CN112580503A

CN112580503A - Operation correction method, device, equipment and storage medium

Info

Publication number: CN112580503A
Application number: CN202011497238.1A
Authority: CN
Inventors: 邓辉斌; 彭兴
Original assignee: Shenzhen Yuande Education Technology Co ltd
Current assignee: Shenzhen Yuande Education Technology Co ltd
Priority date: 2020-12-17
Filing date: 2020-12-17
Publication date: 2021-03-30

Abstract

The invention provides a job correction method, a device, equipment and a storage medium, wherein the job correction method comprises the following steps: acquiring pictures of jobs to be changed in batches through terminal equipment; segmenting the picture one by one through feature point identification, and dividing the picture into a plurality of sub-pictures; dividing the sub-picture into a question part and an answer part based on font identification, and associating the question part with the answer part; when the sub-picture is corrected, the subject part of the sub-picture is used for searching in a preset picture library to obtain a standard answer corresponding to the subject part of the sub-picture; acquiring a topic type to which a topic part of a sub-picture belongs, and loading a corresponding correction method based on the topic type; and finishing correction on the theme part of the sub-picture by using a correction method. The job correction method can load the corresponding correction method based on the title type, expand the correctable title type and improve the job correction efficiency.

Description

Operation correction method, device, equipment and storage medium

Technical Field

The invention relates to the technical field of education, in particular to a method and a device for homework correction and a storage medium of a device.

Background

The current correction technology on the market mainly aims at objective questions such as selection questions and judgment questions. By means of mobile phone, tablet, writing board and other equipment, the answer of the selected question is judged to set specific mark, and the student selects the specific mark to complete the job answering. The current basis can be analyzed and judged for specific recognition, thereby achieving correction of the subject. However, in actual work, the coverage of question types is wide, and the selection and judgment of questions are far from sufficient.

Accordingly, the prior art is yet to be improved and developed.

Disclosure of Invention

The invention mainly aims to solve the technical problems that the existing operation correction method only can meet the correction of selection and judgment questions and cannot meet the requirements of users.

A first aspect of the present invention provides a job approval method, including:

acquiring pictures of jobs to be changed in batches through terminal equipment;

performing topic-by-topic segmentation on the picture through feature point identification, and dividing the picture into a plurality of sub-pictures;

dividing each sub-picture into a question part and an answer part based on font identification, and associating the question part and the answer part of each sub-picture;

when the correction is carried out on each sub-picture, the topic part of each sub-picture is used for searching in a preset picture library to obtain a standard answer corresponding to the topic part of each sub-picture;

acquiring the topic type of the topic part of each sub-picture, and loading a corresponding correction method based on the topic type;

and finishing the correction of the answer part of each sub-picture by the correction method.

Optionally, in a first implementation manner of the first aspect of the present invention, the obtaining a topic type to which the topic part of each sub-picture belongs, and the loading a corresponding correction method based on the topic type includes:

if the question type of the question part of the sub-picture is an objective question, directly comparing whether the symbol or letter in the answer part is the same as the symbol or letter of the standard answer;

if the topic type of the topic part of the sub-picture is a subjective topic, continuing to judge the subject to which the topic part belongs, and loading a corresponding comparison method according to the subject to which the topic part belongs.

Optionally, in a second implementation manner of the first aspect of the present invention, if the topic type to which the topic part of the sub-picture belongs is a subjective topic, continuing to determine the subject to which the topic part belongs, and loading a corresponding comparison method according to the subject to which the topic part belongs includes:

and if the subject to which the question part belongs is English or Chinese and the question part is a composition question, judging whether the meaning of the text in the answer part is the same as that of the text in the standard answer or not through semantic recognition.

Optionally, in a third implementation manner of the first aspect of the present invention, if the topic type to which the topic part of the sub-picture belongs is a subjective topic, continuing to determine the subject to which the topic part belongs, and loading a corresponding comparison method according to the subject to which the topic part belongs further includes:

and if the subject to which the question part belongs is mathematical, physical or chemical and the answer part is a calculation formula, judging whether the calculation formula in the answer part is the same as the calculation formula in the standard answer or not through image identification.

Optionally, in a fourth implementation manner of the first aspect of the present invention, if the subject to which the question portion belongs is math, physics, or chemistry, and the answer portion is a calculation formula, determining, by image recognition, whether the calculation formula in the answer portion is the same as the calculation formula in the standard answer includes:

and if the similarity between the calculation formula in the answer part and the calculation formula in the standard answer is greater than a preset threshold value, judging that the calculation formula in the answer part is correct.

Optionally, in a fifth implementation manner of the first aspect of the present invention, the font corresponding to the question portion is a printed font, and the font corresponding to the question answering portion is a handwritten font.

Optionally, in a sixth implementation manner of the first aspect of the present invention, the terminal device includes a mobile phone, a tablet computer, and a tablet.

The second aspect of the present invention also provides a job correction apparatus, including:

the acquisition module is used for acquiring pictures of the jobs to be changed through the terminal equipment;

the first segmentation module is used for segmenting the picture into a plurality of sub-pictures one by one through feature point identification;

the second segmentation module is used for segmenting each sub-picture into a question part and an answer part based on font identification and associating the question part and the answer part of each sub-picture;

the retrieval module is used for retrieving the topic part of each sub-picture in a preset picture library when correcting each sub-picture to obtain a standard answer corresponding to the topic part of each sub-picture;

the loading module is used for acquiring the topic type of the topic part of each sub-picture and loading a corresponding correction method based on the topic type;

and the correcting module is used for finishing correcting the answer part of each sub-picture by the correcting method.

A third aspect of the present invention provides a job correction apparatus including: a memory having instructions stored therein and at least one processor, the memory and the at least one processor interconnected by a line;

the at least one processor invokes the instructions in the memory to cause the job modification device to perform any of the job modification methods described above.

A third aspect of the present invention provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements a job batching method as described in any one of the above.

Has the advantages that: the invention provides a job correction method, a device, equipment and a storage medium, wherein the job correction method comprises the following steps: acquiring pictures of jobs to be changed in batches through terminal equipment; segmenting the picture one by one through feature point identification, and dividing the picture into a plurality of sub-pictures; dividing the sub-picture into a question part and an answer part based on font identification, and associating the question part with the answer part; when the sub-picture is corrected, the subject part of the sub-picture is used for searching in a preset picture library to obtain a standard answer corresponding to the subject part of the sub-picture; acquiring a topic type to which a topic part of a sub-picture belongs, and loading a corresponding correction method based on the topic type; and finishing correction on the theme part of the sub-picture by using a correction method. The job correction method can load the corresponding correction method based on the title type, expand the correctable title type and improve the working efficiency of the user.

Drawings

FIG. 1 is a schematic diagram of an embodiment of a method for validating a job;

FIG. 2 is a schematic diagram of an embodiment of a job approval apparatus according to the present invention;

FIG. 3 is a schematic diagram of an embodiment of a job batching device according to the present invention.

Detailed Description

The embodiment of the invention provides a method, a device, equipment and a storage medium for correcting operation.

The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

For convenience of understanding, a specific flow of an embodiment of the present invention is described below, and referring to fig. 1, a first aspect of the present invention provides a job approval method, where the job approval method includes:

s100, acquiring pictures of jobs to be changed through terminal equipment;

in this embodiment, the premise of picture processing is that picture input is available, and picture input can be realized through electronic devices such as a tablet, a tablet computer, and a mobile phone, and the picture input is divided into two types, one type is picture whole input, answers and questions are on the same picture, and the other type is picture cutting or tablet single-question answering, so that automatic association between the answers and the questions is realized.

S200, segmenting the picture one by one through feature point identification, and dividing the picture into a plurality of sub-pictures;

in this embodiment, taking the whole picture as an example, the answer and the title are on the same picture, so that the picture needs to be cut according to the picture features one by one, and when the feature points are identified, the labels (such as "1", "2", etc.) of the titles can be used as the feature points for identifying different titles.

S300, dividing each sub-picture into a question part and an answer part based on font identification, and associating the question part and the answer part of each sub-picture;

in this embodiment, after the picture is cut one-by-one, the handwriting and print recognition technology needs to be used again to realize the segmentation of the question stem and the answer area of the sub-picture.

S400, when correcting each sub-picture, retrieving the topic part of each sub-picture in a preset picture library to obtain a standard answer corresponding to the topic part of each sub-picture;

in this embodiment, a question bank server is preset, a large number of test question pictures are stored in the question bank server to form a picture bank, the test question pictures in the picture bank can also be generated from daily correction pictures, during correction, a question part obtained by splitting a sub-picture is searched in the picture bank to match a correct question, and the test question pictures include answers to the test questions.

S500, obtaining a topic type to which the topic part of each sub-picture belongs, and loading a corresponding correction method based on the topic type;

in this embodiment, the question type of the question is obtained according to the question part corresponding to the picture, if the question is an objective question, the answer comparison is directly performed to judge, if the question is a subjective question, an OCR recognition technology is firstly used to recognize the answer part into a text, and the OCR recognition result is compared with the text of a standard answer.

S600, finishing correction of the answer part of each sub-picture by the correction method.

Specifically, the current correction technology in the market mainly aims at objective questions, such as selection questions and judgment questions, and due to the fact that the questions of the selection and judgment types have the characteristic of quick correction, a plurality of questions are modified into selection or judgment questions from writing questions through question type modification in the market at present, and the problem of correction of partial questions is solved to a certain extent. However, the investigation aspects of different themes are different, and the diversification of the themes is continuously explored. The scheme can effectively expand the programming to realize the automatic correction and the auxiliary correction of the title types, and improve the working efficiency of the user.

In an optional implementation manner of the first aspect of the present invention, the obtaining a topic type to which the topic part of each sub-picture belongs, and the loading a corresponding correction method based on the topic type includes:

in this embodiment, the answers to the objective questions are generally letters (e.g., a \ B \ C \ D) and symbols (e.g., √ and ×), and since the evaluation of the objective questions is simple, it is only necessary to compare whether the symbols or letters in the answer part are the same as those in the standard answer.

In this embodiment, it is also possible to not determine whether the question portion is a subjective question or an objective question, that is, identify the content of the question answering portion as a text, compare the text with the text of the standard answer, and determine that the text is an objective question if the comparison is consistent and the number of words is less than a certain threshold. If the contrast is different, the subjective question is considered, and the subsequent steps are required.

In an optional implementation manner of the first aspect of the present invention, if the topic type to which the topic part of the sub-picture belongs is a subjective topic, the method of continuously determining the subject to which the topic part belongs, and loading a corresponding comparison method according to the subject to which the topic part belongs includes:

In this embodiment, if the comparison is different or not suitable for the question type using text comparison, such as composition questions of english and chinese, diagnosis can be performed through the AI semantic understanding function, and various rules are set to realize preliminary scoring.

In an alternative implementation manner of the first aspect of the present invention, when the text included in the answer section is recognized, the handwritten character is changed greatly due to different handwriting habits of people, and strokes and parts are stuck frequently. Therefore, we recognize the line first, and then recognize the character. The specific algorithm is as follows: the achievement list image is horizontally projected, and the pixel line projected by each line of text is the number of black pixels of the pixel line. If the projected image is a text, the number of black pixels is large, and the horizontal projection result is that the horizontal projection of the pixel row is 1, and if the projected pixel row is mostly 0, we can assume that the ith pixel row is a row gap, and the horizontal projection of the pixel row is 0. By using the method, all the projections are carried out, the gaps among the texts can be projected, and each text line is judged and segmented according to the number of black pixels of the pixel line, so that the full version of text segmentation is completed.

In an alternative implementation of the first aspect of the present invention, since the segmentation of the characters is generally performed on the extracted text, the segmentation of the characters is performed on the basis of the segmentation of the previous lines. Because the sizes of different handwritten characters are inconsistent due to the difference between Chinese characters and the difference between the handwritten characters, before character segmentation is carried out, the size of each character is integrally adjusted by the text, so that the characters are consistent in size and become characters with uniform sizes.

The size normalization of Chinese characters is generally performed by a projection method, and the specific algorithm for normalizing Chinese characters is as follows:

and converting the picture to be detected into a gray image, and performing binarization processing on the obtained gray image to obtain a binary image.

Suppose that the pixel point is mxy ═ Xi, j, where (1 ≦ i ≦ l, 1 ≦ j ≦ w), l and w indicate the length and width of M, Xi, j is the pixel point in the ith row and the jth column, the normalized chinese character size is (l2, w2), and the original size is (l, w).

Assuming that the original character size/normalized Chinese character size is percent, l, w, l ', w' need to satisfy:

{l/l′＝percent，w/w′＝percent

and converting l '═ l/percent and w' ═ w/percent of all Chinese character pixels in the original Chinese character, namely completing the normalization of the Chinese character.

When the coordinate conversion is performed, l 'and w' in the result obtained by l '═ l/percent and w' ═ w/percent have decimal parts, and the decimal parts need to be rounded up and rounded.

In an optional implementation manner of the first aspect of the present invention, if the topic type to which the topic part of the sub-picture belongs is a subjective topic, continuing to determine the subject to which the topic part belongs, and loading a corresponding comparison method according to the subject to which the topic part belongs further includes:

In an optional implementation manner of the first aspect of the present invention, if the subject to which the question portion belongs is mathematic, physical, or chemical, and the answer portion is a calculation formula, determining, by image recognition, whether the calculation formula in the answer portion is the same as the calculation formula in the standard answer includes:

In this embodiment, for mathematics, physics, and chemistry disciplines, which contain formulas, formula recognition technology can be used for recognition, the recognition result is compared with the standard answer, and if the result is consistent, correction is completed. If the answer is different from the standard answer, the similarity comparison is performed by using the picture recognition technology. Setting a fixed threshold, and if the similarity is greater than the threshold, prejudging that the similarity is correct; if the similarity is smaller than the threshold value, the judgment is wrong. The user can set the threshold value by himself according to the situation, and manual score correction is supported.

In an optional implementation manner of the first aspect of the present invention, the font corresponding to the question portion is a printed font, and the font corresponding to the answer portion is a handwritten font.

In an optional implementation manner of the first aspect of the present invention, when content included in the question portion and the answer portion is obtained, an image is obtained by an OCR technology, then preprocessing such as Hough line detection, transverse projection, median filtering technology and the like is performed on the image, an interference line segment in the image, for example, a horizontal line on a workbook, is blanked, then layout analysis is performed on the image, question and answer information in the image are extracted, and the extracted information block is segmented by a connected domain method. And then, carrying out radical segmentation on the extracted information block through size normalization of the Chinese characters and Chinese character refinement, extracting character features through density features, grid features and peripheral features of the characters, and matching through the density features, the grid features and the peripheral features of the characters and the density features, the grid features and the peripheral features of the characters in the template to obtain a recognition result.

In an optional implementation manner of the first aspect of the present invention, the terminal device includes a mobile phone, a tablet computer, and a tablet.

In conclusion, the method is mainly characterized in that the method supports the pre-correction of subjective questions, particularly questions comprising formulas and graphs; the whole correcting system can basically cover the correcting of the current common question types, and comprehensively applies an OCR recognition technology, a formula recognition technology, a semantic understanding technology, a picture recognition technology and a picture comparison technology. The invention compares the similarity of the pictures answered by the students and the pictures corresponding to the standard answers by using a picture comparison technology, sets a certain threshold value and carries out primary judgment on the answers of the students. The method comprises the steps of putting pictures of all students in a picture library, matching the pictures of the whole answer picture library by using standard answers of corresponding questions, marking the similarity of each picture, classifying all the pictures according to different similarity rules, realizing that a teacher or a training institution modifies the answers of a plurality of students at the same time, and bringing more beneficial technical effects compared with the prior art.

Referring to fig. 2, a second aspect of the present invention provides a job correction apparatus including:

the acquisition module 10 is used for acquiring pictures of the jobs to be changed through the terminal equipment;

the first segmentation module 20 is configured to segment the picture topic by feature point identification, and divide the picture into a plurality of sub-pictures;

a second segmentation module 30, configured to segment each of the sub-pictures into a question portion and an answer portion based on font identification, and associate the question portion and the answer portion of each of the sub-pictures;

a retrieval module 40, configured to, when modifying each sub-picture, retrieve the topic part of each sub-picture in a preset picture library to obtain a standard answer corresponding to the topic part of each sub-picture;

a loading module 50, configured to obtain a topic type to which the topic part of each sub-picture belongs, and load a corresponding correction method based on the topic type;

and the correcting module 60 is configured to complete correcting the answer portion of each sub-picture by the correcting method.

In an optional implementation manner of the second aspect of the present invention, the job modification apparatus further includes:

the first comparison module is used for directly comparing whether the symbols or letters in the answer part are the same as the symbols or letters of the standard answer or not if the question type of the question part of the sub-picture is an objective question;

and the second comparison module is used for continuously judging the subjects to which the subject parts belong if the subject types to which the subject parts of the sub-pictures belong are subjective subjects, and loading a corresponding comparison method according to the subjects to which the subject parts belong.

In an optional implementation manner of the second aspect of the present invention, the second comparing module is further configured to determine, through semantic recognition, whether a meaning of a text in the answer part is the same as a meaning of a text in the standard answer, if the subject to which the question part belongs is english or chinese and the question part is a composition question.

In an optional implementation manner of the second aspect of the present invention, the second comparing module is further configured to determine, through image recognition, whether a calculation formula in the question answering portion is the same as a calculation formula in the standard answer if the subject to which the question answering portion belongs is mathematical, physical, or chemical and the question answering portion is a calculation formula.

In an optional implementation manner of the second aspect of the present invention, the second comparing module is further configured to determine that the calculation formula in the answer section is correct if the similarity between the calculation formula in the answer section and the calculation formula in the standard answer is greater than a preset threshold.

In an optional implementation manner of the second aspect of the present invention, the font corresponding to the question portion is a printed font, and the font corresponding to the answer portion is a handwritten font.

In an optional implementation manner of the second aspect of the present invention, the terminal device includes a mobile phone, a tablet computer, and a tablet.

Fig. 2 above describes a job modification apparatus in an embodiment of the present invention in detail from the perspective of a modular functional entity, and a job modification apparatus in an embodiment of the present invention is described in detail from the perspective of hardware processing.

Fig. 3 is a schematic structural diagram of a job modifying apparatus according to an embodiment of the present invention, which may include one or more processors 70 (CPUs) (e.g., one or more processors) and a memory 80, and one or more storage media 90 (e.g., one or more mass storage devices) for storing applications or data, based on the fact that the job modifying apparatus may generate relatively large differences due to different configurations or performances. The memory and storage medium may be, among other things, transient or persistent storage. The program stored on the storage medium may include one or more modules (not shown), each of which may include a series of instruction operations for implementing an apparatus for automatically requesting comments for an order. Still further, the processor may be configured to communicate with a storage medium to execute a series of instruction operations in the storage medium on the job modification apparatus.

The work wholesale device may also include one or more power supplies 100, one or more wired or wireless network interfaces 110, one or more input-output interfaces 120, and/or one or more operating systems, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will appreciate that the configuration of the work batching device shown in FIG. 3 does not constitute a limitation of the work batching device and may include more or less components than those shown, or some components in combination, or a different arrangement of components.

The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, or a volatile computer-readable storage medium, having stored therein instructions, which, when executed on a computer, cause the computer to perform the steps of the job modification method.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses, and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A job approval method, comprising:

acquiring pictures of jobs to be changed in batches through terminal equipment;

2. The job approval method according to claim 1, wherein the obtaining of the topic type to which the topic part of each of the sub-pictures belongs, and the loading of the corresponding approval method based on the topic type comprises:

3. The method according to claim 2, wherein if the topic type of the topic part of the sub-picture is a subjective topic, the method continues to determine the subject to which the topic part belongs, and the loading a corresponding comparison method according to the subject to which the topic part belongs includes:

4. The method according to claim 2, wherein if the topic type of the topic part of the sub-picture is a subjective topic, the method continues to determine the subject to which the topic part belongs, and the method for loading the corresponding contrast according to the subject to which the topic part belongs further comprises:

5. The method according to claim 4, wherein if the subject to which the question portion belongs is mathematical, physical, chemical, and the answer portion is a calculation formula, determining whether the calculation formula in the answer portion is the same as the calculation formula in the standard answer by image recognition comprises:

6. The job approval method according to claim 1, wherein a font corresponding to the question portion is a printed font, and a font corresponding to the answer portion is a handwritten font.

7. The job approval method according to claim 1, wherein the terminal device includes a mobile phone, a tablet computer, and a handwriting board.

8. A job approval apparatus, characterized by comprising:

9. A job correction apparatus, characterized by comprising: a memory having instructions stored therein and at least one processor, the memory and the at least one processor interconnected by a line;

the at least one processor invoking the instructions in the memory to cause the job approval apparatus to perform the job approval method of any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the job batching method as recited in any one of claims 1-7.