A method and system for automatic quality inspection of digital archives based on image recognition

Through image recognition technology and quality inspection classification model, the quality problems in digital archives are automatically analyzed and marked, and the inefficiency and inconsistency of quality inspection caused by manual intervention in the existing technology are solved, and fast and accurate automated quality inspection and Consistency quality inspection report is generated.

CN117894027BActive Publication Date: 2024-11-26ZHEJIANG HUANHUA TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410095071.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-23
Publication Date
2024-11-26
Estimated Expiration
2044-01-23

AI Technical Summary

Technical Problem

The existing digital archive automatic quality inspection methods require manual intervention when facing complex archives, resulting in inefficiency and subjectivity and inconsistency of quality inspection results, affecting the credibility and consistency of archives, and lacking effective solutions.

Method used

Automatic quality inspection methods based on image recognition are adopted, through image acquisition, image recognition technology, character recognition and quality inspection classification models, digital archives that do not meet quality standards are automatically analyzed and marked, and quality inspection reports are generated to achieve automated quality inspection.

Benefits of technology

It realizes rapid and accurate identification of quality problems, reduces the time and cost of manual quality inspection, improves quality inspection efficiency, and generates consistent quality inspection reports through the quality inspection classification model to ensure timely understanding and improvement of archive quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117894027B_ABST
    Figure CN117894027B_ABST
Patent Text Reader

Abstract

The present invention discloses an automatic quality inspection method and system for digital archives based on image recognition, and relates to the technical field of digital archive quality inspection. The method comprises the following steps: acquiring image data of a digital archive to be inspected through an image acquisition device; analyzing the digital archive image data using image recognition technology, and marking digital archives that do not meet the quality standards according to preset quality standards; analyzing digital archives that do not meet the quality standards, and using a pre-built digital archive quality inspection classification model for classification and judgment; and forming a quality inspection report according to the judgment result of the digital archive quality inspection classification model. The present invention can quickly and accurately identify quality problems in digital archives through image recognition technology, which helps to timely discover potential quality defects, realize automated quality inspection of digital archives to be inspected, reduce the time and labor cost of manual quality inspection, and improve the efficiency of quality inspection.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of digital archive quality inspection, and in particular to an automatic quality inspection method and system for digital archives based on image recognition. Background Art

[0002] Digital archives are a new form of archival information that has emerged with the continuous development of computer technology, scanning technology, OCR technology, digital photography technology (including audio and video recording), database technology, multimedia technology and storage technology. This transformation transforms archival resources of various carriers into digital form, and through network connection, uses computer systems for efficient management. The characteristics of digital archives are that they are stored in digital form, have network connection and multimedia characteristics, form an orderly structured archival information library, and realize resource sharing and timely utilization.

[0003] With the widespread use of digital archives, it is essential to verify and inspect their quality. This process aims to ensure the accuracy and security of digital archives and prevent information distortion. In the digitization process, the use of computer vision technology, image recognition technology and other automated tools can effectively detect and repair potential errors and improve the credibility of digital archives.

[0004] In common automatic quality inspection of digitized archives, when faced with some complex archives, manual intervention may be required to solve problems that automatic quality inspection cannot handle. However, for large-scale collections of digital archives, relying on manual processing may be expensive and time-consuming, resulting in delays in the entire quality inspection process, which is not conducive to completing the task in a timely manner and reduces the efficiency of automatic quality inspection. In addition, different people may have different understandings and ways of dealing with the same problem, resulting in subjectivity and inconsistency in the quality inspection results, affecting the consistency and credibility of the digital archives.

[0005] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0006] In view of the problems in the related art, the present invention proposes a method and system for automatic quality inspection of digital archives based on image recognition to overcome the above technical problems existing in the existing related art.

[0007] To this end, the specific technical solution adopted by the present invention is as follows:

[0008] According to one aspect of the present invention, a method for automatic quality inspection of digital archives based on image recognition is provided. The method for automatic quality inspection of digital archives based on image recognition comprises the following steps:

[0009] S1. Acquire image data of the digital file to be detected through an image acquisition device;

[0010] S2. Use image recognition technology to analyze digital archive image data and mark digital archives that do not meet the quality standards according to the preset quality standards;

[0011] S3, analyze the digital archives that do not meet the quality standards and use the pre-built digital archive quality inspection classification model to make classification judgments;

[0012] S4. Generate a quality inspection report based on the judgment results of the digital archive quality inspection classification model.

[0013] Optionally, acquiring the image data of the digital archive to be detected by an image acquisition device includes the following steps:

[0014] S21, preprocessing the image data of the digital archive to be detected;

[0015] S22, using digital archive character recognition technology on the pre-processed image data to extract text data in the digital archive;

[0016] S23. Check the text data according to the preset quality standards and mark the digital files that do not meet the quality standards.

[0017] Optionally, using a digital archive character recognition technology on the pre-processed image data to extract text data from the digital archive includes the following steps:

[0018] S221, using image processing technology to detect the pre-processed image data to obtain a text area image in the image data;

[0019] S222, using a digital segmentation algorithm to segment the text region image and extract character features in the text region image;

[0020] S223. Using the trained digital archive character recognition model, the extracted character features are recognized and converted into text data.

[0021] Optionally, segmenting the text region image using a digital segmentation algorithm and extracting character features from the text region image comprises the following steps:

[0022] S2221, graying the text area image;

[0023] S2222, performing binary processing on the text area image after grayscale processing according to a preset digital archive adaptive threshold;

[0024] S2223, using a digital segmentation algorithm to identify and mark connected areas in the text area image after binary processing, grouping adjacent pixels into a connected area, and performing border detection on the connected area;

[0025] S2224. Based on the border detection result, the text area image is segmented into individual character images and character features in the text area image.

[0026] Optionally, a digital segmentation algorithm is used to identify and mark connected areas in the text area image after binary processing, and adjacent pixels are grouped into a connected area. The calculation formula for border detection of the connected area is:

[0027]

[0028] In the formula, RD(P,G) represents the loss function of the digital archive at the coordinate (P,G);

[0029] p i Represents the i-th pixel of the connected area;

[0030] g i Represents the i-th pixel of the adjacent connected area;

[0031] represents partial derivative;

[0032] β represents a hyperparameter;

[0033] N represents the total number of input pixels;

[0034] length(p) represents the length of the border of the connected region.

[0035] Optionally, checking the text data according to a preset quality standard and marking the digital archives that do not meet the quality standard includes the following steps:

[0036] S231, checking the text data according to a preset quality standard, wherein the checking content includes text format, text content and keyword matching;

[0037] S232. Use digital archive inspection algorithms to determine whether text data meets quality standards;

[0038] S233. Mark text data that does not meet quality standards.

[0039] Optionally, analyzing the digital archives that do not meet the quality standards and using a pre-built digital archive quality inspection classification model for classification judgment includes the following steps:

[0040] S31, analyzing the digital files that do not meet the quality standards, extracting problem feature data, and converting the digital files that do not meet the quality standards into feature vectors;

[0041] S32, inputting the feature vector into a pre-built digital archive quality inspection classification model to obtain the classification probability of the digital archive by the digital archive quality inspection classification model;

[0042] S33. Determine the classification of digital archives that do not meet the quality standards based on the classification probability of the digital archive quality inspection classification model.

[0043] Optionally, analyzing the digitized archives that do not meet the quality standards, extracting problem feature data, and converting the digitized archives that do not meet the quality standards into feature vectors includes the following steps:

[0044] S311. Define classification problem features based on quality standards and the types of problems that need to be detected;

[0045] S312, extracting corresponding classification problem feature data from the digital archives that do not meet the quality standards according to the defined classification problem features;

[0046] S313, converting the classification problem feature data into a feature vector.

[0047] Optionally, judging the classification of digital archives that do not meet the quality standards according to the classification probability of the digital archive quality inspection classification model includes the following steps:

[0048] S331. Preset a threshold value according to the classification probability of the digital archive quality inspection classification model;

[0049] S332, if the classification probability of the digitized archive that does not meet the quality standard exceeds a preset threshold, then determining that the digitized archive belongs to this category;

[0050] S333. If the classification probability of the digital archives that do not meet the quality standards does not exceed the preset threshold, it is judged that it cannot be judged.

[0051] According to another aspect of the present invention, there is also provided a digital archive automatic quality inspection system based on image recognition, the digital archive automatic quality inspection system based on image recognition comprises an image data acquisition module, a problem marking module, a classification judgment module and a quality inspection report generation module;

[0052] An image data acquisition module, used to acquire image data of a digital file to be detected through an image acquisition device;

[0053] The problem marking module is used to analyze the digital archive image data using image recognition technology and mark the digital archives that do not meet the quality standards according to the preset quality standards;

[0054] The classification judgment module is used to analyze digital archives that do not meet the quality standards and use the pre-built digital archive quality inspection classification model for classification judgment;

[0055] The quality inspection report generation module is used to generate a quality inspection report based on the judgment results of the digital archive quality inspection classification model.

[0056] The beneficial effects of the present invention are:

[0057] 1. The present invention can quickly and accurately identify quality problems in digital files through image recognition technology, which helps to timely discover potential quality defects, realize automated quality inspection of digital files to be inspected, reduce the time and labor cost of manual quality inspection, and improve the efficiency of quality inspection.

[0058] 2. The present invention can classify and judge digital archives that do not meet quality standards through the pre-constructed digital archive quality inspection classification model, and generate corresponding quality inspection reports, which helps to timely understand the quality status of digital archives and take corresponding improvement measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0060] Figure 1 is a flow chart of a method for automatic quality inspection of digital archives based on image recognition according to an embodiment of the present invention;

[0061] Figure 2 The present invention is a block diagram of a digital archive automatic quality inspection system based on image recognition according to an embodiment of the present invention.

[0062] In the figure:

[0063] 1. Image data acquisition module; 2. Problem marking module; 3. Classification and judgment module; 4. Quality inspection report generation module. DETAILED DESCRIPTION

[0064] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in the field should be able to understand other possible implementations and advantages of the present invention. The components in the figures are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0065] According to an embodiment of the present invention, a method and system for automatic quality inspection of digital archives based on image recognition are provided.

[0066] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to an embodiment of the present invention, the automatic quality inspection method for digital archives based on image recognition comprises the following steps:

[0067] S1. Acquire image data of the digital file to be detected through an image acquisition device.

[0068] It should be noted that the digital files to be inspected should be placed within the range that the image acquisition device can capture. The parameters of the image acquisition device, including exposure time, focus, resolution, etc., should be adjusted as needed to ensure that clear and accurate image data is obtained. The acquired image data should be saved in a computer or storage device for subsequent digital file inspection and processing.

[0069] S2. Use image recognition technology to analyze digital archive image data and mark digital archives that do not meet the quality standards based on preset quality standards.

[0070] The process of acquiring the image data of the digital archive to be detected by the image acquisition device includes the following steps:

[0071] S21, pre-processing the image data of the digital file to be detected.

[0072] It should be noted that the image data is read from the digital archive to be inspected, and a denoising algorithm is used to perform denoising on the image data of the digital archive to be inspected, so as to eliminate unnecessary detail information such as noise points and artifacts in the image.

[0073] S22. Using digital archive character recognition technology on the pre-processed image data, extract text data from the digital archive.

[0074] The process of using digital archive character recognition technology to extract text data from the digital archive includes the following steps:

[0075] S221. Detect the pre-processed image data using image processing technology to obtain a text area image in the image data.

[0076] S222. Use a digital segmentation algorithm to segment the text region image and extract character features in the text region image.

[0077] The process of segmenting the text region image using a digital segmentation algorithm and extracting character features from the text region image includes the following steps:

[0078] S2221, grayscale the text area image.

[0079] S2222. According to a preset digitized archive adaptive threshold, binary processing is performed on the text area image after grayscale processing.

[0080] S2223. Use a digital segmentation algorithm to identify and mark connected areas in the text area image after binary processing, group adjacent pixels into a connected area, and perform border detection on the connected area.

[0081] Among them, the digital segmentation algorithm is used to identify and mark the connected areas in the text area image after binary processing, and the adjacent pixels are combined into a connected area. The calculation formula for the border detection of the connected area is:

[0082]

[0083] In the formula, RD(P,G) represents the loss function of the digital archive at the coordinate (P,G);

[0084] p i Represents the i-th pixel of the connected area;

[0085] g i Represents the i-th pixel of the adjacent connected area;

[0086] represents partial derivative;

[0087] β represents a hyperparameter;

[0088] N represents the total number of input pixels;

[0089] length(p) represents the length of the border of the connected region.

[0090] S2224. Based on the border detection result, the text area image is segmented into individual character images and character features in the text area image.

[0091] S223. Using the trained digital archive character recognition model, the extracted character features are recognized and converted into text data.

[0092] S23. Check the text data according to the preset quality standards and mark the digital files that do not meet the quality standards.

[0093] Among them, checking the text data according to the preset quality standards and marking the digital files that do not meet the quality standards include the following steps:

[0094] S231. Check the text data according to a preset quality standard, wherein the check content includes text format, text content and keyword matching.

[0095] S232. Use digital archive inspection algorithms to determine whether text data meets quality standards.

[0096] S233. Mark text data that does not meet quality standards.

[0097] S3. Analyze the digital archives that do not meet the quality standards and use the pre-built digital archive quality inspection classification model to make classification judgments.

[0098] It should be noted that the digital archive quality inspection classification model is a decision tree. Digital archives of different quality standards are collected to form a sample data set, and the sample data set is preprocessed; relevant feature data is extracted from the preprocessed digital archive sample data set; the sample data set is divided into a training set and a test set. According to the characteristics of the problem and the feature data, the training set is used to train the decision tree model, and the test set is used to evaluate the decision tree model.

[0099] Among them, analyzing the digital archives that do not meet the quality standards and using the pre-built digital archive quality inspection classification model for classification and judgment includes the following steps:

[0100] S31. Analyze the digital archives that do not meet the quality standards, extract problem feature data, and convert the digital archives that do not meet the quality standards into feature vectors.

[0101] The steps of analyzing the digital archives that do not meet the quality standards, extracting problem feature data, and converting the digital archives that do not meet the quality standards into feature vectors include the following steps:

[0102] S311. Define classification problem features based on quality standards and the types of problems that need to be detected.

[0103] S312. According to the defined classification problem characteristics, corresponding classification problem characteristic data are extracted from the digital archives that do not meet the quality standards.

[0104] S313, converting the classification problem feature data into a feature vector.

[0105] S32. Input the feature vector into a pre-built digital archive quality inspection classification model to obtain the classification probability of the digital archive by the digital archive quality inspection classification model.

[0106] It should be noted that the feature vectors are normalized so that they are within the same scale range. The processed feature vectors are used as input into the digital archive quality inspection and classification model. The feature vectors are inferred to obtain the classification probability of the digital archives, and the classification probability of the digital archives is extracted from the output of the digital archive quality inspection and classification model.

[0107] S33. Determine the classification of digital archives that do not meet the quality standards based on the classification probability of the digital archive quality inspection classification model.

[0108] Among them, judging the classification of digital archives that do not meet the quality standards according to the classification probability of the digital archive quality inspection classification model includes the following steps:

[0109] S331. Preset a threshold value based on the classification probability of the digital archive quality inspection classification model.

[0110] S332. If the classification probability of the digital archive that does not meet the quality standard exceeds a preset threshold, it is determined that the digital archive belongs to this category.

[0111] S333. If the classification probability of the digital archives that do not meet the quality standards does not exceed the preset threshold, it is judged that it cannot be judged.

[0112] S4. Generate a quality inspection report based on the judgment results of the digital archive quality inspection classification model.

[0113] It should be noted that, based on the output of the digital archive quality inspection classification model, the model's classification results for the digital archives are judged; based on the judgment results, the content that the quality inspection report needs to include is determined; and the specific content of the quality inspection report is recorded and written.

[0114] like Figure 2 As shown, according to another aspect of the present invention, there is also provided a digital archive automatic quality inspection system based on image recognition, the digital archive automatic quality inspection system based on image recognition comprises an image data acquisition module 1, a problem marking module 2, a classification judgment module 3 and a quality inspection report generation module 4;

[0115] An image data acquisition module 1 is used to acquire image data of a digital file to be detected through an image acquisition device;

[0116] Problem marking module 2 is used to analyze the digital archive image data using image recognition technology and mark the digital archives that do not meet the quality standards according to the preset quality standards;

[0117] Classification and judgment module 3 is used to analyze digital archives that do not meet the quality standards and use the pre-built digital archive quality inspection classification model for classification and judgment;

[0118] The quality inspection report generation module 4 is used to generate a quality inspection report based on the judgment result of the digital archive quality inspection classification model.

[0119] In summary, with the help of the above technical solution of the present invention, through image recognition technology, quality problems in digital archives can be quickly and accurately identified, which helps to timely discover potential quality defects, realize automated quality inspection of digital archives to be inspected, reduce the time and labor costs of manual quality inspection, and improve the efficiency of quality inspection. Through the pre-constructed digital archive quality inspection classification model, digital archives that do not meet the quality standards can be classified and judged, and corresponding quality inspection reports can be generated, which helps to timely understand the quality status of digital archives and take corresponding improvement measures.

[0120] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An automatic quality inspection method for digital archives based on image recognition. It is characterized in that The automatic quality inspection method for digital archives based on image recognition includes the following steps: S1. Acquire image data of the digital file to be detected through an image acquisition device; S2. Use image recognition technology to analyze digital archive image data and mark digital archives that do not meet the quality standards according to the preset quality standards; S3, analyze the digital archives that do not meet the quality standards and use the pre-built digital archive quality inspection classification model to make classification judgments; S4. Form a quality inspection report based on the judgment results of the digital archive quality inspection classification model; The method of acquiring the image data of the digital archive to be detected by the image acquisition device comprises the following steps: S21, preprocessing the image data of the digital archive to be detected; S22, using digital archive character recognition technology on the pre-processed image data to extract text data in the digital archive; S23, checking the text data according to the preset quality standards, and marking the digital files that do not meet the quality standards; The method of extracting text data from the digitized archives by using the digitized archives character recognition technology on the pre-processed image data comprises the following steps: S221, using image processing technology to detect the pre-processed image data to obtain a text area image in the image data; S222, using a digital segmentation algorithm to segment the text region image and extract character features in the text region image; S223, using the trained digital archive character recognition model to recognize the extracted character features and convert them into text data; The method of segmenting the text area image using a digital segmentation algorithm and extracting character features in the text area image comprises the following steps: S2221, graying the text area image; S2222, performing binary processing on the text area image after grayscale processing according to a preset digital archive adaptive threshold; S2223, using a digital segmentation algorithm to identify and mark connected areas in the text area image after binary processing, grouping adjacent pixels into a connected area, and performing border detection on the connected area; S2224, based on the frame detection result, segmenting the text area image into individual character images and character features in the text area image; The digital segmentation algorithm is used to identify and mark the connected areas in the text area image after binary processing, and the adjacent pixels are combined into a connected area. The calculation formula for border detection of the connected area is: In the formula, RD(P,G) represents the loss function of the digital archive at the coordinate (P,G); p i Represents the i-th pixel of the connected area; g i Represents the i-th pixel of the adjacent connected area; represents partial derivative; β represents a hyperparameter; N represents the total number of input pixels; length(p) represents the border length of the connected region.

2. According to claim 1, a method for automatic quality inspection of digital archives based on image recognition, It is characterized in that The checking of text data according to the preset quality standards and marking of digital files that do not meet the quality standards include the following steps: S231, checking the text data according to a preset quality standard, wherein the checking content includes text format, text content and keyword matching; S232. Use digital archive inspection algorithms to determine whether text data meets quality standards; S233. Mark text data that does not meet quality standards.

3. According to claim 1, a method for automatic quality inspection of digital archives based on image recognition, It is characterized in that The analysis of the digital archives that do not meet the quality standards and the classification judgment using the pre-built digital archive quality inspection classification model include the following steps: S31, analyzing the digital files that do not meet the quality standards, extracting problem feature data, and converting the digital files that do not meet the quality standards into feature vectors; S32, inputting the feature vector into a pre-built digital archive quality inspection classification model to obtain the classification probability of the digital archive by the digital archive quality inspection classification model; S33. Determine the classification of digital archives that do not meet the quality standards based on the classification probability of the digital archive quality inspection classification model.

4. According to claim 3, a method for automatic quality inspection of digital archives based on image recognition, It is characterized in that The analyzing of the digital archives that do not meet the quality standards, extracting the problem feature data, and converting the digital archives that do not meet the quality standards into feature vectors comprises the following steps: S311. Define classification problem features based on quality standards and the types of problems that need to be detected; S312, extracting corresponding classification problem feature data from the digital archives that do not meet the quality standards according to the defined classification problem features; S313, converting the classification problem feature data into a feature vector.

5. According to claim 3, a method for automatic quality inspection of digital archives based on image recognition, It is characterized in that The method of determining the classification of digital archives that do not meet the quality standards according to the classification probability of the digital archive quality inspection classification model comprises the following steps: S331. Preset a threshold value according to the classification probability of the digital archive quality inspection classification model; S332, if the classification probability of the digitized archive that does not meet the quality standard exceeds a preset threshold, then determining that the digitized archive belongs to this category; S333. If the classification probability of the digital archives that do not meet the quality standards does not exceed the preset threshold, it is judged that it cannot be judged.

6. An automatic quality inspection system for digital archives based on image recognition, used to implement an automatic quality inspection method for digital archives based on image recognition as claimed in any one of claims 1 to 5, It is characterized in that The digital archive automatic quality inspection system based on image recognition includes an image data acquisition module, a problem marking module, a classification judgment module and a quality inspection report generation module; The image data acquisition module is used to acquire the image data of the digital archive to be detected through an image acquisition device; The problem marking module is used to analyze the digital archive image data using image recognition technology and mark the digital archives that do not meet the quality standards according to the preset quality standards; The classification judgment module is used to analyze digital archives that do not meet the quality standards and use the pre-built digital archive quality inspection classification model for classification judgment; The quality inspection report generation module is used to generate a quality inspection report based on the judgment result of the digital archive quality inspection classification model.