Image analysis positioning method

Abstract

A multi-dimensional barcode combination image analysis and positioning method, belonging to the field of image processing recognition and automated data acquisition technology, employs a combination of multi-dimensional barcodes and answer data areas according to fixed geometric rules. The start and end codes use Data Matrix QR codes. Answer boxes of fixed size and spacing are arranged regularly between the barcodes. The barcodes use 10×10 units and six-digit encoding. The first three digits represent the barcode combination sequence number (001–999), the fourth and fifth digits represent the number of answer boxes (01–99), and the sixth digit represents the classification and sorting direction of the barcode combination. Through image analysis, the method automatically positions the starting standard barcode, automatically positions the answer data area within the rule combination, and accurately positions and analyzes the collected answer data. Editing the barcode answer sheet and scanning the collected data eliminates the need for manual positioning of the answer data area. This method is simple, efficient, and standardized, with universal software, lowering the barrier to entry for scanning and grading, and has a wide range of applications.

Description

Technical Field

[0001] This invention belongs to the field of image analysis and processing and automated data acquisition technology. In particular, it provides an image analysis and positioning method that uses multidimensional standard barcodes (including QR codes and 3D codes with unit color encoding) and data image regions to be combined according to geometric rules. The method achieves automatic positioning of data image regions and acquisition of image data through the analysis and positioning of multidimensional barcodes. Background Technology

[0002] In daily work, study, and life, we often encounter various questionnaires and exam papers. Answer sheet data and content need to be collected quickly. Scanning and image analysis to collect data for grading are effective and commonly used methods. However, in practical applications, the answer data or content is set in specific locations on the exam paper. These data locations need to correspond with the scanning and grading software system to achieve correct data analysis and collection. For example, the answer area for a question in a questionnaire is within the area on the exam paper starting from the top left corner as the origin O(0,0), with coordinates (4th quadrant) from (50, 50) to (200, 100). If image analysis and recognition are used to collect data, this area must either be pre-defined by the grading system developer (such as the answer sheet system of an optical mark reader) or the grading system user must specifically mark or set the parameters of this area during the scanning and grading operation (such as the scan cutout import position information of an online grading system). Questionnaires and answer sheets vary greatly, which makes the current marking system have a high threshold and limitations. Teachers generally cannot edit answer sheets and questionnaires themselves. It requires dedicated personnel to manage and operate them, which is cumbersome, labor-intensive, and costly.

[0003] Patent No. 201310468859.0 discloses "a barcode with optional question function and its application." The optional question barcode is used as a selection mark and score marker on the answer sheet. The barcode data setting and collection do not require positioning, eliminating the barrier to entry. However, this one-dimensional optional question barcode can only perform one-dimensional positioning and parsing, not two-dimensional positioning and parsing, thus limiting its use. Invention application No. 201911112700.9 discloses a method for collecting written content from test papers and exercises. However, this method requires manual input of the coordinates of the answer information area, which actually increases the workload in practical use. This method lacks both advancement and practicality.

[0004] Therefore, there is a need for a method that allows all teachers to easily and freely edit exam papers and answer sheets (answer areas), and for the scanning and grading software system to automatically identify and determine the location of the answer areas during scanning and grading (without requiring manual positioning), thereby realizing a multi-dimensional barcode image analysis and positioning method for data image positioning and analysis and collecting answer data. Summary of the Invention

[0005] The purpose of this invention is to provide an image analysis and localization method for a combination of multidimensional barcodes and answer data frames. The multidimensional barcode adopts a Data Matrix QR code (hereinafter referred to as DM barcode), whose encoding defines the barcode itself and the answer data area combined according to geometric rules. By using existing image analysis technology, the origin of the DM barcode is automatically analyzed and located, the answer data image area in the rule combination is automatically located, and the answer data is located, analyzed, and collected.

[0006] The technical solution adopted in this invention is as follows: A DM barcode is used, with a structure of 10×10 to 20×20 dots, or 8×18 or 8×32 dots. Its encoding defines the answer data area of ​​the barcode itself and the dotted or written answer boxes according to geometric rules. The size and geometric structure of the barcode and the answer area are fixed. During image analysis and reading, the starting position of the DM barcode is automatically determined by known image analysis technology. Then, through the geometric combination rules between the barcode and the answer area, the software system automatically calculates and determines the position and range of each dotted or written answer box, performing precise image analysis of the answer area without manual operation, and collecting answer data.

[0007] Preferably, the DM barcode uses 10×10 unit encoding.

[0008] Preferably, the combination rule for the DM barcode and the answer box is: starting DM barcode, interval, rule-ordered dotted or written answer box, interval, and ending DM barcode.

[0009] Preferably, the DM barcode uses a six-digit code. The first, second, and third digits are the serial number of the barcode combination (001-999), the fourth and fifth digits are the number of answer boxes (01-99), and the sixth digit is the classification of the barcode combination, marking five types of answer box combinations with fixed geometric structures: 1 is a forward-ordered dot-matrix answer box combination (e.g., ABCD order), 2 is a forward-ordered writing answer box combination, and 3, 4, and 5 are currently left vacant for future use. For the same combination classification, when the last digit of the barcode is 6, 7, 8, 9, or 0, the answer boxes are arranged in reverse order. When the last digit of the barcode is 6, it is a reverse-ordered dot-matrix answer box combination (e.g., DCBA), which facilitates bidirectional positioning and parsing. That is, the correspondence between the last digits of the start code and the end code is: 1-6, 2-7, 3-8, 4-9, 5-0, distinguishing the forward and reverse order of the answer box group readings within the barcode.

[0010] Preferably, the size of the DM barcode and the answer area is set as a multiple of the size of the multidimensional barcode unit.

[0011] Preferably, the width of the DM barcode unit is set to 1U = 0.5mm, the height and width of the barcode are both 5mm, the spacing is 4U, the border thickness of the dot-type answer box is 0.8U, and the inner size of the box is 3U × 8.4U (1.5 × 4.2mm), that is, the size of the answer box is 2.3 × 5mm; the border thickness of the writing-type answer box is 0.8U, and the inner size of the box is 16U × 20U, that is, the inner size of the box is 8 × 10mm.

[0012] Preferably, the dotted answer boxes are of a fixed size and are arranged in order from A to Z or 1 to 26, with fixed intervals between them.

[0013] Preferably, the writing-style answer boxes are of a fixed size, arranged in order, with no limit on the spacing between them.

[0014] As shown above, the DM barcode and all the answer boxes are combinations of fixed geometric rules. Therefore, once the position of the DM barcode is determined, the positions of all the answer boxes can be automatically calculated and determined by the software system.

[0015] Preferably, the answer box with barcode combination number 800-999 can be defined as an extended function box as needed.

[0016] Preferably, when the multidimensional barcode and the answer box are combined, and the written answer box is of any other size, it can be scaled according to the actual application needs, and the size of its answer range is determined to be a multiple of the unit.

[0017] For example, in question 2 of a test paper, there is a 4-choose-1 question with options A, B, C, and D. When editing the answer sheet, the system generates a DataMatrix barcode and 4 answer boxes with dots. The starting code is 002041 (the ending code is 002046). The last digit 1 indicates that it is the first type of dot-type answer box in a forward order. The first three digits 002 are the question number (question 2), and the fourth digit indicates that there are 4 selection boxes (4 answer boxes in total), in the order of A, B, C, and D. When analyzing the scanned image of the answer sheet, the upper left corner of the answer sheet image is set as the starting origin O(0,0). Using known image analysis technology, the lower left corner coordinates of the Data Matrix barcode are automatically identified as (x0, y0). Therefore, the positions and sizes of the marked answer boxes A, B, C, and D are determined by geometric combination rules. For example, the coordinates of the lower left corner of answer box B are (x0+10U+4U+4.6U+4U, y0), which is (x0+11.3, y0), and the coordinates of the upper right corner of B are (x0+11.3+4.6U, y0-10U), which is (x0+13.6, y0-0.5). The system performs positioning image analysis on A, B, C, and D to accurately identify and collect the answer data.

[0018] The above technical solution can also accurately locate and analyze the readings of the writing-style answer box.

[0019] The image analysis and positioning method provided by this invention uses a combination of multidimensional barcodes and answer data areas according to geometric rules. It first automatically analyzes and positions the multidimensional barcode using known technology, and then automatically positions the data image areas within the combination according to the combination rules, analyzing the answer data within the collected areas. This allows the system to generate multidimensional barcodes and answer boxes on demand when using a marking system, achieving positioning of all answer boxes. Whether editing answer sheets or scanning and marking, there is no need for manual preset or import of answer box position parameters, greatly reducing the system's usage threshold and operational difficulty. This enables a wider range of teachers and more types of answer sheets to use the scanning and marking system, collect answer data, manage big data, improve efficiency, and reduce costs. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the DM barcode combination dotted answer box of the present invention.

[0021] Figure 2 This is a schematic diagram of the DM barcode combined writing answer box of the present invention.

[0022] Figure 3 This is a schematic diagram of the technical solution of the present invention used as an examination answer sheet in Example 1.

[0023] Figure 4 This is a schematic diagram illustrating the implementation of the technical solution of the present invention when used as an examination answer sheet, involving a combination of multiple sets of dotted answer boxes.

[0024] In the diagram, 1 is the starting DM barcode (represented by the dashed box area); 2 is the dotted answer box; 3 is the interval; 4 is the ending DM barcode (represented by the dashed box area); 5 is the handwritten answer box; 6 is the question number; 7 is the respondent information area; 8 is the answer question area; 9 is the answer selection area; and 10 is the score annotation area.

[0025] Figure 1 In the code, Mark 1 is the DM barcode start code, coded as 002041. The first three digits 002 indicate question 2, the following 04 indicate that there are 4 answer boxes to fill in, and the last digit 1 indicates that it is a combination of type 1 barcode and answer box, ordered in ascending order by letters A, B, C, D. Mark 2 is the answer box, which is filled in by the examinee to select the answer. Mark 3 is a set of intervals. Mark 4 is the stop code 002046. The last digit 6 of the stop code corresponds to the last digit 1 of the start code, indicating the same combination. When reading from the end, the answer boxes are arranged in reverse order by D, C, B, A. Mark 6 is the question number indicator, which is convenient for examinees and examiners to identify.

[0026] Figure 2In the code, marker 1 is the starting DM barcode, coded as 007022. The last digit 2 indicates a forward-ordered, handwritten answer box. Marker 5 is a handwritten answer box. Marker 4 is the ending code 007027. The last digit 7 corresponds to the last digit 2 of the starting code, indicating that the answer box data is arranged in reverse order. The entire coding combination represents the second type, forward-ordered, handwritten answer box, question 7, which has two handwritten boxes.

[0027] The working principle and beneficial effects of the technical solution of this invention are as follows:

[0028] In this invention, during answer sheet editing, the software system generates a combination of DM barcodes and answer boxes as needed, eliminating the need for manual setting of answer box positions. During scanning and grading, there is no need for manual adjustment of the imported answer box position parameters; the software system automatically parses and determines the origin position of the DM barcode, and then automatically calculates and confirms the positions and ranges of all relevant answer boxes based on the DM barcode origin position using combination rules. This achieves precise positioning and analysis of the answer data image and data collection. This significantly reduces the barrier to entry and cost of using the grading system, improves the accuracy of answer data, and increases grading efficiency. Detailed Implementation

[0029] Figure 3 This is a schematic diagram of the technical solution of the present invention used as an examination answer sheet in Example 1.

[0030] During the answer sheet editing process, the software system generates a combination of a DM barcode and an answer box for each question as needed, serving as the area for candidates to answer questions and for teachers to annotate and score. Figure 3 In the middle, mark 7 is the candidate information area, which uses a writing answer box with question number 999 as a special function box extension for candidates to write their student ID or admission ticket number, making writing and collection more convenient. Mark 8 is the question area of ​​the exam paper, mark 9 is the candidate's area for filling in the answer selection mark, and mark 10 is the scoring and annotation area for the examiner, where the examiner gives scores or text annotations.

[0031] In the candidate's designated answer selection area (marked 9), there are four questions (1-4 in total). Each question has a corresponding combination of a DM barcode (Category 1, with four answer selection boxes: A, B, C, and D) and an answer selection box. The codes for these four DM barcodes are as follows: Question 1, 001041 / 001046; Question 2, 002041 / 002046; Question 3, 003041 / 003046; Question 4, 004041 / 004046. The DM barcode encoding uses a 10U×10U square dot matrix. The total width of the marked dot frame is 4.6U, and the spacing of marked 3 is 4U. When scanning and parsing the answer sheet image (assuming the upper left corner of the answer sheet image is the origin O), the software system first uses known technology to automatically determine the coordinates of the starting barcode origin O of the DM barcode combination in the four questions: (x1, y1), (x2, y2), (x3, y3), (x4, y4). Then, based on the barcode combination rules, it automatically calculates and determines... The location and area of ​​the answer selection boxes for all four questions are as follows: For example, in question 1, the height of the answer selection box is 10U. The coordinates of the lower left corner of point A are (x1+10U+4U, y1), and the coordinates of the upper right corner of point A are (x1+10U+4U+4.6U, y1-10U). The coordinates of the lower left corner of point B are (x1+10U+4U+4.6U+4U, y1), and so on. After image analysis and interpretation, the answer selection box B for question 1 is blacked out, indicating that the candidate's answer for this question is B. For these four questions, scanning and analysis reveal the answer choices as B, B, A, and C, respectively.

[0032] Figure 4 This is a schematic diagram illustrating the implementation of the technical solution of the present invention when used as an examination answer sheet, involving a combination of multiple sets of dotted answer boxes.

[0033] Actual questionnaires often have many sets of objective multiple-choice questions (marked answer boxes). To reduce the area occupied by the combined barcode and enhance its practicality, a set of DM barcodes and multiple sets of answer boxes can be combined according to geometric rules to determine the position of each answer box within the specified range of the set of barcodes (for all questions), so as to locate, analyze and collect data.

[0034] Figure 4 In this system, answer boxes numbered 801 to 899 are used as a combination of dotted answer boxes. These boxes serve as special function boxes that can be defined as needed, corresponding to multiple sets of optional questions within the area between a set of start and end barcodes. Fill in the answer box group The system uses a geometric combination of rules to determine the position of each question and each marked box within a given area. The starting and ending barcodes are 810041 / 810046, corresponding to serial number 810. The area contains 10 question combinations, each with four marked boxes (A, B, C, D). During scanning and grading, the software automatically identifies and determines the coordinates of the barcode origin O for this combination area as (x0, y0). The software system then... Fill in the answer box groupThe combination rules are used to calculate and determine the position of each marked box for the 10 questions in the area, and then perform image positioning and analysis reading. For example, for question 1, the coordinates of the upper left corner of marked box A are (x0+10U+W, y0-10U), and the coordinates of the lower right corner of marked box A are (x0+18U+W, y0-10U+4U). Similarly, the coordinates of the upper left corner of marked box D for question 10 can be determined as (x0+10U+94U+2W-8U, y0-10U+28U). The width of W can be preset and fixed by the software system or adjusted during marking.

[0035] The above are merely two preferred embodiments of the present invention and are not intended to limit the technical solutions of the present invention. Any modifications or equivalent substitutions made within the technical solutions of the present invention are included within the protection scope of the present invention.

[0036] By employing the technical solution of this invention, during the generation of exam papers and answer sheets and the data scanning and collection process, the automatic positioning of the barcode origin in the barcode combination, through the barcode geometric combination rules, achieves automatic positioning of all answer boxes, eliminating the need for manual positioning of answer boxes. The system locates, analyzes, and reads the answer data from the answer boxes, achieving precise data collection. All teachers can participate; it is simple and fast. Furthermore, the DM barcode and answer box combination has the advantages of standardization and universality, making it suitable for exam marking in various schools and for various exam questionnaires. It has a low barrier to entry, high efficiency and low cost, is easy to implement and promote, and facilitates big data management, resulting in good economic and social benefits.

Claims

1. An image analysis and localization method, characterized in that: The multidimensional barcode adopts the DM barcode (1), with a structure of 10×10~20×20 dots, or 8×18, 8×32 dots. Its encoding limits the answer data area of ​​the barcode itself and the dotted answer box (2) or the written answer box (5) to be combined according to geometric rules. The geometric structure of the barcode and the answer area is fixed, the size is fixed, and it is a multiple of the size of the DM barcode unit. When reading the image, the starting position of the DM barcode (1) or DM barcode (4) is automatically determined by the known image analysis technology. Then, through the geometric combination rules between the barcode and the answer area, the software system automatically calculates and determines the position and range of each dotted answer box (2) or each written answer box (5), and performs image analysis of the answer area without manual operation and with precise positioning to collect answer data. Furthermore, the DM barcode (1) or DM barcode (4) uses a 10×10 unit six-digit code. The first, second, and third digits are the serial number of the barcode combination 001 to 999, the fourth and fifth digits are the number of answer boxes 01 to 99, and the sixth digit is the classification of the barcode combination, marking five types of answer box combinations with fixed geometric structures. 1 is a forward-ordered dotted answer box combination, 2 is a forward-ordered writing answer box combination, and when the last digit of the barcode is 6, 7, 8, 9, or 0, the answer boxes are arranged in reverse order. When the last digit of the barcode is 6, it is a reverse-ordered dotted answer box combination, which facilitates bidirectional positioning and analysis. That is, the last digit of the start code and the end code correspond to: 1-6, 2-7, 3-8, 4-9, 5-0, which distinguishes the forward and reverse order of the answer box group readings within the barcode. Furthermore, the dotted answer boxes (2) are of a fixed size and are arranged in order from A to Z or 1 to 26 with a fixed interval (3) between them, while the written answer boxes (5) are of a fixed size and are arranged in order with no limit on the interval between them.

2. The image parsing and localization method according to claim 1, characterized in that: Set the unit width of DM barcode (1) or DM barcode (4) to 1U = 0.5mm, the height and width of the barcode to 10U, the width of the interval (3) to 4U, the frame thickness of the dotted answer box (2) to 0.8U, and the inner size of the box to 3U × 8.4U, and the frame thickness of the writing answer box (5) to 0.8U, and the inner size of the box to 16U × 20U.

3. The image analysis and localization method according to claim 1, characterized in that: The answer box combination with redundant barcode serial numbers (not limited to 800-999 as described in the instruction manual) can be defined as an extended function box as needed. The writing answer box with serial number 999 can be used as a writing box for the candidate's student ID or admission ticket number.

4. The image parsing and localization method according to claim 1, characterized in that: A set of on-demand defined extended function boxes may contain a combination of multiple sets of dotted answer boxes (2) or writing answer boxes (5) with fixed size and structure and regular arrangement in the area between the DM start code (1) and the DM end code (4).

Images