Paper textbook publishing digitalization method based on artificial intelligence

Abstract

The invention relates to the technical field of image recognition, in particular to a paper textbook publishing digitization method based on artificial intelligence, and the method comprises the steps: extracting sample image information of different key regions of paper textbooks of all subjects, and building a database; by determining the feature influence factors of the key areas of the paper textbooks of different subjects, the sensitive feature characterization parameters of the key areas of the paper textbooks of different subjects can be calculated, so that the sensitive feature sorting sequence of the key areas of the paper textbooks of different subjects can be quickly determined; whether each key area of the uploaded paper textbook is abnormal or not can be accurately acquired by acquiring the actual image of the paper textbook uploaded by the user side and the sample image of the corresponding key area, so that the auditing and publishing precision of the paper textbook is improved; the problems that in the prior art, due to the fact that sensitive features of different key areas of paper textbooks of different subject types are different, the checking and publishing efficiency and precision of the paper textbooks of corresponding subjects are low are solved.

Description

Technical Field

[0001] The present invention relates to the field of image recognition technology, and in particular, to a digitalization method for paper textbook publishing based on artificial intelligence. Background Art

[0002] The rise of artificial intelligence technology has brought new breakthroughs to the recognition of paper textbooks. Deep learning algorithms, such as convolutional neural networks (CNNs), have achieved remarkable results in the field of image recognition. Through training with a large amount of labeled data, the model can automatically learn the features in the image, thereby realizing the efficient recognition of paper textbooks. In teaching systems, artificial intelligence image recognition technology is widely used in aspects such as digitalization of textbook content, integration of teaching resources, and intelligent tutoring. For example, by identifying knowledge points in textbooks, personalized learning suggestions and tutoring are provided for students.

[0003] With the development of digital technology, the recognition technology of paper textbooks has also been continuously improving. In the early stage, it was mainly based on simple optical character recognition (OCR) technology. By scanning and recognizing the text on paper textbooks, it was converted into an editable text format. Although this method can process text information, its processing ability for complex content such as images and charts in textbooks is limited. Later, recognition technologies based on image processing emerged, which can comprehensively process the page layout, images, text, etc. of textbooks, improving the accuracy and integrity of recognition. For example, through image segmentation technology, different elements on the page are separated and recognized and analyzed respectively.

[0004] Chinese Patent Publication No.: CN113569540B discloses a method and device for generating test papers based on social science textbooks. The method includes identifying a paper social science textbook to obtain an electronic document corresponding to the textbook; searching for multiple text segments (constituted by one or consecutive multiple sentences in the electronic document) that conform to text features in the electronic document, and using a semantic recognition model to identify the text segments containing professional terms among the multiple text segments; generating at least one question according to each text segment containing professional terms; respectively extracting multiple knowledge points from the table of contents and each question in the electronic document; combining the knowledge points and the position information of the text segments used to generate questions, and constructing a knowledge graph through knowledge fusion; according to the specified test paper generation parameters, searching in the knowledge graph to obtain a set of test questions that meet the test paper generation parameters, and combining the multiple questions included in the set of test questions into a test paper. This solution effectively improves the efficiency of automatically generating test papers by using textbooks to generate questions.

[0005] Chinese Patent Publication No.: CN114220305A discloses a teaching system based on artificial intelligence image recognition technology, including an image input module, an AI recognition module, a teaching knowledge point input module, a modification and input module, a label and index module, a touch panel terminal, a data interaction module, and a knowledge point display module. The present invention applies AI technology and image recognition technology to a modern teaching system. Students sort out the knowledge points in relevant textbooks. For the relevant difficult points or unknown knowledge points in the textbooks, the image recognition technology and AI technology are used to recognize and calculate the difficult points or unknown knowledge points on the paper textbooks, and finally obtain corresponding answers from the database, without the user spending time to sort out and manually input.

[0006] However, the following problems still exist in the prior art: In the existing methods of recognizing paper textbooks and the methods of artificial intelligence image recognition teaching systems, the data representativeness of different key areas of paper textbooks of different subject types in the complex dimension of image recognition is not considered. In actual situations, some paper textbooks of certain subject types have obvious features in some key areas, and the regional features of some paper textbooks of certain subject types are more similar to the corresponding key areas of other paper textbooks. If the same analysis method is used to review whether various paper textbooks are published, when facing a large amount of data information on the review and publication of paper textbooks, the review and publication speed is slow and the efficiency is low. Summary of the Invention

[0007] Therefore, the present invention provides a digital method for the publication of paper textbooks based on artificial intelligence to overcome the problem in the prior art that the data representativeness of different key areas of paper textbooks of different subject types in the complex dimension of image recognition is not considered, and if the same analysis method is used to review whether various paper textbooks are published, it will lead to a slow review and publication speed and low efficiency when facing a large amount of data information on the review and publication of paper textbooks.

[0008] To achieve the above object, the present invention provides a digital method for the publication of paper textbooks based on artificial intelligence, including: Extracting sample images of different key areas of paper textbooks of different subject types to determine the characteristic influence factors of each key area of paper textbooks of different subject types; Calculating the sensitive feature representation parameters of each subject type of paper textbook according to the characteristic influence factors of each key area of paper textbooks of different subject types to determine the sensitive feature sorting sequence of each key area of a single subject type of paper textbook; Obtaining the sample image of the paper textbook uploaded by the user terminal and the corresponding textbook subject type, obtaining the sensitive feature representation parameters of each key area of the paper textbook of the corresponding subject type according to the actual textbook subject type, dividing the abnormal tendency label of the paper textbook, and analyzing the actual image of the paper textbook according to the determination result, including, If it is the first abnormal tendency label, obtain the sensitive feature sorting sequence of each key area of the paper-based textbook of the corresponding subject type, determine the comparison order of the local sample images of the key areas of the paper-based textbook to be published according to the sensitive feature sorting sequence, and call and compare the key areas of the paper-based textbook of the corresponding subject type with the corresponding key areas of the actual paper-based textbook in turn to determine the similarity, so as to determine whether there is an abnormality; If it is the second abnormal tendency label, obtain the complete sample image of the paper-based textbook and compare it with the sample image of the corresponding complete paper-based textbook to determine the similarity, so as to determine whether there is an abnormality.

[0009] Furthermore, the process of determining the characteristic influence factors of each key area of the paper-based textbooks of different subject types includes Compare the sample images of each key area of a single subject type with the local sample images of the paper-based textbooks of the corresponding key areas of other subject types respectively; Solve the color saturation and obtain the edge complexity of each key area of the paper-based textbook of a single subject type.

[0010] Furthermore, obtaining the edge complexity of each key area of the paper-based textbook of a single subject type includes Calibrate the edge contour lines of each key area of the paper-based textbook of a single subject type and extract the edge texture density; Calculate the coincidence degree between the edge contour line in the paper-based textbook and the reference contour line, and take the ratio of the coincidence degree to the reference contour coincidence threshold as the first edge complexity influence factor; Calculate the ratio of the edge texture density in the paper-based textbook to the reference texture density threshold as the second edge complexity influence factor; Determine the sum of the first edge complexity influence factor and the second edge complexity influence factor as the edge complexity.

[0011] Furthermore, the process of calculating the sensitive feature characterization parameter of the paper-based textbook of a single subject type includes Obtain the characteristic influence factors of the paper-based textbook of a single subject type, including the average color saturation and the edge complexity; Determine the ratio of a predetermined color saturation threshold to the average color saturation as the first characteristic influence factor; Determine the ratio of the edge complexity to a predetermined edge complexity threshold as the second characteristic influence factor; Determine the sum of the first characteristic influence factor and the second characteristic influence factor as the sensitive feature characterization parameter.

[0012] Furthermore, the process of determining the sensitive feature sorting sequence of each key area of the paper-based textbook of a single subject type includes Determine the sensitive feature characterization parameters corresponding to each key area of the paper textbook of a single subject type; Arrange the sensitive feature characterization parameters in descending order to obtain the sensitive feature sorting sequence.

[0013] Further, the process of dividing the abnormal tendency labels of the paper textbook includes, Extract the significant sensitive feature characterization parameters of each key area of the paper textbook of the corresponding subject type; If the significant sensitive feature characterization parameter is greater than the predetermined sensitive feature characterization parameter threshold, it is determined as the first abnormal tendency label; If the significant sensitive feature characterization parameter is less than or equal to the predetermined sensitive feature characterization parameter threshold, it is determined as the second abnormal tendency label.

[0014] Further, determining the comparison order for the actual image of the paper textbook to be published according to the sensitive feature sorting sequence includes, Sequentially determine the corresponding key areas according to the sensitive feature sorting sequence, and compare the key areas with the actual image; Among them, each sensitive feature characterization parameter corresponds one-to-one with the key area serial number.

[0015] Further, sequentially call the key areas of the paper textbook of the corresponding subject type and compare them with the corresponding key areas of the actual paper textbook according to the comparison order to determine the similarity to determine whether there is an abnormality, including, If the similarity corresponding to the actual image of each key area is less than or equal to the predetermined key area similarity threshold, it is determined that there is an abnormality.

[0016] Further, obtain the complete sample image of the paper textbook and compare it with the sample image of the corresponding complete paper textbook to determine the similarity to determine whether there is an abnormality, including, Obtain the complete sample image of the paper textbook uploaded by the user terminal and compare it with the complete sample image of the paper textbook of the corresponding subject type to determine the similarity; If the similarity between the complete sample image of the current paper textbook and the sample image of the corresponding complete paper textbook is less than or equal to the predetermined complete sample image similarity threshold, it is determined that there is an abnormality.

[0017] Further, the sample image of the paper textbook uploaded by the user terminal needs to include the complete sample image of the paper textbook and the local sample images of each key area of the paper textbook.

[0018] Compared with the prior art, the present invention establishes a database by extracting the sample image information of different key regions of paper textbooks of each subject type. By determining the characteristic influencing factors of each key region of paper textbooks of different subject types, the sensitive characteristic representation parameters of each key region of paper textbooks of each subject type can be calculated, so as to quickly determine the sensitive characteristic sorting sequence of each key region of paper textbooks of each subject type, improving the review efficiency before the publication of paper textbooks. At the same time, by obtaining the actual image of the paper textbook uploaded by the user terminal and the sample image of the corresponding key region, the significant sensitive characteristic representation parameters of each key region of the uploaded paper textbook can be accurately obtained, improving the review and publication accuracy of paper textbooks, and overcoming the problems in the prior art that the different sensitive characteristics of different key regions of paper textbooks of different subject types lead to low efficiency and low accuracy in determining the review and publication of the corresponding subject paper textbooks. By determining the comparison order of the actual images of each key region of the paper textbook, and sequentially calling the actual image of the corresponding key region and the sample image of the corresponding key region for comparison according to the comparison order, it can accurately determine whether there is an abnormality in the paper textbook uploaded by the user terminal, and determine whether to enable publication.

[0019] In particular, the present invention compares the sample images of different key regions of a single-subject paper textbook with the image information of the corresponding key regions of other-subject paper textbooks, comprehensively considering the differences in the publication review of paper textbooks in different key regions. In actual situations, some key regions of paper textbooks of certain subjects have relatively obvious sensitive characteristics, and some sensitive characteristics of the key regions of paper textbooks of some subject types are relatively similar to the corresponding key regions of other subject types. Therefore, the sensitive characteristic representativeness of the sample images of each key region is different. In some cases, whether there is an abnormality can be identified based on the sample images of the key regions with obvious sensitive characteristic representativeness. Therefore, the present invention considers determining the sensitive characteristic representation parameters of each key region, providing data support for the way of selecting sensitive characteristic representation parameters when analyzing the review and publication of the actual images of the uploaded paper textbooks, and then adaptively selecting for the actual review and publication analysis, reducing the review processing volume while ensuring accuracy, and improving the review efficiency of paper textbook publication.

[0020] In particular, the present invention provides an important feature dimension for whether there is an abnormality in the publication review of paper textbooks by obtaining the color saturation and edge complexity of each key region of a single-subject type paper textbook. The edge complexity reflects the structural characteristics of each key region of the paper textbook, and the attribute factors of different subject types may cause the edge complexity of different subject types of paper textbooks to be different. Combining the edge complexity with the color saturation characterizes the influence of the sensitive characteristics of the key region and the data representativeness.

[0021] In particular, the present invention analyzes the actual images in the case where the characterization parameter of the significantly sensitive feature is greater than the threshold benchmark of the predetermined sensitive feature characterization parameter. In the above case, there are key areas with prominent sensitive features in the paper textbooks representing the corresponding subject types, and the data representativeness is strong. Therefore, by determining the comparison order of the actual images of each key area of the paper textbooks for the subject type according to the sorting sequence of the sensitive features, and sequentially calling the actual images of the corresponding key areas and the local sample images of the corresponding key areas for comparison according to the comparison order, the noise data introduced by the comparison of the full-area sample images can be reduced. Moreover, by preferentially analyzing the actual images corresponding to the areas with strong image representativeness, the efficiency of the publication review of paper textbooks can be improved on the premise of ensuring reliability, and the accuracy of the publication review of paper textbooks can be guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 FIG. is a flowchart of the steps of the digital method for publishing paper textbooks based on artificial intelligence according to an embodiment of the present invention; Figure 2 FIG. is a flowchart of the steps of obtaining the edge complexity of each key area of a single subject type paper textbook according to an embodiment of the present invention; Figure 3 FIG. is a flowchart of the steps of determining the characteristic influence factors of each key area of paper textbooks of different subject types according to an embodiment of the present invention; Figure 4 FIG. is a logical determination diagram for dividing the abnormal tendency labels of paper textbooks according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] In order to make the objectives and advantages of the present invention clearer, the present invention will be further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0024] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present invention and do not limit the protection scope of the present invention.

[0025] Please refer to Figure 1 shown, which is a flowchart of the steps of the digital publishing of paper textbooks based on artificial intelligence according to an embodiment of the present invention. The present invention provides a digital method for publishing paper textbooks based on artificial intelligence, including: Step S1, extracting sample images of different key areas of paper textbooks of different subject types to determine the characteristic influence factors of each key area of paper textbooks of different subject types; Step S2: Calculate the sensitive feature characterization parameters of each type of paper-based textbook according to the characteristic influence factors of each key area of the paper-based textbooks of different subject types, so as to determine the sensitive feature sorting sequence of each key area of a single type of paper-based textbook; Step S3: Obtain the sample image of the paper-based textbook uploaded by the user terminal and the corresponding textbook subject type, and obtain the sensitive feature characterization parameters of each key area of the paper-based textbook of the corresponding subject type according to the actual textbook subject type, so as to divide the abnormal tendency label of the paper-based textbook. Analyze the actual image of the paper-based textbook according to the judgment result, including, If it is the first abnormal tendency label, obtain the sensitive feature sorting sequence of each key area of the paper-based textbook of the corresponding subject type, determine the comparison order of the local sample images of the key areas of the paper-based textbook to be published according to the sensitive feature sorting sequence, and sequentially call the key areas of the paper-based textbook of the corresponding subject type and the corresponding key areas of the actual paper-based textbook for comparison according to the comparison order to determine the similarity, so as to determine whether there is an abnormality; If it is the second abnormal tendency label, obtain the complete sample image of the paper-based textbook and compare it with the sample image of the corresponding complete paper-based textbook to determine the similarity, so as to determine whether there is an abnormality.

[0026] Specifically, the subject types of paper-based textbooks are generally divided into Chinese, mathematics, biology, politics, physics, chemistry, biology, history according to subjects, and can also be divided into grade one, grade two, grade three... grade n (n is a positive integer), or other forms can also be used. In the preferred implementation, the subject division method is adopted, which will not be elaborated here.

[0027] Specifically, the division of the key areas of paper-based textbooks is not limited. In the implementation, it can be divided into formula areas, curve graph areas, table areas, illustration areas, or other forms can also be used, which will not be elaborated here.

[0028] Specifically, the present invention establishes a database by extracting sample image information of different key regions of paper textbooks of various subject types. By determining the characteristic influencing factors of each key region of paper textbooks of different subject types, the sensitive characteristic representation parameters of each key region of paper textbooks of various subject types can be calculated, thereby quickly determining the sensitive characteristic sorting sequence of each key region of paper textbooks of various subject types, improving the review efficiency before the publication of paper textbooks. At the same time, by obtaining the actual images of the paper textbooks uploaded by the user terminal and the sample images of the corresponding key regions, the significant sensitive characteristic representation parameters of each key region of the uploaded paper textbooks can be accurately obtained, improving the review and publication accuracy of paper textbooks, and overcoming the problems in the prior art that the different sensitive characteristics of different key regions of paper textbooks of different subject types lead to low determination efficiency and low accuracy in the review and publication of corresponding subject paper textbooks. By determining the comparison order for the actual images of each key region of the paper textbooks, and sequentially calling the actual images of the corresponding key regions and the sample images of the corresponding key regions for comparison according to the comparison order, it can accurately determine whether there are abnormalities in the paper textbooks uploaded by the user terminal and determine whether to enable publication.

[0029] Specifically, there is no limitation on the determination method of similarity. The corresponding data information processing algorithms or models can be imported into the logic component to implement the corresponding functions. There is no specific limitation on the data information processing algorithms. For example, the method of calculating the similarity index can be used, or a pre-trained deep learning model can be used to extract the feature representation of the data information, and then the cosine similarity between the features can be calculated as the similarity. Of course, other methods can also be used, which will not be elaborated here.

[0030] Please refer to Figure 2 as shown, which is the flowchart of the steps for the present invention embodiment to determine the characteristic influencing factors of each key region of paper textbooks of different subject types. In step S1, the process of determining the characteristic influencing factors of each key region of paper textbooks of different subject types includes comparing the sample images of each key region of a single subject type with the sample images of the corresponding key regions of paper textbooks of other subject types respectively; solving the color saturation and obtaining the edge complexity of each key region of the paper textbooks of a single subject type.

[0031] Specifically, there is no limitation on the color saturation. The OpenCV or Pillow library in Python can be used to calculate the color saturation between the sample images of each key region and the sample images of the corresponding key regions of other subjects, or other forms can also be used, which will not be elaborated here.

[0032] Please refer to Figure 3As shown, it is a flowchart of the steps for obtaining the edge complexity of each key area of a single-subject type paper textbook in an embodiment of the present invention. In step S1, the process of obtaining the edge complexity of each key area of a single-subject type paper textbook includes, Calibrate the edge contour lines of each key area of a single-subject type paper textbook and extract the edge texture density; Calculate the coincidence degree between the edge contour line in the paper textbook and the reference contour line, and take the ratio of the coincidence degree to the reference contour coincidence threshold as the first edge complexity influencing factor; Calculate the ratio of the edge texture density in the paper textbook to the reference texture density threshold as the second edge complexity influencing factor; Determine the sum of the first edge complexity influencing factor and the second edge complexity influencing factor as the edge complexity; Among them, the coincidence degree is the ratio of the total area of the part where the edge contour in the paper textbook coincides with the reference contour to the area of the reference contour.

[0033] Specifically, the reference contour coincidence threshold and the reference texture density threshold are obtained by presetting. Among them, the product of the average value of the ratio of the total area of the coincidence between the edge contour and the basic contour within three months of the historical cycle of this system to the area of the reference contour and the accuracy coefficient is used as the reference contour coincidence threshold, and the accuracy coefficient is in the interval [0.92, 0.98]. The product of the average value of the edge texture density within three months of the historical cycle of this system and the deviation coefficient is used as the reference texture density threshold, and the deviation coefficient is in the interval [0.95, 0.99].

[0034] The present invention provides an important feature dimension for whether there is an abnormality in the publication review of paper textbooks by obtaining the edge contour coincidence degree and the edge texture density of each key area of a single-subject type paper textbook. The edge complexity reflects the structural characteristics of each key area of the paper textbook. The attribute factors of different subject types may cause the edge complexity of different subject types of paper textbooks to be different. Combining the edge contour coincidence degree and the edge texture density to characterize the influence of the complex characteristics of the key area.

[0035] Specifically, the process of calculating the sensitive feature characterization parameters of a single-subject type paper textbook includes, Obtain the feature influence factors of a single-subject type paper textbook, including the average color saturation and the edge complexity; Determine the ratio of a predetermined color saturation threshold to the average color saturation as the first feature influence factor; Determine the ratio of the edge complexity to a predetermined edge complexity threshold as the second feature influence factor; Determine the sum of the first feature influence factor and the second feature influence factor as the sensitive feature characterization parameter.

[0036] In implementation, the predetermined color saturation threshold is set in advance. Specifically, the average color saturation of different key areas of paper textbooks of each subject type is determined in advance, and the color saturation threshold is set as the product of the average color saturation and the precision coefficient, where the precision coefficient is selected within the range of [0.90, 0.95].

[0037] In implementation, the predetermined edge complexity threshold is set in advance. Specifically, the edge complexity of different key areas of paper textbooks of each subject type is determined in advance, the average edge complexity is calculated, and the edge complexity threshold is set as the product of the average edge complexity and the complexity offset coefficient, where the complexity offset coefficient is between the range of [1.10, 1.15].

[0038] In the present invention, by comparing the sample images of different key areas of a single-subject paper textbook with the image information of the corresponding key areas of paper textbooks of other subjects, the differences in the publication review of paper textbooks in different key areas are comprehensively considered. In actual situations, there are obvious sensitive features in some key areas of paper textbooks of some subjects, and some sensitive features in the key areas of paper textbooks of some subject types are more similar to the corresponding key areas of other subject types. Therefore, the sensitive feature representativeness of the sample images of each key area is different. In some cases, whether there is an abnormality can be identified based on the sample images of the key areas with obvious sensitive feature representativeness. Therefore, the present invention considers determining the sensitive feature characterization parameters of each key area, providing data support for the method of selecting sensitive feature characterization parameters when analyzing the actual images of paper textbooks uploaded for subsequent review and publication, and then adaptively selecting for actual review and publication analysis, reducing the review processing volume and improving the efficiency of paper textbook publication review on the premise of ensuring accuracy.

[0039] Specifically, the process of determining the sensitive feature ranking sequence of each key area of a single-subject type paper textbook includes Determine the sensitive feature characterization parameter corresponding to each key area of a single-subject type paper textbook; Arrange the sensitive feature characterization parameters in descending order to obtain the sensitive feature ranking sequence.

[0040] Please refer to Figure 4 As shown, it is a logical determination schematic diagram for dividing the abnormal tendency labels of paper textbooks in an embodiment of the present invention. The process of dividing the abnormal tendency labels of paper textbooks in the present invention includes Extract the significant sensitive feature characterization parameters of each key area of the paper textbook corresponding to the subject type; If the characterization parameter of the significant sensitive feature is greater than the threshold value of the characterization parameter of the predetermined sensitive feature, it is determined as the first abnormal tendency label; If the characterization parameter of the significant sensitive feature is less than or equal to the threshold value of the characterization parameter of the predetermined sensitive feature, it is determined as the second abnormal tendency label.

[0041] In implementation, the characterization parameter of the significant sensitive feature is the maximum characterization parameter of the sensitive feature, and the threshold value of the characterization parameter of the predetermined sensitive feature is in the interval [2.15, 2.45].

[0042] Specifically, determining the comparison order for the actual image of the paper-based textbook to be published according to the sorting sequence of the sensitive features, including, Sequentially determining the corresponding key areas according to the sorting sequence of the sensitive features, and comparing the key areas with the actual image; Among them, each sensitive feature characterization parameter corresponds one-to-one with the key area number.

[0043] In implementation, optionally, Each key area can be assigned a number. Taking the division of four key areas as an example, 1 - formula area, 2 - curve graph area, 3 - table area, 4 - illustration area; formula area, curve graph area, table area, illustration area For example, the characterization parameter of the formula area is 2.5, the characterization parameter of the curve graph area is 2.65, the characterization parameter of the table area is 2.55, and the characterization parameter of the illustration area is 2.45; The corresponding sorting sequence of the key areas is 2.65, 2.55, 2.5, 2.45; The corresponding sorting of the key areas is 2, 3, 1, 4.

[0044] Specifically, sequentially calling the key areas of the paper-based textbook of the corresponding subject type and comparing them with the corresponding key areas of the actual paper-based textbook according to the comparison order to determine the similarity, so as to determine whether there is an abnormality, including, If the similarity corresponding to the actual image of any key area is greater than the threshold value of the similarity of the key area, it is determined that there is no abnormality, and the publication of the paper-based textbook is enabled; If the similarity corresponding to the actual image of each key area is less than or equal to the threshold value of the similarity of the key area, it is determined that there is an abnormality, the publication of the paper-based textbook is not enabled, and the paper-based textbook is corrected and uploaded for analysis again.

[0045] In implementation, the similarity threshold for the key area is obtained by presetting. A number of sample image information of the same key area of paper teaching materials of the same subject type are obtained in advance, the average similarity between the sample image information is determined, and the ratio of the average similarity to the offset coefficient is determined as the similarity threshold for this key area. The local image offset coefficient is selected within the range of [0.85, 0.95].

[0046] Specifically, the complete sample image of the paper teaching material is obtained and compared with the sample image of the corresponding complete paper teaching material to determine the similarity to determine whether there is an abnormality, including The complete sample image of the paper teaching material uploaded by the user terminal is compared with the complete sample image of the paper teaching material of the corresponding subject type to determine the similarity to determine whether to enable the publication of the paper teaching material; If the similarity between the complete sample image of the current paper teaching material and the sample image of the corresponding complete paper teaching material is less than or equal to the preset similarity threshold of the complete sample image, it is determined that there is an abnormality, the publication of the paper teaching material is not enabled, and the paper teaching material is corrected and uploaded for analysis again; If the similarity between the complete sample image of the current paper teaching material and the sample image of the corresponding complete paper teaching material is greater than the preset similarity threshold of the complete sample image, it is determined that there is no abnormality, and the publication of the paper teaching material is enabled.

[0047] Specifically, the sample similarity threshold of the complete sample image of the paper teaching material is obtained by presetting. A number of complete sample image information of all key areas of the paper teaching materials of the same subject are obtained in advance, the average similarity between the complete sample image information is determined, and the ratio of the average similarity to the offset coefficient is determined as the similarity threshold of the key area. The complete image offset coefficient is selected within the range of [1.15, 1.25].

[0048] Specifically, the sample image of the paper teaching material uploaded by the user terminal shall include the complete sample image of the paper teaching material and the local sample images of each key area of the paper teaching material.

[0049] So far, the technical solution of the present invention has been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the protection scope of the present invention.

Claims

1. A digitalization method for paper textbook publishing based on artificial intelligence, characterized in that, Including: Extracting sample images of different key areas of paper textbooks of different subject types to determine the characteristic influence factors of each key area of paper textbooks of different subject types; Calculating the sensitive feature characterization parameters of each subject type of paper textbook based on the characteristic influence factors of each key area of paper textbooks of different subject types to determine the sensitive feature sorting sequence of each key area of a single subject type of paper textbook; Obtaining the sample image of the paper textbook uploaded by the user side and the corresponding textbook subject type, obtaining the sensitive feature characterization parameters of each key area of the paper textbook of the corresponding subject type according to the actual textbook subject type to divide the abnormal tendency label of the paper textbook, and analyzing the actual image of the paper textbook according to the determination result, Including, If it is the first abnormal tendency label, obtaining the sensitive feature sorting sequence of each key area of the paper textbook of the corresponding subject type, determining the comparison order of the local sample images of the key areas of the paper textbook to be published according to the sensitive feature sorting sequence, and sequentially calling the key areas of the paper textbook of the corresponding subject type and the corresponding key areas of the actual paper textbook for comparison according to the comparison order to determine the similarity to determine whether there is an abnormality; If it is the second abnormal tendency label, obtaining the complete sample image of the paper textbook and comparing it with the sample image of the corresponding complete paper textbook to determine the similarity to determine whether there is an abnormality.

2. The digital method for publishing paper textbooks based on artificial intelligence according to claim 1, characterized in that The process of determining the characteristic influence factors of each key area of paper textbooks of different subject types includes, Comparing the sample images of each key area of a single subject type with the local sample images of the paper textbooks of the corresponding key areas of other subject types respectively; Solving the color saturation and obtaining the edge complexity of each key area of a single subject type of paper textbook.

3. The digitalization method for paper textbook publication based on artificial intelligence according to claim 2, wherein Obtaining the edge complexity of each key area of a single subject type of paper textbook includes, Calibrating the edge contour lines of each key area of a single subject type of paper textbook and extracting the edge texture density; Calculating the coincidence degree between the edge contour line in the paper textbook and the reference contour line, and taking the ratio of the coincidence degree to the reference contour coincidence threshold as the first edge complexity influence factor; Calculating the ratio of the edge texture density in the paper textbook to the reference texture density threshold as the second edge complexity influence factor; Determining the sum of the first edge complexity influence factor and the second edge complexity influence factor as the edge complexity.

4. The digitalization method for paper textbook publishing based on artificial intelligence according to claim 1, wherein The process of calculating the sensitive feature characterization parameters of a single subject type of paper textbook includes, Obtaining the characteristic influence factors of a single subject type of paper textbook, including the average color saturation and the edge complexity; Determining the ratio of the predetermined color saturation threshold to the average color saturation as the first characteristic influence factor; Determining the ratio of the edge complexity to the predetermined edge complexity threshold as the second characteristic influence factor; Determining the sum of the first characteristic influence factor and the second characteristic influence factor as the sensitive feature characterization parameter.

5. The digitalization method for paper textbook publication based on artificial intelligence according to claim 1, wherein The process of determining the sensitive feature sorting sequence of each key area of a single subject type of paper textbook includes, Determining the sensitive feature characterization parameters corresponding to each key area of a single subject type of paper textbook; Arrange the sensitive feature characterization parameters from largest to smallest to obtain the sensitive feature sorting sequence.

6. The digital method for publishing paper textbooks based on artificial intelligence according to claim 1, wherein The process of dividing the abnormal tendency labels of paper textbooks includes extracting the significant sensitive feature characterization parameters of each key area of the paper textbook corresponding to the subject type; if the significant sensitive feature characterization parameter is greater than the predetermined sensitive feature characterization parameter threshold, it is determined as the first abnormal tendency label; if the significant sensitive feature characterization parameter is less than or equal to the predetermined sensitive feature characterization parameter threshold, it is determined as the second abnormal tendency label.

7. The digital method for publishing paper textbooks based on artificial intelligence according to claim 5, characterized in that Determine the comparison order for the actual image of the paper textbook to be published according to the sensitive feature sorting sequence, including sequentially determining the corresponding key areas according to the sensitive feature sorting sequence, and comparing the key areas with the actual image; wherein, each sensitive feature characterization parameter corresponds one-to-one with the key area serial number.

8. The digital method for publishing paper textbooks based on artificial intelligence according to claim 1, characterized in that Call the key areas of the paper textbook corresponding to the subject type and the corresponding key areas of the actual paper textbook for comparison in sequence according to the comparison order to determine the similarity to determine whether there is an abnormality, including if the similarity corresponding to the actual image of each key area is less than or equal to the predetermined key area similarity threshold, it is determined that there is an abnormality.

9. The digitalization method for paper-based textbook publication based on artificial intelligence according to claim 1, characterized in that Obtain the complete sample image of the paper textbook and compare it with the sample image of the corresponding complete paper textbook to determine the similarity to determine whether there is an abnormality, including obtaining the complete sample image of the paper textbook uploaded by the user terminal and comparing it with the complete sample image of the paper textbook corresponding to the subject type to determine the similarity; if the similarity between the complete sample image of the current paper textbook and the sample image of the corresponding complete paper textbook is less than or equal to the predetermined complete sample image similarity threshold, it is determined that there is an abnormality.

10. The digital method for publishing paper-based teaching materials based on artificial intelligence according to claim 1, characterized in that, The sample image of the paper textbook uploaded by the user terminal needs to include the complete sample image of the paper textbook and the local sample images of each key area of the paper textbook.

Images