CN113283431B - Answer sheet option area identification method and system - Google Patents
Answer sheet option area identification method and system Download PDFInfo
- Publication number
- CN113283431B CN113283431B CN202110841398.1A CN202110841398A CN113283431B CN 113283431 B CN113283431 B CN 113283431B CN 202110841398 A CN202110841398 A CN 202110841398A CN 113283431 B CN113283431 B CN 113283431B
- Authority
- CN
- China
- Prior art keywords
- option
- area
- option area
- missing
- answer sheet
- 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
Links
Images
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- 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
- G06V10/30—Noise filtering
Abstract
The invention provides an answer sheet option area identification method and system, wherein the method comprises the following steps: acquiring an image of the answer sheet; performing image segmentation on the answer sheet image by using a pre-trained image segmentation model to identify each option area and parameter information thereof on the answer sheet image, wherein the parameter information comprises coordinate information and size information; judging whether a missing option area exists according to the coordinate information of the option area; if so, supplementing the missing option area according to the coordinate information of the adjacent option area of the missing option area and the size information of the option area; and outputting each option area of the identified answer sheet image. The invention successfully solves the problems of less characteristic points of the fuzzy image and higher image noise ratio which are difficult to process by combining the deep learning neural network model and the logic judgment, improves the identification precision of the option area and can well solve the problem of missing identification.
Description
Technical Field
The invention relates to the technical field of image information identification, in particular to a method and a system for identifying option areas of an answer sheet.
Background
In order to accelerate the efficiency of the examination paper marking and reading, answer cards are generally adopted for answer filling at present, and because objective questions such as selected questions, admission card numbers and the like on the answer cards can be matched with images of correct answer card modules to automatically identify the errors, the marking and reading work of the objective questions can be rapidly completed. The precondition of quick reading of objective questions of the answer sheet is that each option area (such as [ A ]) in the image of the answer sheet can be automatically identified.
In the prior art, the current method for identifying the option area is as follows: firstly, acquiring a template answer sheet, and respectively framing an objective question area, a subjective question area and an examination admission card number area on the template answer sheet by a user; and then, carrying out pixel-level image preprocessing on the three regions to obtain a binary image, searching a filling option region according to the contour characteristics and the arrangement characteristics of the filling option, screening according to the perimeter, the area and the like when identifying the option, removing irrelevant contours, synthesizing all shapes into a polygon in a polygon fitting mode, judging the number of edges, removing most numbers and letters, clustering, and finally judging the most frequently appearing contours as the option region.
The above-mentioned method for identifying the option area has the following disadvantages: under the conditions that fuzzy image feature points are few, image noise is large, outline information is difficult to obtain, a large number of isolated objects are often in disagreement with isolated points of the objects after image processing, so that the identification accuracy of an option area is low, identification is easy to miss, the option area is lost, the number of clusters is determined during clustering, and when the position and the characteristic of a cluster center are not clear in advance, an initial value is difficult to set.
Disclosure of Invention
Based on this, the invention aims to provide an answer sheet option area identification method and system, so as to solve the technical problems that the existing option area identification method is low in identification precision and easy to miss identification.
The method for identifying the option area of the answer sheet according to the embodiment of the invention comprises the following steps:
acquiring an image of the answer sheet;
carrying out image segmentation on the answer sheet image by using a pre-trained image segmentation model so as to identify each option area and parameter information thereof in each objective question area on the answer sheet image, wherein the parameter information comprises coordinate information and size information, and the image segmentation model is obtained based on deep learning neural network training;
judging whether a missing option area exists according to the coordinate information of the option area;
if so, supplementing the missing option area in a corresponding position according to the coordinate information of the adjacent option area of the missing option area and the size information of the option area so as to identify and output each option area of the answer sheet image;
and if not, outputting each option area of the answer sheet image obtained by identification.
In addition, the answer sheet option area identification method according to the above embodiment of the present invention may further have the following additional technical features:
further, the step of determining whether there is a missing option area according to the coordinate information of the option area includes:
determining rank information of each option area in each objective question area according to the coordinate information of the option areas;
counting the number of the option areas of each row or each column, and when the counted number of the option areas of any current row or current column is lower than a preset number, determining that the missing option areas exist in the current row or current column;
and determining the position of the missing option area in the current row or the current column according to the coordinate information of the option area in the current row or the current column.
Further, the step of determining rank information of each option area in each objective question area according to the coordinate information of the option area comprises:
according to the coordinate information of the option areas, the option areas with the abscissa variation value smaller than a first threshold value are listed as the same row, and the option areas with the ordinate variation value smaller than a second threshold value are listed as the same column;
and carrying out row sorting on the option areas of each row according to a rule that the abscissa increases from left to right, and carrying out column sorting on the option areas of each column according to a rule that the ordinate increases from top to bottom, so as to determine the row and column information of each option area.
Further, the step of supplementing the missing option area at a corresponding position according to the coordinate information of the adjacent option area of the missing option area and the size information of the option area includes:
determining target coordinate information of the missing option area according to the coordinate information of the adjacent option area of the missing option area;
calculating the size average value of the option area according to the size information of the option area;
and displaying a preset option area frame according to the size average value, and inserting the preset option area frame into a position corresponding to the target coordinate information.
Further, before the step of outputting each option area of the answer sheet image, the method further comprises:
calculating the black pixel proportion of each option area;
and determining a filled option area according to the black pixel proportion of the option area.
Further, after the step of outputting each option area of the answer sheet image, the method further includes:
and storing each option area of the answer sheet image and the answer sheet image in a correlation manner.
Further, before the step of performing image segmentation on the answer sheet image by using the pre-trained image segmentation model, the method further includes:
acquiring a frame selection area in the answer sheet image to determine the objective question area;
preprocessing the objective question area;
wherein, the preprocessing is one or more of denoising processing, data enhancement processing and resolution adjustment processing.
According to the embodiment of the invention, the system for identifying the option area of the answer sheet comprises the following components:
the image acquisition module is used for acquiring an image of the answer sheet;
the image segmentation module is used for carrying out image segmentation on the answer sheet image by utilizing a pre-trained image segmentation model so as to identify each option area and parameter information thereof in each objective question area on the answer sheet image, wherein the parameter information comprises coordinate information and size information;
the missing judgment module is used for judging whether the missing option area exists according to the coordinate information of the option area;
the area identification module is used for supplementing the missing option area in the corresponding position according to the coordinate information of the adjacent option area of the missing option area and the size information of the option area when the missing option area is judged to exist, so as to identify and output each option area of the answer sheet image; and outputting each option area of the answer sheet image obtained by identification when judging that no missing option area exists.
The present invention further provides a computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, implements the above-mentioned answer sheet option area identification method.
The invention also provides an answer sheet option area identification device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the answer sheet option area identification method.
Compared with the prior art: the method comprises the steps of segmenting an option area of an image by using a deep learning model, enabling the obtained option area and parameter information thereof to be more accurate, solving the problem that a large number of isolated targets are not matched with isolated points of the targets after image processing, judging and analyzing the position of each option area identified by the model through logic, and completing the part of model missing identification according to coordinate information of adjacent option areas and size information of the option areas, thereby obtaining a complete option area and ensuring that the option areas are not missed for identification.
Drawings
Fig. 1 is a flowchart of an answer sheet option area identification method according to a first embodiment of the invention;
fig. 2 is an image of an answer sheet provided in an embodiment of the present invention;
fig. 3 is a flowchart illustrating an answer sheet option area identification method according to a second embodiment of the present invention;
fig. 4 is a schematic structural diagram of an answer sheet option area identification system in a third embodiment of the invention;
fig. 5 is a schematic structural diagram of an answer sheet option area identification device in a fourth embodiment of the present invention.
The following detailed description will further illustrate the invention in conjunction with the above-described figures.
Detailed Description
To facilitate an understanding of the invention, the invention will now be described more fully with reference to the accompanying drawings. Several embodiments of the invention are presented in the drawings. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
It will be understood that when an element is referred to as being "secured to" another element, it can be directly on the other element or intervening elements may also be present. When an element is referred to as being "connected" to another element, it can be directly connected to the other element or intervening elements may also be present. The terms "vertical," "horizontal," "left," "right," and the like as used herein are for illustrative purposes only.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
Example one
Referring to fig. 1, a method for identifying option areas of an answer sheet according to a first embodiment of the present invention is shown, and the method specifically includes steps S01-S05.
Step S01, an image of the answer sheet is acquired.
The answer sheet image can be captured by a camera or a scanner, and when the answer sheet image is captured, the answer sheet is ensured to be placed flatly and have no obvious substances on the surface to reduce the noise interference of subsequent images as much as possible.
Step S02, performing image segmentation on the answer sheet image by using a pre-trained image segmentation model, so as to identify each option area and parameter information thereof in each objective question area on the answer sheet image, where the parameter information includes coordinate information and size information.
In some cases of the present embodiment, the image segmentation model is trained based on a deep learning neural network. During specific training, batch objective question area images can be collected in advance, the objective question area images can be obtained by intercepting answer sheet images, each option area in each objective question area image is marked manually (selected in a box), so that batch training samples are prepared, the training samples are the manually marked objective question area images, and then the training samples are used as input of a deep learning neural network to train a model, so that an image segmentation model is obtained through training.
Alternatively, as an alternative implementation manner, a batch of answer sheet images may be collected in advance, and then an objective question area (as shown in fig. 2) is manually selected from a middle frame of the answer sheet images, and then each option area in the objective question area images is manually selected from a middle frame of the objective question area images, so as to prepare a batch of training samples, where the training samples are the manually labeled answer sheet images.
In addition, when the deep learning neural network model is trained, a unified two-dimensional coordinate system construction rule can be preset, so that the model can construct a unified two-dimensional coordinate system in an input image based on the construction rule, and thus the image segmentation model obtained by training can accurately find the coordinate information of each option area based on a standard two-dimensional coordinate system, for example, the coordinate of an option area frame (generally [ or () ]) is determined, and the size information of each option area is further determined.
And step S03, judging whether the missing option area exists according to the coordinate information of the option area.
It should be noted that, because the option areas of the same objective question area are arranged in a row-column regular manner (as shown in fig. 2), for this reason, according to the coordinate information of each option area, the position of the missing option area can be determined, for example, the horizontal coordinate difference value between the 2 nd and 3 rd option areas in the first row is significantly larger than the horizontal distance between the other two adjacent option areas, the missing option area between the 2 nd and 3 rd option areas in the first row is determined, and the number of the missing option areas between the two option areas can be determined according to the horizontal coordinate difference value between the two option areas.
When it is determined that there is a missing option area, step S04 is executed, and when it is determined that there is no missing option area, the identification representing all the option areas is successful, step S05 is executed.
Step S04, according to the coordinate information of the adjacent option area of the missing option area and the size information of the option area, the missing option area is supplemented in the corresponding position to identify and output each option area of the answer sheet image.
In some optional embodiments of this embodiment, according to the coordinate information of the adjacent option area of the missing option area and the size information of the option area, the step of padding the missing option area in the corresponding position may specifically include:
step S041, determining target coordinate information of the missing option area according to the coordinate information of the adjacent option area of the missing option area;
step S042, calculating the size average value of the option area by the size information of the option area, namely calculating the frame size average value of the option area;
and S043, displaying a preset option area (such as [ C ] or ()) frame according to the size average value, and inserting the preset option area frame into a position corresponding to the target coordinate information. Because the subsequent matching of the answer sheet images is only the matching between the option areas, the option areas only need to be marked, and the corresponding option values in the option areas are not needed, and if the options [ B ], the corresponding option areas are only marked as [ B ], so that the data processing amount is greatly reduced, and the identification efficiency is improved. The option area mark is to extract a border outline of the option area so as to identify the corresponding option area by the border outline.
In step S041, if the missing option region is located in the middle of the objective subject region, the missing option region is inevitably located between two adjacent option regions in the same row or in the same column, for example, if the options in each row are [ a ], [ B ], [ C ], [ D ], if the option [ B ] is missing, it may be determined that an option region is missing between the option [ a ] and the option [ C ], the median of the abscissa of the option [ a ] and the median of the abscissa of the option [ C ] may be determined, and the ordinate of the option [ B ] may be the ordinate of any one of the option [ a ] and the option [ C ] or the mean of the two;
if the missing option area is located at the corner position of the objective topic area, the missing option area is necessarily adjacent to two different rows of option areas, for example, when the first row option [ a ] is missing, the missing option area is adjacent to the first row option [ B ] and the second row option [ a ], and the abscissa of the second row option [ a ] is the abscissa of the first row option [ a ], and the ordinate of the first row option [ B ] is the ordinate of the first row option [ a ].
And step S05, outputting each option area of the answer sheet image obtained by identification.
In summary, the answer sheet option area identification method in the above embodiments of the invention divides the option area of the image by using the deep learning model, the obtained option areas and the parameter information thereof are more accurate, the problem that a large number of isolated targets and isolated points with inconsistent targets exist after the image processing is solved, the position of each option area identified by the analysis model is judged through logic, completing the part of the model which is not identified according to the coordinate information of the adjacent option area and the size information of the option area, therefore, a complete option area is obtained, and the option area is ensured not to be identified in a missing way, so the invention successfully solves the problems of less characteristic points of the fuzzy image and larger image noise ratio and difficult processing by combining the deep learning neural network model and the logic judgment, the identification precision of the option area is improved, and the problem of missed identification can be well solved.
Example two
Referring to fig. 3, a method for identifying option areas of an answer sheet according to a second embodiment of the present invention is shown, and the method specifically includes steps S11 to S18.
Step S11, obtaining an answer sheet image, obtaining a frame selection area in the answer sheet image to determine the objective question area, and then preprocessing the objective question area.
Wherein, the preprocessing is one or more of denoising processing, data enhancement processing and resolution adjustment processing.
In the specific implementation, the frame selection area in the answer sheet image can be manually selected by the operator, and can also be automatically selected based on the reference point given by the operator.
Step S12, performing image segmentation on the answer sheet image by using a pre-trained image segmentation model to identify each option area and parameter information thereof in each objective question area on the answer sheet image, where the parameter information includes coordinate information and size information.
The image segmentation model is obtained based on deep learning neural network training.
And step S13, determining the rank information of each option area in each objective question area according to the coordinate information of the option areas.
In specific implementation, the step S13 can be specifically implemented by the following detailed steps:
step S131, according to the coordinate information of the option areas, listing the option areas with the abscissa variation value smaller than a first threshold value as a same row, and listing the option areas with the ordinate variation value smaller than a second threshold value as a same column;
and S132, carrying out row sorting on the option areas in each row according to a rule that the horizontal coordinate is sequentially increased from left to right, and carrying out column sorting on the option areas in each column according to a rule that the vertical coordinate is sequentially increased from top to bottom so as to determine the row and column information of each option area.
The first threshold and the second threshold can be preset for answer sheets of different versions, and the answer sheets of different versions can be identified based on answer sheet identifiers (such as codes) in the answer sheet images, so that the first threshold and the second threshold corresponding to the current version are automatically selected; or the first threshold value and the second threshold value can be manually input by an operator; or the first threshold value and the second threshold value can be dynamically determined by calculating the abscissa mean value and the ordinate mean value of each option area.
Step S14, performing quantity statistics on the option areas in each row or each column, and determining that there is a missing option area in any current row or current column when the statistical quantity of the option areas in any current row or current column is lower than a preset quantity.
The preset number may be preset, or may be manually input by an operator, or a value with the largest statistical number may be used as the preset number, for example, after counting that the number of 7 rows of option areas in 8 rows is 4, and the number of only one row of option areas is 3, the preset number is 4 by default.
In addition, it should be noted that, since the option areas of the same objective problem area are arranged in a row-column regular manner (as shown in fig. 2), for this reason, under the normal recognition condition, the number of option areas in each row or column is equal, whereas as long as the number of option areas in one row or column is less than that in other rows or columns, the option area in the row or column is necessarily missing.
Step S15, determining the position of the missing option area in the current row or the current column according to the coordinate information of the option area in the current row or the current column.
Specifically, the position of the missing option area in the current row or the current column may be determined according to the abscissa change of the option area in the current row or the ordinate change of the current column, and after the position of the missing option area is determined, the option area adjacent to the missing option area may be determined.
Step S16, according to the coordinate information of the adjacent option area of the missing option area and the size information of the option area, the missing option area is added in the corresponding position.
And step S17, calculating the black pixel proportion of each option area, and determining the filled option area according to the black pixel proportion of the option area.
It should be noted that, because the examinee usually applies the 2B pencil to perform full-blacking or partial-blacking in the corresponding option area of the answer sheet, the black pixel ratio of the option area to be filled is greater than the threshold, so that the option area to be filled by the examinee can be determined based on the black pixel ratio, and then the answering condition is identified, which is convenient for the examinee to review by matching the filled option area with the image of the answer sheet module recording the correct answer.
Therefore, the embodiment determines the filled option area by adopting a black pixel proportion mode, and compared with the traditional clustering mode, a series of problems of clustering can be solved, and more accurate identification information can be obtained.
And step S18, outputting each option area of the answer sheet image obtained by identification, and storing each option area of the answer sheet image and the answer sheet image in a correlation manner.
Outputting each option area of the identified answer sheet image actually means outputting all the answer sheet images marked by the option areas, and performing associated storage with the original answer sheet image, so as to perform subsequent tracing and performing spot check on identification accuracy by an operator.
EXAMPLE III
Another aspect of the present invention further provides an answer sheet option area identification system, referring to fig. 4, which is an answer sheet option area identification system according to a third embodiment of the present invention, and the answer sheet option area identification system includes:
the image acquisition module 11 is used for acquiring an image of the answer sheet;
an image segmentation module 12, configured to perform image segmentation on the answer sheet image by using a pre-trained image segmentation model, so as to identify each option area and parameter information thereof in each objective question area on the answer sheet image, where the parameter information includes coordinate information and size information;
a missing judgment module 13, configured to judge whether a missing option area exists according to the coordinate information of the option area;
the area identification module 14 is configured to, when it is determined that a missing option area exists, supplement the missing option area at a corresponding position according to the coordinate information of an adjacent option area of the missing option area and the size information of the option area, so as to identify and output each option area of the answer sheet image; and outputting each option area of the answer sheet image obtained by identification when judging that no missing option area exists.
Further, in some optional embodiments of the present invention, the missing determining module 13 includes:
the rank information determining unit is used for determining rank information of each option area in each objective question area according to the coordinate information of the option areas;
the quantity counting unit is used for counting the quantity of the option areas of each row or each column, and when the counting quantity of the option areas of any current row or current column is lower than the preset quantity, the missing option areas in the current row or current column are determined;
a position determining unit, configured to determine a position of the missing option area in the current row or the current column according to coordinate information of the option area in the current row or the current column.
Further, in some optional embodiments of the present invention, the row-column information determining unit is further configured to list, according to the coordinate information of the option area, the option area with the abscissa variation value smaller than the first threshold as a same row, and list, with the ordinate variation value smaller than the second threshold, the option area as a same column; and carrying out row sorting on the option areas of each row according to a rule that the abscissa increases from left to right, and carrying out column sorting on the option areas of each column according to a rule that the ordinate increases from top to bottom, so as to determine the row and column information of each option area.
Further, in some alternative embodiments of the present invention, the area identification module 14 includes:
a coordinate determination unit, configured to determine target coordinate information of the missing option area according to coordinate information of an option area adjacent to the missing option area;
the mean value calculating unit is used for calculating the mean value of the size of the option area according to the size information of the option area;
and the frame inserting unit is used for displaying a preset option area frame according to the size average value and inserting the preset option area frame into a position corresponding to the target coordinate information.
Further, in some optional embodiments of the present invention, the answer sheet option area identification system further includes:
the filling option area determining module is used for calculating the black pixel proportion of each option area; and determining a filled option area according to the black pixel proportion of the option area.
Further, in some optional embodiments of the present invention, the answer sheet option area identification system further includes:
the image processing module is used for acquiring a frame selection area in the answer sheet image so as to determine the objective question area; preprocessing the objective question area; wherein, the preprocessing is one or more of denoising processing, data enhancement processing and resolution adjustment processing.
The functions or operation steps of the modules and units when executed are substantially the same as those of the method embodiments, and are not described herein again.
In summary, the answer sheet option area recognition system in the above embodiments of the invention divides the option area of the image by using the deep learning model, the obtained option areas and the parameter information thereof are more accurate, the problem that a large number of isolated targets and isolated points with inconsistent targets exist after the image processing is solved, the position of each option area identified by the analysis model is judged through logic, completing the part of the model which is not identified according to the coordinate information of the adjacent option area and the size information of the option area, therefore, a complete option area is obtained, and the option area is ensured not to be identified in a missing way, so the invention successfully solves the problems of less characteristic points of the fuzzy image and larger image noise ratio and difficult processing by combining the deep learning neural network model and the logic judgment, the identification precision of the option area is improved, and the problem of missed identification can be well solved.
Example four
Referring to fig. 5, the answer sheet option area identification device according to a fourth embodiment of the present invention includes a memory 20, a processor 10, and a computer program 30 stored in the memory and executable on the processor, where the processor 10 executes the computer program 30 to implement the answer sheet option area identification method described above.
The answer sheet option area identification device may specifically be a computer, an electronic paper marking device, and the like, and the processor 10 may be a Central Processing Unit (CPU), a controller, a microcontroller, a microprocessor, or another data Processing chip in some embodiments, and is configured to run program codes or process data stored in the memory 20, for example, execute an access restriction program.
The memory 20 includes at least one type of readable storage medium, which includes a flash memory, a hard disk, a multimedia card, a card type memory (e.g., SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, and the like. Memory 20 may in some embodiments be an internal storage unit of the answer sheet option area identification device, for example a hard disk of the answer sheet option area identification device. The memory 20 may also be an external storage device of the answering Card option area identification device in other embodiments, such as a plug-in hard disk provided on the answering Card option area identification device, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and so on. Further, the memory 20 may also include both an internal storage unit of the answer sheet option area identification device and an external storage device. The memory 20 may be used not only to store application software installed in the answer sheet option area identification device and various types of data, but also to temporarily store data that has been output or is to be output.
It is noted that the configuration shown in fig. 5 does not constitute a limitation of the answer sheet option area identification device, and in other embodiments, the answer sheet option area identification device may include fewer or more components than those shown, or some components may be combined, or a different arrangement of components.
In summary, the answer sheet option area identification device in the above embodiments of the present invention, by using the deep learning model to segment the option area of the image, the obtained option areas and the parameter information thereof are more accurate, the problem that a large number of isolated targets and isolated points with inconsistent targets exist after the image processing is solved, the position of each option area identified by the analysis model is judged through logic, completing the part of the model which is not identified according to the coordinate information of the adjacent option area and the size information of the option area, therefore, a complete option area is obtained, and the option area is ensured not to be identified in a missing way, so the invention successfully solves the problems of less characteristic points of the fuzzy image and larger image noise ratio and difficult processing by combining the deep learning neural network model and the logic judgment, the identification precision of the option area is improved, and the problem of missed identification can be well solved.
An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the method for identifying option areas of an answer sheet as described above.
Those of skill in the art will understand that the logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be viewed as implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (7)
1. An answer sheet option area identification method is characterized by comprising the following steps:
acquiring an image of the answer sheet;
carrying out image segmentation on the answer sheet image by using a pre-trained image segmentation model so as to identify each option area and parameter information thereof in each objective question area on the answer sheet image, wherein the parameter information comprises coordinate information and size information, and the image segmentation model is obtained based on deep learning neural network training;
judging whether a missing option area exists according to the coordinate information of the option area;
if so, supplementing the missing option area in a corresponding position according to the coordinate information of the adjacent option area of the missing option area and the size information of the option area so as to identify and output each option area of the answer sheet image;
if not, outputting each option area of the answer sheet image obtained by identification;
the step of judging whether the missing option area exists according to the coordinate information of the option area comprises the following steps:
determining rank information of each option area in each objective question area according to the coordinate information of the option areas;
counting the number of the option areas of each row or each column, and when the counted number of the option areas of any current row or current column is lower than a preset number, determining that the missing option areas exist in the current row or current column;
determining the number of the missing option areas and the positions of the missing option areas in the current row or the current column according to the abscissa change of the option areas of the current row or the ordinate change of the current column;
the step of supplementing the missing option area in the corresponding position according to the coordinate information of the adjacent option area of the missing option area and the size information of the option area comprises the following steps:
determining target coordinate information of the missing option area according to the coordinate information of the adjacent option area of the missing option area;
calculating the size average value of the option area according to the size information of the option area;
displaying a preset option area frame according to the size average value, and inserting the preset option area frame into a position corresponding to the target coordinate information;
when the missing option area is in the first row or the last row of the objective problem area, taking the ordinate of any one of two adjacent option areas in the same row as the ordinate of the missing option area, and taking the median of the abscissas of the two adjacent option areas in the same row as the abscissa of the missing option area;
when the missing option area is in the first column or the last column of the objective problem area, taking the abscissa of any one of two adjacent option areas in the same column as the abscissa of the missing option area, and taking the median of the ordinates of the two adjacent option areas in the same column as the ordinate of the missing option area;
when the missing option area is in the middle of the objective question area, taking the ordinate of any one of two adjacent option areas in the same row as the ordinate of the missing option area, and taking the abscissa median of the two adjacent option areas in the same row as the abscissa of the missing option area;
if the missing option area is located at the corner position of the objective question area, taking the ordinate of the option area in the same row as the missing option area in two adjacent option areas as the ordinate of the missing option area, and taking the abscissa of the option area in the same column as the abscissa of the missing option area.
2. The method for identifying option areas of an answer sheet according to claim 1, wherein the step of determining the rank information of each option area in each objective question area according to the coordinate information of the option area comprises:
according to the coordinate information of the option areas, the option areas with the abscissa variation value smaller than a first threshold value are listed as the same row, and the option areas with the ordinate variation value smaller than a second threshold value are listed as the same column;
and carrying out row sorting on the option areas of each row according to a rule that the abscissa increases from left to right, and carrying out column sorting on the option areas of each column according to a rule that the ordinate increases from top to bottom, so as to determine the row and column information of each option area.
3. The answer sheet option area identification method according to claim 1, further comprising, before the step of outputting the option areas of the answer sheet image:
calculating the black pixel proportion of each option area;
and determining a filled option area according to the black pixel proportion of the option area.
4. The answer sheet option area identification method according to claim 1 or 3, wherein after the step of outputting each option area of the answer sheet image, further comprising:
and storing each option area of the answer sheet image and the answer sheet image in a correlation manner.
5. The answer sheet option area recognition method of claim 1, wherein before the step of image-segmenting the answer sheet image using a pre-trained image segmentation model, further comprising:
acquiring a frame selection area in the answer sheet image to determine the objective question area;
preprocessing the objective question area;
wherein, the preprocessing is one or more of denoising processing, data enhancement processing and resolution adjustment processing.
6. An answer sheet option area identification system, the system comprising:
the image acquisition module is used for acquiring an image of the answer sheet;
the image segmentation module is used for carrying out image segmentation on the answer sheet image by utilizing a pre-trained image segmentation model so as to identify each option area and parameter information thereof in each objective question area on the answer sheet image, wherein the parameter information comprises coordinate information and size information;
the missing judgment module is used for judging whether the missing option area exists according to the coordinate information of the option area;
the area identification module is used for supplementing the missing option area in the corresponding position according to the coordinate information of the adjacent option area of the missing option area and the size information of the option area when the missing option area is judged to exist, so as to identify and output each option area of the answer sheet image; and the answer sheet is used for outputting each option area of the answer sheet image obtained by identification when judging that no missing option area exists;
the deletion judging module comprises:
the rank information determining unit is used for determining rank information of each option area in each objective question area according to the coordinate information of the option areas;
the quantity counting unit is used for counting the quantity of the option areas of each row or each column, and when the counting quantity of the option areas of any current row or current column is lower than the preset quantity, the missing option areas in the current row or current column are determined;
a position determining unit, configured to determine a position of the missing option area in the current row or the current column according to coordinate information of the option area in the current row or the current column;
wherein the region identification module comprises:
a coordinate determination unit, configured to determine target coordinate information of the missing option area according to coordinate information of an option area adjacent to the missing option area;
the mean value calculating unit is used for calculating the mean value of the size of the option area according to the size information of the option area;
the frame inserting unit is used for displaying a preset option area frame according to the size average value and inserting the preset option area frame into a position corresponding to the target coordinate information;
when the missing option area is in the first row or the last row of the objective problem area, taking the ordinate of any one of two adjacent option areas in the same row as the ordinate of the missing option area, and taking the median of the abscissas of the two adjacent option areas in the same row as the abscissa of the missing option area;
when the missing option area is in the first column or the last column of the objective problem area, taking the abscissa of any one of two adjacent option areas in the same column as the abscissa of the missing option area, and taking the median of the ordinates of the two adjacent option areas in the same column as the ordinate of the missing option area;
when the missing option area is in the middle of the objective question area, taking the ordinate of any one of two adjacent option areas in the same row as the ordinate of the missing option area, and taking the abscissa median of the two adjacent option areas in the same row as the abscissa of the missing option area;
if the missing option area is located at the corner position of the objective question area, taking the ordinate of the option area in the same row as the missing option area in two adjacent option areas as the ordinate of the missing option area, and taking the abscissa of the option area in the same column as the abscissa of the missing option area.
7. The answer sheet option area identification system of claim 6, further comprising:
the filling option area determining module is used for calculating the black pixel proportion of each option area; and determining a filled option area according to the black pixel proportion of the option area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110841398.1A CN113283431B (en) | 2021-07-26 | 2021-07-26 | Answer sheet option area identification method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110841398.1A CN113283431B (en) | 2021-07-26 | 2021-07-26 | Answer sheet option area identification method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113283431A CN113283431A (en) | 2021-08-20 |
| CN113283431B true CN113283431B (en) | 2021-11-26 |
Family
ID=77287252
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110841398.1A Active CN113283431B (en) | 2021-07-26 | 2021-07-26 | Answer sheet option area identification method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113283431B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113469147B (en) * | 2021-09-02 | 2021-12-17 | 北京世纪好未来教育科技有限公司 | Answer sheet identification method and device, electronic equipment and storage medium |
| 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 |
| CN115292542B (en) * | 2022-10-09 | 2023-01-20 | 江西风向标智能科技有限公司 | Test question searching method, system, storage medium and equipment based on test question matching chart |
| 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 |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104866847A (en) * | 2015-05-20 | 2015-08-26 | 李宇 | Rapid and robust answering card identifying method |
| CN105989347A (en) * | 2015-02-28 | 2016-10-05 | 科大讯飞股份有限公司 | Intelligent marking method and system of objective questions |
| CN106778750A (en) * | 2017-01-17 | 2017-05-31 | 深圳市海云天科技股份有限公司 | A kind of answering card identification point accurate positioning method and system |
| CN108171297A (en) * | 2018-01-24 | 2018-06-15 | 谢德刚 | A kind of answer card identification method and device |
| CN109165642A (en) * | 2018-06-13 | 2019-01-08 | 泉州市雄蓝教育科技有限公司 | A kind of enhanced answering card intelligent identification Method based on deep learning |
| 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 |
| CN112733644A (en) * | 2020-12-30 | 2021-04-30 | 广东德诚大数据科技有限公司 | Filling point identification system and method based on scanning identification |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2016224585A (en) * | 2015-05-28 | 2016-12-28 | 株式会社Ｊｖｃケンウッド | Road surface sign detection device, fault element detection device, lane detection device, lane detection method, and program |
| CN108932508B (en) * | 2018-08-13 | 2022-03-18 | 杭州大拿科技股份有限公司 | Method and system for intelligently identifying and correcting subjects |
| CN110647879B (en) * | 2019-08-06 | 2023-04-07 | 安徽学云教育科技有限公司 | Answer sheet objective question identification method |
| CN111626280B (en) * | 2020-04-13 | 2021-09-07 | 北京邮电大学 | Method and device for identifying answer sheet without positioning point |
-
2021
- 2021-07-26 CN CN202110841398.1A patent/CN113283431B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105989347A (en) * | 2015-02-28 | 2016-10-05 | 科大讯飞股份有限公司 | Intelligent marking method and system of objective questions |
| CN104866847A (en) * | 2015-05-20 | 2015-08-26 | 李宇 | Rapid and robust answering card identifying method |
| CN106778750A (en) * | 2017-01-17 | 2017-05-31 | 深圳市海云天科技股份有限公司 | A kind of answering card identification point accurate positioning method and system |
| CN108171297A (en) * | 2018-01-24 | 2018-06-15 | 谢德刚 | A kind of answer card identification method and device |
| CN109165642A (en) * | 2018-06-13 | 2019-01-08 | 泉州市雄蓝教育科技有限公司 | A kind of enhanced answering card intelligent identification Method based on deep learning |
| 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 |
| CN112733644A (en) * | 2020-12-30 | 2021-04-30 | 广东德诚大数据科技有限公司 | Filling point identification system and method based on scanning identification |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113283431A (en) | 2021-08-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113283431B (en) | Answer sheet option area identification method and system | |
| CN108171297B (en) | Answer sheet identification method | |
| CN110060237B (en) | Fault detection method, device, equipment and system | |
| CN106033535B (en) | Electronic paper marking method | |
| CN101122953A (en) | Picture words segmentation method | |
| US20150278626A1 (en) | Character recognition device and character segmentation method | |
| CN110598566A (en) | Image processing method, device, terminal and computer readable storage medium | |
| CN104298947A (en) | Method and device for accurately positioning two-dimensional bar code | |
| CN115620075B (en) | Method, system and equipment for generating data set for leukocyte classification model | |
| CN111915635A (en) | Test question analysis information generation method and system supporting self-examination paper marking | |
| CN113688838B (en) | Red handwriting extraction method and system, readable storage medium and computer equipment | |
| CN113762274B (en) | Answer sheet target area detection method, system, storage medium and equipment | |
| CN113283439B (en) | Intelligent counting method, device and system based on image recognition | |
| CN109635796B (en) | Questionnaire recognition method, device and equipment | |
| CN114821554A (en) | Image recognition method, electronic device, and storage medium | |
| CN111507344A (en) | Method and device for recognizing characters from image | |
| JP2017521011A (en) | Symbol optical detection method | |
| CN115359302A (en) | Coin identification method, system and storage medium | |
| CN112861843A (en) | Method and device for analyzing selection frame based on feature image recognition | |
| CN112861836B (en) | Text image processing method, text and card image quality evaluation method and device | |
| CN113392819A (en) | Batch academic image automatic segmentation and labeling device and method | |
| CN117409428B (en) | Test paper information processing method, system, computer equipment and storage medium | |
| CN112052723A (en) | Literacy card, and desktop scene STR method and device based on image recognition | |
| CN114639108B (en) | Appraising mark identification method, system, storage medium and equipment of subjective question | |
| JP2004094427A (en) | Slip image processor and program for realizing the same device |
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 | ||
| CP01 | Change in the name or title of a patent holder | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 330000 Building 1, maiyuan Road, Nanchang Economic and Technological Development Zone, Jiangxi Province Patentee after: Jiangxi wind vane Intelligent Technology Co.,Ltd. Address before: 330000 Building 1, maiyuan Road, Nanchang Economic and Technological Development Zone, Jiangxi Province Patentee before: JIANGXI VANEDUCATION TECHNOLOGY Inc. |
|
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Liu Fan Inventor after: Xie Degang Inventor before: Liu Fan |