CN111814606A - An automatic scoring system and implementation method for technical image processing and pattern recognition - Google Patents

Abstract

The invention relates to the technical field of image processing, in particular to an automatic marking system for technical image processing and mode identification and an implementation method thereof.A specific position on a test paper and an answer sheet is marked in advance respectively, and the marking can be identified to quickly distinguish whether the scanned image is the test paper or the answer sheet, so that automatic marking in different modes is carried out; in the answer sheet mode, answers are obtained by identifying the positions of color blocks with specific gray values, and the answers are obtained by collecting and identifying handwriting in the test paper model, so that the method and the device can be suitable for more test scenes.

Description

An automatic scoring system and implementation method for technical image processing and pattern recognition

technical field

The invention relates to the technical field of image processing, in particular to an automatic scoring system and an implementation method for technical image processing and pattern recognition.

Background technique

The current test paper automatic identification technology is mainly aimed at automatic identification of answer sheets. Although this method saves time for marking and improves the accuracy of marking, it also has certain problems: waste of paper, and because candidates need to smear on answer sheets. , so it takes extra time, and it is easy for candidates to smear the answer in the wrong position. At the same time, the existing test does not necessarily use the answer sheet mode for different scenarios, such as some small test exams, due to cost constraints, It is impossible to use answer sheets, and the system for identifying answer sheets cannot be used, so teachers can only manually correct objective questions, which is time-consuming and labor-intensive.

SUMMARY OF THE INVENTION

In view of this, the purpose of the present invention is to provide an automatic scoring system and implementation method for technical image processing and pattern recognition, which can be compatible with answer sheet mode and test paper mode for automatic scoring, thereby adapting to more test scenarios.

In order to achieve the above object, the present invention adopts the following technical solutions:

An automatic marking system and implementation method for technical image processing and pattern recognition of the present invention, comprising the steps of:

S101, mark blank test papers and answer sheets in advance;

S102, use an optical scanner to scan the test paper or the answer sheet and convert it into an image, and determine whether the image is the test paper or the answer sheet by identifying the mark in the image;

S103, if the image is an answer sheet, convert the image into a gray-scale histogram, first obtain the serial number of the answer sheet, and then obtain the corresponding answer under the serial number by identifying the information of the graffiti area of the specific gray value under the corresponding position of the serial number;

S104, if the graphic is a test paper, first obtain the question serial number on the test paper, and then obtain the answer by collecting the handwriting at the position corresponding to the question serial number, and recognizing the handwriting;

S105, input the standard answer, compare the automatically obtained answer with the standard answer, and obtain the test score.

Further, the marks in S101 are graphics or text information printed on specific positions of blank test papers and answer sheets.

Further, the specific steps of S103 include:

S10301, first convert the image of the answer sheet into a gray-scale histogram, scan the answer sheet from left to right, and from top to bottom according to a specific interval to collect and identify the numbers, letters and handwriting on the answer sheet;

S10302, determine the arrangement of the numbers from small to large, if the numbers are arranged in multiple horizontal rows, then collect the letters directly below or directly above all the numbers, and if the numbers are arranged vertically in multiple columns, collect all the numbers directly to the left The letter in the square or right position;

S10303, if it is identified that there are letter segments arranged horizontally or vertically in the area, the area is the answer filling area, if the letter segments in the answer filling area are arranged horizontally, identify the color of the specific gray value in the horizontal letter segment Block, judge the letter represented by the color block by judging the letters on the left and right of the color block, if the letter segments are arranged vertically, judge the letter represented by the color block by judging the letters above and below the color block;

S10304, if it is recognized that there are numbers arranged horizontally or vertically in the area, but there are no letters directly above, below, to the left and right of the numbers, it is determined that the area is the test number filling area, and the identification test is passed. The position identification test number of the color block of the specific gray value in the number filling area;

S10305, extract the numbers arranged in order in the answer filling area as the serial number, and output the identified letters as the answer of the objective question corresponding to the serial number;

S10306, if it is recognized that there is handwriting in the area, it is determined that the area is an area for filling in the answer of the subjective question, and the image of the area is directly saved and output to the display device.

Further, the specific steps of S104 include:

S10401, scan and identify the image of the test paper line by line;

S10402, obtain the serial number of each question in the test paper image, the rule is: if the leftmost number in the first row of any single printed text field is a number, then determine that the number is a serial number;

S10403, obtain a handwriting closest to the printed text field where the serial number is located, and identify it, so as to determine whether it is a main body or an objective question;

S10404, if the identified handwriting contains only one or more letter information, it is judged as an objective question, and the comparison result is output by comparing the letter information under the question serial number and the pre-stored standard answer;

S10405, if the recognized handwriting contains text information, it is determined as a subjective question, and the image of the saved handwriting can be directly output to the display device.

The present invention also provides an automatic marking system for technical image processing and pattern recognition, comprising:

The scanning unit is used to scan the actual test paper or answer sheet, so as to obtain the image of the answer sheet or test sheet;

a processing unit for receiving the image from the scanning unit, and processing, identifying, comparing and outputting the image;

The display unit is used to display handwritten handwriting of subjective questions that cannot be directly judged.

The beneficial effects of the present invention are as follows: an automatic marking system and implementation method for technical image processing and pattern recognition of the present invention, by pre-marking specific positions on the test paper and the answer sheet respectively, by identifying the mark, the scanned image can be quickly distinguished Whether it is a test paper or an answer sheet, so as to carry out automatic scoring in different modes; in the answer sheet mode, the answer is obtained by identifying the position of the color block with a specific gray value, and the answer is obtained by collecting and recognizing handwriting under the test paper model, so that this The invention can be applied to more test scenarios.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present application more clearly, the following drawings will briefly introduce the drawings that need to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore do not It should be regarded as a limitation of the scope. For those of ordinary skill in the art, other related drawings can also be obtained according to these drawings without creative work:

Fig. 1 is the method flow chart of the present invention;

Fig. 2 is the detailed step flow chart of step S103 of the present invention;

FIG. 3 is a detailed flow chart of step S104 of the present invention.

Detailed ways

In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments It is a part of the embodiments of the present application, but not all of the embodiments. The components of the embodiments of the present application generally described and illustrated in the drawings herein may be arranged and designed in a variety of different configurations.

Thus, the following detailed description of the embodiments of the application provided in the accompanying drawings is not intended to limit the scope of the application as claimed, but is merely representative of selected embodiments of the application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

It should be noted that like numerals and letters refer to like items in the following figures, so once an item is defined in one figure, it does not require further definition and explanation in subsequent figures. Meanwhile, in the description of the present application, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

It should be noted that, in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any relationship between these entities or operations. any such actual relationship or sequence exists. Moreover, the terms "comprising", "comprising" or any other variation thereof are intended to encompass a non-exclusive inclusion such that a process, method, article or device that includes a list of elements includes not only those elements, but also includes not explicitly listed or other elements inherent to such a process, method, article or apparatus.

The existing automatic scoring system is generally aimed at answer sheets, using an optical scanner to scan the answer sheet to obtain images and grayscale images, and judge the correctness of the answer by judging the position of the pencil smear on the answer sheet, but this method is not applicable. In order to directly identify the test paper, in some schools or small exams, the answer sheet cannot be applied due to cost constraints, so a wide range of applications is required, which can automatically mark both through the answer sheet and directly through the test paper. Automatic marking system and implementation method.

As shown in Figure 1-Figure 2: a technical image processing and pattern recognition automatic scoring system and implementation method of the present embodiment, including steps,

S101, mark blank test papers and answer sheets in advance, and print specific marks at specific positions when printing test papers and answer sheets, for example, print the letter "S" in the upper left corner of the test paper, and print the letter "S" in the upper left corner of the answer sheet "D";

S102, use an optical scanner to scan the test paper or answer sheet and convert it into an image, and determine whether the image is a test paper or an answer sheet by identifying the marks in the image, first convert the test paper into an image by using an optical scanner, and mark it when sending it to the processing device Recognition and processing equipment pre-stored information of marks and specific marks representing test papers or answer sheets, and by comparing the marks, it can be identified whether the image is an answer sheet or a test sheet;

S103: Judging by marking, if the image is an answer sheet, convert the image into a gray-scale histogram, first obtain the serial number of the answer sheet, and then obtain the information of the scribbling area with a specific gray value under the corresponding position of the serial number to obtain the Corresponding answers, generally speaking, the objective part of the answer sheet needs to be answered with 2B pencils, so the gray value of the color blocks answered with 2B pencils is certain, and the color blocks used to detect the gray value of a specific range are The distribution position of the answer can be obtained, and the question right and wrong and the final score can be obtained after comparing with the marked distribution position;

The specific implementation process includes:

S10301, the image is scanned by an optical scanner and sent to a processing device. The processing device first converts the answer sheet image into a gray-scale histogram, and scans the answer sheet line by line from left to right and top to bottom according to a specific spacing Collect and identify the numbers, letters and handwriting on the answer sheet. The numbers can be Arabic numerals, Chinese numerals or Roman numerals. The letters are mainly "A", "B", "C" and other letters of the objective question options. body, the processing equipment can be easily recognized according to the existing image recognition technology, and the recognition of handwriting requires the establishment of a computer neural network to continuously learn and improve the accuracy;

S10302, determine the arrangement of the numbers from small to large. If the numbers are arranged in multiple horizontal rows, it means that the numbers are arranged horizontally. Generally speaking, in this case, the letters of the options are located directly below the serial numbers, and the processing equipment is allowed to collect and identify The letters directly below or directly above all numbers, and whether the corresponding number is really the serial number of the topic is inferred by whether there is a letter in the collection result. If the numbers are arranged vertically in multiple columns, collect the positive left or positive The letter in the right position, and the corresponding number is the same as the serial number of the title;

S10303, if it is identified that there are letter segments arranged horizontally or vertically in the area, the area is the answer filling area, if the letter segments in the answer filling area are arranged horizontally, identify the color of the specific gray value in the horizontal letter segment Block, judge the letter represented by the color block by judging the letters on the left and right of the color block. For example, the letter "A" and "C" are on the left and right of the position of the color block, then it is judged that the color block represents the letter "B", if The letter segments are arranged vertically, and the letters represented by the color block are judged by judging the letters above and below the color block; for example, the letters "A" and "C" are directly above and below the position of the color block, then it is determined that the color block represents the letter "B" .

S10304, if it is recognized that there are numbers arranged horizontally or vertically in the area, but there are no letters directly above, below, to the left and right of the numbers, it is determined that the area is the test number filling area, and the identification test is passed. The position identification test number of the color block of the specific gray value in the number filling area;

S10305, extract the numbers arranged in order in the answer filling area as the serial number, output the identified letters as the answer of the objective question corresponding to the serial number, extract the numbers in the test number filling area, and associate the identified test number, question serial number and answer correspondingly stand up;

S10306, if it is recognized that there is handwriting in the area, determine that the area is a filling area for subjective question answers, directly save the image of the area and output it to the display device, and the display device displays the handwriting for the teacher to watch and mark the paper.

S104, if the graphic is a test paper, first obtain the question serial number on the test paper, and then obtain the answer by collecting the handwriting at the position corresponding to the question serial number, and recognizing the handwriting to obtain the answer, when it is judged that the graphic is the test paper, adopt the method of printing and handwriting respectively to obtain the answer , The method of training machine to recognize handwriting can use the method of machine learning, establish a neural network, input a large number of handwriting into the neural network, and continuously train the handwriting model, so that the processing device can accurately recognize the handwriting and printing. body to obtain handwritten answers.

In this embodiment, the specific implementation steps of S104 include:

S10401, scan and recognize the image of the test paper line by line, refer to S10301 for the specific scanning and recognition process, but recognize printed body and handwriting;

S10402, obtain the serial number of each question in the test paper image, the rule is: if the leftmost number in the first row of any single printed text field is a number, then determine the number as the serial number, because the method of line-by-line scanning is adopted , so set the scanning spacing to be the same as the font size. After scanning a line of text, it will be a paragraph without text when scanning the next line;

S10403: Obtain a handwriting closest to the printed text field where the serial number is located, and identify it to determine whether it is a main body or an objective question. Generally speaking, the end of the objective part of the question is a bracket for filling in the answer, scan the For the printed text field of the question, the handwritten handwriting can be determined as the handwritten answer of the question by scanning the handwritten handwriting under the condition that the text is uninterrupted;

S10404, if the identified handwriting contains only one or more letter information, it is judged as an objective question, and the comparison result is output by comparing the letter information under the question serial number and the pre-stored standard answer;

S10405, if the recognized handwriting contains text information, it is determined as a subjective question, and the image of the saved handwriting can be directly output to the display device.

S105, input the standard answer, compare the automatically obtained answer with the standard answer, obtain the test score, compare the automatically obtained answer with the input standard answer, multiply the correct answer by the weight of the score, and then obtain the score of the objective question. final score.

In this embodiment, the mark in S101 is the graphic or text information printed on the specific position of the blank test paper and the answer sheet.

The present invention also provides an automatic marking system for technical image processing and pattern recognition, comprising:

The scanning unit is used to scan the actual test paper or answer sheet to obtain the image of the answer sheet or test sheet, generally an optical scanner;

The processing unit is used to receive the image from the scanning unit and process, identify, compare and output the image, which is generally a computer host connected to the optical scanner through a data line;

The display unit is used to display the handwriting of subjective questions that cannot be directly judged, and is generally a computer monitor.

The beneficial effects of the present invention are as follows: an automatic marking system and implementation method for technical image processing and pattern recognition of the present invention, by pre-marking specific positions on the test paper and the answer sheet respectively, by identifying the mark, the scanned image can be quickly distinguished Whether it is a test paper or an answer sheet, so as to carry out automatic scoring in different modes; in the answer sheet mode, the answer is obtained by identifying the position of the color block with a specific gray value, and the answer is obtained by collecting and recognizing handwriting under the test paper model, so that this The invention can be applied to more test scenarios.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be Modifications or equivalent substitutions without departing from the spirit and scope of the technical solutions of the present invention should be included in the scope of the claims of the present invention.

Claims (5)

1. a realization method of the automatic marking of technical image processing and pattern recognition, is characterized in that: comprise steps, S101, mark blank test papers and answer sheets in advance; S102, use an optical scanner to scan the test paper or the answer sheet and convert it into an image, and determine whether the image is the test paper or the answer sheet by identifying the mark in the image; S103, if the image is an answer sheet, convert the image into a gray-scale histogram, first obtain the serial number of the answer sheet, and then obtain the corresponding answer under the serial number by identifying the information of the graffiti area of the specific gray value under the corresponding position of the serial number; S104, if the graphic is a test paper, first obtain the question serial number on the test paper, and then obtain the answer by collecting the handwriting at the position corresponding to the question serial number, and recognizing the handwriting; S105, input the standard answer, compare the automatically obtained answer with the standard answer, and obtain the test score. 2 . The method for realizing technical image processing and pattern recognition according to claim 1 , wherein the marking in S101 is graphics or text information printed on specific positions of blank test papers and answer sheets. 3 . 3. The realization method of a kind of technical image processing and pattern recognition according to claim 1, is characterized in that: the concrete steps of described S103 comprise: S10301, first convert the image of the answer sheet into a gray-scale histogram, scan the answer sheet from left to right, and from top to bottom according to a specific interval to collect and identify the numbers, letters and handwriting on the answer sheet; S10302, determine the arrangement of the numbers from small to large, if the numbers are arranged in multiple horizontal rows, then collect the letters directly below or directly above all the numbers, and if the numbers are arranged vertically in multiple columns, collect all the numbers directly to the left The letter in the square or right position; S10303, if it is identified that there are letter segments arranged horizontally or vertically in the area, the area is the answer filling area, if the letter segments in the answer filling area are arranged horizontally, identify the color of the specific gray value in the horizontal letter segment Block, judge the letter represented by the color block by judging the letters on the left and right of the color block, if the letter segments are arranged vertically, judge the letter represented by the color block by judging the letters above and below the color block; S10304, if it is recognized that there are numbers arranged horizontally or vertically in the area, but there are no letters directly above, below, to the left and right of the numbers, it is determined that the area is the test number filling area, and the identification test is passed. The position identification test number of the color block of the specific gray value in the number filling area; S10305, extract the numbers arranged in order in the answer filling area as the serial number, and output the identified letters as the answer of the objective question corresponding to the serial number; S10306, if it is recognized that there is handwriting in the area, it is determined that the area is an area for filling in the answer of the subjective question, and the image of the area is directly saved and output to the display device. 4. the automatic marking system and realization method of a kind of technical image processing and pattern recognition according to claim 1, is characterized in that: the concrete steps of described S104 comprise: S10401, scan and identify the image of the test paper line by line; S10402, obtain the serial number of each question in the test paper image, the rule is: if the leftmost number in the first row of any single printed text field is a number, then determine that the number is a serial number; S10403, obtain a handwriting closest to the printed text field where the serial number is located, and identify it, so as to determine whether it is a main body or an objective question; S10404, if the identified handwriting contains only one or more letter information, it is judged as an objective question, and the comparison result is output by comparing the letter information under the question serial number and the pre-stored standard answer; S10405, if the recognized handwriting contains text information, it is determined as a subjective question, and the image of the saved handwriting can be directly output to the display device. 5. an automatic marking system for technical image processing and pattern recognition, characterized in that: comprising: The scanning unit is used to scan the actual test paper or answer sheet, so as to obtain the image of the answer sheet or test sheet; a processing unit for receiving the image from the scanning unit, and processing, identifying, comparing and outputting the image; The display unit is used to display the handwritten handwriting of subjective questions that cannot be directly judged by comparison.

Images