A subject cross scanning identification and extraction method based on a paper review system

Abstract

The application discloses a subject cross scanning recognition and extraction method based on a paper reading system. First, abnormal subject images and subject attribution are recognized through paper surface image detection. Further, different abnormal subject images are processed according to the subject attribution and the scanning mode of the subject. While accurately recognizing and extracting abnormal subject images of each subject, the method can realize the supplement of missing scanning paper surface images of each subject, thereby timely making up for operation errors in paper reading scanning, reducing the workload of manual checking, and improving the efficiency and quality of intelligent paper reading.

Description

Technical Field

[0001] This invention relates to the field of intelligent marking technology, and more specifically to a cross-disciplinary scanning, recognition, and extraction method based on a marking system. Background Technology

[0002] Currently, intelligent marking is being used more and more widely. During the intelligent marking process, due to the large number of students in schools and the tight time constraints of scanning, operational errors may occur. One common error is that exam papers from other subjects are mixed in while scanning papers from one subject, resulting in many abnormal papers in that subject, while the papers mixed in from the other subject are missed in scanning that subject. Therefore, how to effectively extract the papers from other subjects that have been mixed in, and avoid missed scanning and duplicate scanning, is a problem that urgently needs to be solved in current intelligent marking. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention aims to provide a cross-disciplinary scanning and identification extraction method based on a marking system.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A method for cross-disciplinary scanning and identification extraction based on a marking system includes the following steps:

[0006] S1. Pre-set the identification data template and scanning method for each subject. The identification data template includes the identification content of the subject and its coordinate position on the paper image and / or the standard text of the subject. The scanning method includes single-sided scanning and double-sided scanning.

[0007] S2. Scan the exam papers or answer sheets for each subject to be graded to obtain the paper images for each subject. Store the paper images for each subject in their respective scan folders. Name the paper images for each subject in the order of scanning.

[0008] S3. Perform image detection on all paper images for each subject to determine whether each paper image belongs to the current subject. If a paper image is determined not to belong to the current subject, mark it as an abnormal subject image. Specifically, determine whether each paper image belongs to the current subject by detecting whether the preset coordinate position in the paper image contains the identification content of the current subject, and / or by comparing the text similarity between the text extracted from the paper image and the standard text corresponding to the current subject to whether it exceeds a preset similarity threshold.

[0009] S4. Based on the identification data templates corresponding to other subjects, identify the subjects to which the abnormal subject images belong, and record them as abnormal subjects. Set up an independent abnormal subject image folder for each abnormal subject under the current subject, and extract the abnormal subject images belonging to the same abnormal subject from the scanning folder of the current subject and store them in the corresponding abnormal subject image folder.

[0010] S5. Based on the scanning methods of the current subject and the abnormal subject, adopt the corresponding exception handling method:

[0011] For abnormal subjects that use the same scanning method as the current subject, the abnormal subject images in the corresponding abnormal subject image folder can be directly added to the corresponding subject's scan folder as supplementary scan images for that abnormal subject.

[0012] For abnormal subjects whose scanning method differs from the current subject, if the current subject is scanned double-sided, then for abnormal subjects scanned single-sided, blank pages are first detected and removed from the abnormal subject images in the abnormal subject image folder. The abnormal subject images with blank pages removed are then added to the corresponding subject's scan folder as supplementary scan images. If the current subject is scanned single-sided, then for abnormal subjects scanned double-sided, staff are prompted to locate the corresponding exam paper or answer sheet to be graded based on the filename of the abnormal subject image, and then re-scan double-sided to obtain supplementary scan images, which are then added to the corresponding subject's scan folder.

[0013] Furthermore, the preset identification content is one or more of the following: preset target keywords, preset subject reference icons, and preset QR codes.

[0014] Furthermore, the image folders for each abnormal subject are named using the subject name.

[0015] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0016] The present invention also provides a computer device, including a processor and a memory, wherein the memory is used to store a computer program; and the processor is used to execute the computer program to implement the above-described method.

[0017] The beneficial effects of this invention are as follows: First, this invention identifies abnormal subject images and their subject affiliation through paper image detection. Then, based on the subject affiliation and the scanning method of the subject, different abnormal subject images are processed differently. While accurately identifying and extracting abnormal subject images for each subject, it can supplement the paper images that were missed in scanning for each subject. This can promptly make up for operational errors that occur during the marking and scanning process, reduce the workload of manual verification, and improve the efficiency and quality of intelligent marking. Detailed Implementation

[0018] The present invention will be further described below. It should be noted that this embodiment is based on the present technical solution and provides detailed implementation methods and specific operation processes, but the protection scope of the present invention is not limited to this embodiment.

[0019] This embodiment provides a cross-disciplinary scanning and identification extraction method based on a marking system, including the following steps:

[0020] S1. Pre-set the identification data templates and scanning methods for each subject. The identification data templates include the subject's identifier content and its coordinate position on the exam paper image, and / or the subject's standard text. The scanning methods include single-sided scanning and double-sided scanning. It should be noted that when the exam paper or answer sheet for a subject is single-sided, the scanning method is single-sided scanning; when the exam paper or answer sheet for a subject is double-sided, the scanning method is double-sided scanning. The preset identifier content can be one or more of preset target keywords, preset subject identifier icons, and preset QR codes.

[0021] S2. Scan the exam papers or answer sheets for each subject to be graded to obtain the paper images for each subject. Store the paper images for each subject in their respective scan folders. Name the paper images for each subject in the order of scanning.

[0022] S3. Perform image detection on all paper images for each subject to determine whether each paper image belongs to the current subject. If a paper image is determined not to belong to the current subject, mark it as an abnormal subject image. Specifically, determine whether each paper image belongs to the current subject by detecting whether the preset coordinate position in the paper image contains the identification content of the current subject, and / or by comparing whether the text similarity between the text extracted from the paper image and the standard text corresponding to the current subject exceeds a preset similarity threshold.

[0023] S4. Based on the identification data templates for other subjects, identify the subjects to which the abnormal subject images belong (not belonging to the current subject) belong, and record them as abnormal subjects. Create a separate abnormal subject image folder for each abnormal subject under the current subject. Extract abnormal subject images belonging to the same abnormal subject from the current subject's scan folder and store them in the corresponding abnormal subject image folder. It should be noted that since multiple subjects' exam papers or answer sheets may be mixed in during the scanning of a single subject, the number of abnormal subjects may be one or more.

[0024] S5. Based on the scanning methods of the current subject and the abnormal subject, adopt the corresponding exception handling method:

[0025] For abnormal subjects that use the same scanning method as the current subject, the abnormal subject images in their corresponding abnormal subject image folder can be directly added to the corresponding subject's scan folder as supplementary scan images. It should be noted that when an image of a test paper or answer sheet from one subject appears as an abnormal subject image in another subject, it is considered a missed image in its original subject and needs to be supplemented. Therefore, adding the supplementary scan images of all abnormal subjects under the current subject to the corresponding subject's scan folder as supplementary scan images will complete the missing scan for that subject.

[0026] For abnormal subjects whose scanning method differs from the current subject, if the current subject is scanned double-sided, then for abnormal subjects scanned single-sided, blank pages are first detected and removed from the abnormal subject images in the abnormal subject image folder. The abnormal subject images with blank pages removed are then added to the corresponding subject's scan folder as supplementary scan images. If the current subject is scanned single-sided, then for abnormal subjects scanned double-sided, staff are prompted to locate the corresponding exam paper or answer sheet to be graded based on the filename of the abnormal subject image, and then re-scan double-sided to obtain supplementary scan images, which are then added to the corresponding subject's scan folder.

[0027] In this embodiment, the abnormal subject image folders corresponding to each abnormal subject are named using the subject name.

[0028] For those skilled in the art, various corresponding changes and modifications can be made based on the above technical solutions and concepts, and all such changes and modifications should be included within the protection scope of the claims of this invention.

Claims

1. A method for cross-disciplinary scanning, recognition, and extraction based on a marking system, characterized in that, Includes the following steps: S1. Pre-set the identification data templates and scanning methods for each subject. The identification data templates include the identification content of the subject and its coordinate position on the paper image and / or the standard text of the subject. The scanning methods include single-sided scanning and double-sided scanning. S2. Scan the exam papers or answer sheets for each subject to be graded to obtain the paper images for each subject. Store the paper images for each subject in their respective scan folders. Name the paper images for each subject in the order of scanning. S3. Perform image detection on all paper images for each subject to determine whether each paper image belongs to the current subject. If a paper image is determined not to belong to the current subject, mark it as an abnormal subject image. Specifically, determine whether each paper image belongs to the current subject by detecting whether the preset coordinate position in the paper image contains the identification content of the current subject, and / or by comparing the text similarity between the text extracted from the paper image and the standard text corresponding to the current subject to whether it exceeds a preset similarity threshold. S4. Based on the identification data templates corresponding to other subjects, identify the subjects to which the abnormal subject images belong, and record them as abnormal subjects. Set up an independent abnormal subject image folder for each abnormal subject under the current subject, and extract the abnormal subject images belonging to the same abnormal subject from the scanning folder of the current subject and store them in the corresponding abnormal subject image folder. S5. Based on the scanning method of the current subject and the abnormal subject, adopt the corresponding exception handling method: For abnormal subjects that use the same scanning method as the current subject, the abnormal subject images in the corresponding abnormal subject image folder can be directly added to the corresponding subject's scan folder as supplementary scan images for that abnormal subject. For abnormal subjects whose scanning method is different from that of the current subject, if the current subject is scanned in double-sided scanning, then for abnormal subjects whose scanning method is scanned in single-sided scanning, first perform blank page detection on the abnormal subject images in the abnormal subject image folder and remove the blank pages. The abnormal subject images with blank pages removed are added to the corresponding subject's scan folder as supplementary scan images for the corresponding abnormal subject. If the current subject is scanned using a single-sided scan method, then for abnormal subjects that are scanned using a double-sided scan method, staff will be prompted to find the corresponding exam paper or answer sheet to be graded based on the file name of the image of the abnormal subject, and then re-scan the double-sided image to obtain a supplementary image to be added to the corresponding subject's scan folder.

2. The method according to claim 1, characterized in that, The preset identification content is one or more of the following: preset target keywords, preset subject reference icons, and preset QR codes.

3. The method according to claim 1, characterized in that, The image folders for each abnormal subject are named after the subject.

4. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-3.

5. A computer device, characterized in that, It includes a processor and a memory, the memory being used to store a computer program; the processor being used to execute the computer program to implement the method according to any one of claims 1-3.