Score identification method and device based on score bar and storage medium
By identifying the position and image of the scoring bar, combining the object detection algorithm and image stitching technology, the score values in the grid where the scoring traces are located are accurately identified, solving the problem of identifying errors in the existing technology and achieving higher recognition accuracy.
Patent Information
- Application Number
- CN202411493742.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-24
- Publication Date
- 2025-03-28
AI Technical Summary
The prior art has identification errors in identifying the correction traces on the scoring bar. This is mainly due to the large differences in marking methods of different teachers, which leads to errors when calculating the midpoint coordinates of the correction traces to obtain the score value.
By obtaining the image of the test paper to be identified, identifying the position of the score bar and extracting its image, combining the object detection algorithm and image stitching technology, identifying the score traces and determining the score values in the grid where it is located, thereby accurately obtaining the score results of the test questions.
This method can more accurately identify the score values on the scoring bar, reduce identification errors, and improve the accuracy of score recognition.
Smart Images
Figure CN119723593A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent education technology, and in particular to a scoring recognition method, device and storage medium based on a scoring strip. Background Art
[0002] As education informatization continues to advance, question cards or teaching aids with scoring strips are also increasingly used by school students. Figure 1 As shown in the figure, there are generally multiple grids on the score strip, and each grid is printed with a score value. When marking, the teacher uses a red pen to draw a line on the corresponding grid to mark the score value of the corresponding test question. Conventional methods for identifying score strips are to first identify the mark traces on the image, and then calculate the position of the coordinates of the midpoints of each mark trace on the score strip. However, different teachers have different marking methods. Calculating the midpoint of the mark trace to determine its position on the score strip to obtain the corresponding score value is prone to recognition errors. Summary of the invention
[0003] In view of the deficiencies in the prior art, the present invention aims to provide a scoring recognition method, device and storage medium based on a scoring strip.
[0004] In order to achieve the above object, the present invention adopts the following technical solution:
[0005] A scoring recognition method based on a scoring bar comprises the following steps:
[0006] S1. Obtaining a test paper image to be identified, wherein the test paper image to be identified includes at least one test question with a scoring strip; the scoring strip has a plurality of grids, each of which is printed with a score value; when marking the test paper, marking the grid of the scoring strip corresponding to the score of the test question with a designated color;
[0007] S2, identifying the position of each scoring strip in the test paper image to be identified, and extracting the image of each scoring strip to form a scoring strip atlas;
[0008] S3. Identify the scoring traces in the image of the scoring bar, and further identify the score values in the grids where the scoring traces are located as the score results of the corresponding test questions, and associate and bind with the corresponding test questions.
[0009] Furthermore, in step S2, the position of the scoring bar is identified by any of the following methods:
[0010] (1) Presetting the coordinate data of each scoring strip on the test paper template, then locating the area of each scoring strip on the test paper image to be identified according to the coordinate data of each scoring strip, and extracting the image of each scoring strip;
[0011] (2) The scoring stripes in the test paper image to be identified are detected and located through the target detection algorithm, and then the images of each scoring stripe are extracted.
[0012] Furthermore, in step S2, after forming the scoring bar atlas, the scoring bar images in the scoring bar atlas are spliced, and then in step S3, all the scoring traces and the score values in the grids where the scoring traces are located are identified at one time from the spliced image, so as to obtain the score results of the test questions corresponding to each scoring bar at one time.
[0013] Furthermore, in the scoring bar atlas, the images of the scoring bars are arranged in the order of the corresponding test questions, and when the images are spliced, the images of the scoring bars in the scoring bar atlas are spliced in order along the vertical direction.
[0014] Furthermore, when images are stitched together, a preset stitching interval is added between adjacent scoring strips.
[0015] Furthermore, the specific process of step S3 is:
[0016] S3.1. Collect sample data of scoring strip images with scoring traces marked, construct a sample data set, and use the sample data set to train a target detection model;
[0017] S3.2, inputting the image of the scoring strip to be identified into the trained object detection model, and the object detection model detects the position of the scoring trace in the image of the scoring strip;
[0018] S3.2. Identify the grid where the scoring trace is located, and further identify the score value in the grid, so as to obtain the score result of the corresponding test question, and then associate and bind the identified score result with the corresponding test question.
[0019] Furthermore, in step S3, color value detection is directly performed on the image of the scoring bar to obtain the target color value, that is, the position of the scoring mark, and then the grid where the scoring mark is located is further obtained, and the score value in the grid is identified to obtain the corresponding score result of the test question, and then the identified score result and the corresponding test question are associated and bound.
[0020] Furthermore, in step S3, when more than two grids are identified to have scoring traces in a scoring strip, the values on the grids are combined from left to right as the final scoring result.
[0021] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program implements the above method when executed by a processor.
[0022] The present invention also provides a computer device, comprising a processor and a memory, wherein the memory is used to store a computer program; and when the processor is used to execute the computer program, the above method is implemented.
[0023] The beneficial effect of the present invention is that the method of the present invention can more accurately identify the score value marked on the score strip by first identifying the position of the score strip on the test paper image and extracting the image of the score strip, then identifying the correction traces on the image of the score strip, and then identifying the score value in the grid where the correction traces are located, thereby reducing the possibility of recognition errors. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] Figure 1 This is an example of an existing test paper with grading strips. DETAILED DESCRIPTION
[0025] The present invention will be further described below in conjunction with the accompanying drawings. It should be noted that this embodiment is based on the technical solution and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to this embodiment.
[0026] This embodiment provides a scoring recognition method based on a scoring bar, comprising the following steps:
[0027] S1. Obtain a test paper image to be identified, wherein the test paper image to be identified includes at least one test question with a scoring bar. Figure 1 As shown, the scoring strip has multiple grids, each of which is printed with a score value (usually 0-9, which may or may not be repeated). When marking the test paper, the teacher marks the grid of the scoring strip with a designated color (such as crossing out, ticking, etc.) according to the score of the test question.
[0028] S2. Identify the position of each scoring bar in the test paper image to be identified, and extract the image of each scoring bar to form a scoring bar atlas.
[0029] S3. Identify the scoring traces in the image of the scoring bar, and further identify the score values in the grids where the scoring traces are located as the score results of the corresponding test questions, and associate and bind with the corresponding test questions.
[0030] As one implementation method, in step S2, the coordinate data of each scoring strip on the test paper template is pre-set, and then the area of each scoring strip is located on the test paper image to be identified according to the coordinate data of each scoring strip, and the image of each scoring strip is extracted.
[0031] As another implementation, in step S2, the scoring strips in the test paper image to be identified are detected and located by a target detection algorithm, and then the images of each scoring strip are extracted. The scoring strips have a fixed format and obvious features, and the target detection algorithm can detect and locate the scoring strips with high accuracy.
[0032] As an implementation method, in step S2, after forming the scoring bar atlas, the scoring bar images in the scoring bar atlas are spliced, and then in step S3, all the scoring traces and the score values in the grids where the scoring traces are located are identified at one time from the spliced image, so as to obtain the score results of the test questions corresponding to each scoring bar at one time.
[0033] Specifically, in the scoring bar set, the images of each scoring bar are arranged in the order of the corresponding test questions. When splicing the images, the images of each scoring bar in the scoring bar set are spliced in order (either ascending or descending) along the vertical direction, that is, the scoring bar corresponding to the next test question is spliced below the scoring bar corresponding to the next test question. In this way, the score results obtained from each scoring bar can be quickly matched with each test question.
[0034] More specifically, after locating the position of the score strip in the test paper image to be identified, the score strip images corresponding to each test question are extracted one by one in order according to the layout rules of the test paper. If it is a single-page image, it is generally extracted one by one from top to bottom. If it is a multi-page image, it is generally extracted from left to right page by page, and extracted one by one from top to bottom in each page image.
[0035] More specifically, when images are stitched together, a preset stitching interval is added between adjacent scoring bars, so as to prevent the scoring bars from being too close to each other and causing recognition errors.
[0036] As another implementation, in step S2, the images of the scoring strips are not spliced, and in step S3, the scoring results of the images of the scoring strips are directly identified respectively.
[0037] In this embodiment, the specific process of step S3 is:
[0038] S3.1. Collect sample data of scoring strip images with scoring traces marked, construct a sample data set, and use the sample data set to train a target detection model (preferably a Yolo algorithm model);
[0039] S3.2, input the image of the score strip to be identified into the trained object detection model, and the object detection model detects the position of the scoring trace in the image of the score strip (such as the slash drawn by the teacher on the score value with a red pen);
[0040] S3.2, identify the grid where the scoring trace is located, and further identify the score value in the grid, so as to obtain the score result of the corresponding test question, and then associate and bind the identified score result with the corresponding test question. Specifically, the minimum border of the scoring trace is identified to determine the range of the grid where it is located.
[0041] As another implementation, in step S3, color value detection is directly performed on the image of the scoring bar (preferably opencv detects red), the target color value, i.e., the position of the scoring mark, is obtained, and then the grid where the scoring mark is located is further obtained, and the score value in the grid is identified, thereby corresponding to the score result of the test question, and then the identified score result is associated and bound with the corresponding test question.
[0042] In this embodiment, in step S3, when more than two grids are identified to have scoring traces in a scoring strip, the values on the grids are combined from left to right as the final scoring result.
[0043] For those skilled in the art, various corresponding changes and modifications can be made according to the above technical solutions and concepts, and all of these changes and modifications should be included in the protection scope of the claims of the present invention.
Claims
1. A scoring recognition method based on scoring strips, characterized in that: The steps include: S1. Obtaining a test paper image to be identified, wherein the test paper image to be identified includes at least one test question with a scoring strip; the scoring strip has a plurality of grids, each of which is printed with a score value; when marking the test paper, marking the grid of the scoring strip corresponding to the score of the test question with a designated color; S2, identifying the position of each scoring strip in the test paper image to be identified, and extracting the image of each scoring strip to form a scoring strip atlas; S3. Identify the scoring traces in the image of the scoring bar, and further identify the score values in the grids where the scoring traces are located as the score results of the corresponding test questions, and associate and bind with the corresponding test questions.
2. The method according to claim 1, characterized in that In step S2, the position of the scoring bar is identified by any of the following methods: (1) Presetting the coordinate data of each scoring strip on the test paper template, then locating the area of each scoring strip on the test paper image to be identified according to the coordinate data of each scoring strip, and extracting the image of each scoring strip; (2) The scoring stripes in the test paper image to be identified are detected and located through the target detection algorithm, and then the images of each scoring stripe are extracted.
3. The method according to claim 1, characterized in that In step S2, after forming the scoring bar atlas, the scoring bar images in the scoring bar atlas are spliced, and then in step S3, all the scoring traces and the score values in the grids where the scoring traces are located are identified at one time from the spliced image, so as to obtain the score results of the test questions corresponding to each scoring bar at one time.
4. The method according to claim 3, characterized in that In the scoring strip atlas, the images of each scoring strip are arranged in the order of the corresponding test questions. When the images are spliced, the images of each scoring strip in the scoring strip atlas are spliced in order along the vertical direction.
5. The method according to claim 3, characterized in that: When stitching images, add a preset stitching interval between adjacent scoring strips.
6. The method according to claim 1, characterized in that The specific process of step S3 is: S3.
1. Collect sample data of scoring strip images with scoring traces marked, construct a sample data set, and use the sample data set to train a target detection model; S3.2, inputting the image of the scoring strip to be identified into the trained object detection model, and the object detection model detects the position of the scoring trace in the image of the scoring strip; S3.
2. Identify the grid where the scoring trace is located, and further identify the score value in the grid, so as to obtain the score result of the corresponding test question, and then associate and bind the identified score result with the corresponding test question.
7. The method according to claim 1, characterized in that In step S3, color value detection is directly performed on the image of the scoring strip to obtain the target color value, that is, the position of the scoring mark, and then the grid where the scoring mark is located is further obtained, and the score value in the grid is identified to obtain the corresponding score result of the test question, and then the identified score result is associated and bound with the corresponding test question.
8. The method according to claim 1, characterized in that In step S3, when more than two grids are identified to have scoring traces in a scoring strip, the values on the grids are combined from left to right as the final scoring result.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.
10. A computer device, characterized in that: The method comprises a processor and a memory, wherein the memory is used to store a computer program; when the processor is used to execute the computer program, the method described in any one of claims 1 to 8 is implemented.