION》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111382721A (en) * | 2020-03-20 | 2020-07-07 | 上海乂学教育科技有限公司 | Examination room monitoring system based on artificial intelligence |
| CN112347946A (en) * | 2020-11-10 | 2021-02-09 | 成都兴唐信息技术有限公司 | Method and system for identifying multi-type answer sheet |
| CN112686143A (en) * | 2020-12-29 | 2021-04-20 | 科大讯飞股份有限公司 | Objective question filling recognition method, electronic device and storage medium |
| CN112686143B (en) * | 2020-12-29 | 2023-12-01 | 科大讯飞股份有限公司 | Objective question filling identification method, electronic equipment and storage medium |
| CN113033535A (en) * | 2021-03-11 | 2021-06-25 | 山东智多分教育科技有限公司 | Computer network marking system and marking method |
| CN113177433A (en) * | 2021-03-17 | 2021-07-27 | 北京焦耳科技有限公司 | Test paper scanning identification method, device and medium |
| CN113487701A (en) * | 2021-07-05 | 2021-10-08 | 北京鑫泰昊岳科技有限公司 | Examination input method and system |
| CN113408521B (en) * | 2021-07-16 | 2023-09-05 | 北京南昊科技股份有限公司 | Answer picture identification method, reading device and storage medium |
| CN113408521A (en) * | 2021-07-16 | 2021-09-17 | 北京南昊科技股份有限公司 | Answer picture identification method, reading and amending device and storage medium |
| CN113762274A (en) * | 2021-11-10 | 2021-12-07 | 江西风向标教育科技有限公司 | Answer sheet target area detection method, system, storage medium and equipment |
| CN113762274B (en) * | 2021-11-10 | 2022-02-15 | 江西风向标教育科技有限公司 | Answer sheet target area detection method, system, storage medium and equipment |
| CN116798036A (en) * | 2023-06-27 | 2023-09-22 | 广州市南方人力资源评价中心有限公司 | Method and device for identifying and checking answer sheet objective question identification result |
| CN116798036B (en) * | 2023-06-27 | 2024-04-02 | 网才科技(广州)集团股份有限公司 | Method and device for identifying and checking answer sheet objective question identification result |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110689013A (en) | Automatic marking method and system based on feature recognition | |
| CN110008933B (en) | Universal intelligent marking system and method | |
| CN109993112B (en) | Method and device for identifying table in picture | |
| WO2020259060A1 (en) | Test paper information extraction method and system, and computer-readable storage medium | |
| US11790641B2 (en) | Answer evaluation method, answer evaluation system, electronic device, and medium | |
| WO2018006294A1 (en) | Exam paper reading system, device and method based on pattern recognition technology | |
| CN106033544B (en) | Template matching-based test paper content area extraction method | |
| CN106033535B (en) | Electronic paper marking method | |
| CN110597806A (en) | Wrong question set generation and answer statistics system and method based on reading and amending identification | |
| CN105046200B (en) | Electronic paper marking method based on straight line detection | |
| CN106846961B (en) | Electronic test paper processing method and device | |
| CN109242856A (en) | A kind of papery operation electronization reads and makes comments method and device | |
| CN110956138B (en) | Auxiliary learning method based on home education equipment and home education equipment | |
| CN107832758A (en) | The examination system of intelligence generation answering card and server terminal Automated Assessment | |
| CN108537218A (en) | A kind of identifying processing method and device of answering card | |
| CN107358148B (en) | Anti-cheating network investigation method and device based on handwriting recognition | |
| CN110659584A (en) | Intelligent trace marking system based on image recognition | |
| CN108876195A (en) | A kind of intelligentized teachers ' teaching quality evaluating system | |
| CN106778717A (en) | A kind of test and appraisal table recognition methods based on image recognition and k nearest neighbor | |
| CN112381099A (en) | Question recording system based on digital education resources | |
| CN107067399A (en) | A kind of paper image segmentation processing method | |
| CN111008594A (en) | Error correction evaluation method, related equipment and readable storage medium | |
| CN109284702B (en) | Answer sheet scoring and marking system based on image mode | |
| CN107066939A (en) | A kind of paper cutting process method of online paper-marking system | |
| CN104077562B (en) | A kind of scanning direction determination methods of test paper |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200114 |
|
| WD01 | Invention patent application deemed withdrawn after publication |