Method and system for industrial video data acquisition and identification

Abstract

The invention belongs to the technical field of video data processing, and provides a method and system for industrial video data acquisition and recognition, and the method comprises the steps: obtaining a sequential static data image set, recognizing an original data set, obtaining an industrial production data situation feature set, carrying out the correlation evaluation, correcting abnormal data, and storing the data. According to the method, the industrial production data situation feature set is constructed, so that an identification basis is provided for identifying abnormal data conditions, a correction basis is provided for correcting the abnormal data, and the extraction efficiency and accuracy of the industrial video data are improved; through sequential static data image collection, data acquisition and communication isolation of an industrial production system are realized, and the security of the system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of video data processing, and in particular to a method and system for industrial video data acquisition and recognition. Background Art

[0002] With the continuous development of economy and science and technology, the production process has become increasingly complex, and various process parameters in the production process have increased day by day. The internal relationships are complex. For industrial parameters in the manufacturing process, manual recording is often used, and there are also cases where third-party reading software is implanted in the industrial production system for acquisition, and the recorded data is summarized for future search.

[0003] However, the direct connection of third-party software or devices to the industrial production system further threatens the security of the industrial production system. Relying on manual recording of production situations is very prone to omissions and errors. At the same time, the information transmission cycle of the recorded information is long, and the reliability and timeliness of the data cannot be guaranteed. In summary, there are technical problems of low efficiency and accuracy in industrial video data acquisition and recognition, as well as security problems in data acquisition and recognition in the prior art. Summary of the Invention

[0004] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a method and system for industrial video data acquisition and recognition, so as to solve the technical problems of low efficiency and accuracy in industrial video data acquisition and recognition, as well as the security problems in data acquisition and recognition in the prior art.

[0005] To achieve the above purpose, the present invention provides the following solutions:

[0006] A method for industrial video data acquisition and recognition, comprising:

[0007] Performing industrial video acquisition on a target factory through a video acquisition device to obtain a set of sequential static data images;

[0008] Performing recognition on the set of sequential static data images through an image recognition device to obtain a set of original data;

[0009] Taking industrial production information as a search tag, performing big data analysis on all data feature situations in a pre-recorded video to obtain a set of industrial production data situation feature;

[0010] Performing correlation evaluation on the set of industrial production data situation feature and the set of original data to obtain abnormal data and normal data, and correcting the abnormal data to obtain a set of corrected data;

[0011] Storing the normal data and the set of corrected data in a database.

[0012] Preferably, industrial video of the target factory is collected by a video acquisition device to obtain a set of sequential static data images, including:

[0013] The video acquisition device is connected in series to the video signal loop of the target factory to obtain industrial video;

[0014] The industrial video is sent to the display screen of the target factory through the video output port of the video acquisition device;

[0015] The video acquisition device is used to continuously copy the video images transmitted to the display screen and perform digital processing on the video images to obtain industrial static data images;

[0016] A plane rectangular coordinate system is established, and the industrial static data images are placed in the first quadrant of the plane rectangular coordinate system;

[0017] The central coordinates and sizes of each individual data image in the industrial static data image are determined through the plane rectangular coordinate system;

[0018] The individual data images are named according to the central coordinates to obtain the name identifiers of each individual data image;

[0019] The industrial static data image is intercepted according to the size to obtain each individual data image;

[0020] By setting the interception frequency of each individual data image, the time identifier of each individual data image is obtained;

[0021] The name identifiers are classified to obtain a set of similar static data images;

[0022] The individual data images in the set of similar static data images are sorted sequentially according to the time identifier to obtain a set of sequential static data images.

[0023] Preferably, the sequential static data image set is recognized by an image recognition device to obtain an original data set, including:

[0024] The sequential static data image set is recognized to obtain recognition data;

[0025] The data name and data time of the recognition data are set according to the name identifier and time identifier of the sequential static data image set;

[0026] The recognition data with the same data name is sorted sequentially according to the data time to obtain the original data set.

[0027] Preferably, the video signal loop is HDMI.

[0028] Preferably, the industrial production data situation feature set includes: the upper limit of single-item data and the accuracy of single-item data.

[0029] Preferably, by performing an association evaluation on the industrial production data situation feature set and the original data set, abnormal data and normal data are obtained, and the abnormal data is corrected to obtain a corrected data set, including:

[0030] Construct a data recognition model;

[0031] Input the original data set and the industrial production data situation feature set into the data recognition model for threshold and accuracy comparison to obtain a recognition result;

[0032] Determine the data in the original data set whose recognition result satisfies any item in the judgment criteria as the normal data;

[0033] Determine the data in the original data set whose recognition result does not satisfy any item in the judgment criteria as the abnormal data; the judgment criteria are the upper limit and accuracy in the industrial production data situation feature set;

[0034] Calculate the deviation coefficient of the abnormal data according to the industrial production data situation feature set;

[0035] Use the deviation coefficient to calculate and correct the abnormal data to obtain the corrected data set.

[0036] Preferably, the absolute value of the deviation coefficient is the difference between the number of digits of the abnormal data and the accuracy of the single-item data.

[0037] Preferably, a system for industrial video data acquisition and recognition includes:

[0038] A video acquisition module, configured to perform industrial video acquisition on a target factory through a video acquisition device to obtain a set of sequential static data images;

[0039] An image recognition module, configured to recognize the set of sequential static data images through an image recognition device to obtain an original data set;

[0040] A data situation feature module, configured to perform big data analysis on all data feature situations in a pre-recorded video using industrial production information as a search tag to obtain an industrial production data situation feature set;

[0041] A data correction module, configured to perform an association evaluation on the industrial production data situation feature set and the original data set to obtain abnormal data and normal data, and correct the abnormal data to obtain a corrected data set;

[0042] A data transmission module, configured to store the normal data and the set of corrected data in a database.

[0043] The present invention discloses the following technical effects:

[0044] The present invention provides a method and a system for industrial video data acquisition and recognition. By constructing a set of industrial production data situation characteristics, the technical problems of low efficiency and accuracy in industrial video data acquisition and recognition in the prior art are solved, and a recognition basis for identifying abnormal data situations and a correction basis for correcting abnormal data are realized; through the sequential static data image set, the security problems of data acquisition and recognition are solved, and communication isolation between data acquisition and the industrial production system is realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0046] Figure 1 It is a schematic flowchart of the industrial video data acquisition and recognition provided by the embodiment of the present invention;

[0047] Figure 2 It is a schematic structural diagram of the system for industrial video data acquisition and recognition provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0049] The purpose of the present invention is to provide a method and a system for industrial video data acquisition and recognition, and solve the technical problems of low efficiency and accuracy in industrial video data acquisition and recognition in the prior art and the security problems of data acquisition and recognition.

[0050] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the drawings and specific embodiments.

[0051] Figure 1The flowchart diagram provided by the embodiments of the present invention for industrial video data acquisition and recognition is as follows. Figure 1 As shown, the present invention provides a method for industrial video data acquisition and recognition, including:

[0052] Step 100: Perform industrial video acquisition on the target factory through a video acquisition device to obtain a set of sequential static data images;

[0053] Step 200: Recognize the set of sequential static data images through an image recognition device to obtain an original data set;

[0054] Step 300: Use industrial production information as a search tag to perform big data analysis on all data feature situations in the pre-recorded video to obtain an industrial production data situation feature set;

[0055] Step 400: Through the correlation evaluation of the industrial production data situation feature set and the original data set, obtain abnormal data and normal data, and correct the abnormal data to obtain a corrected data set;

[0056] Step 500: Store the normal data and the corrected data set in a database.

[0057] Further, performing industrial video acquisition on the target factory through a video acquisition device to obtain a set of sequential static data images includes:

[0058] Connect the video acquisition device into the video signal loop of the target factory to obtain industrial video;

[0059] Use the video output port of the video acquisition device to send the industrial video to the display screen of the target factory;

[0060] Use the video acquisition device to continuously copy the video screen transmitted to the display screen and perform digital processing on the video screen to obtain industrial static data images;

[0061] Establish a plane rectangular coordinate system and place the industrial static data images in the first quadrant of the plane rectangular coordinate system;

[0062] Determine the center coordinates and sizes of each individual data image in the industrial static data images through the plane rectangular coordinate system;

[0063] Name each individual data image according to the center coordinates to obtain the name identifier of each individual data image;

[0064] Perform image cropping on the industrial static data images according to the sizes to obtain each individual data image;

[0065] Obtain the time identifier of each individual data image by setting the cropping frequency of each individual data image.

[0066] Classify the name identifiers to obtain a set of static data images of the same category;

[0067] Sort the individual data images in the set of static data images of the same category according to the time identifier to obtain a set of sequential static data images.

[0068] Specifically, use an image recognition device to recognize the set of sequential static data images to obtain an original data set, including:

[0069] Recognize the set of sequential static data images to obtain recognition data;

[0070] Set the data name and data time for the recognition data according to the name identifier and time identifier of the set of sequential static data images;

[0071] Sort the recognition data with the same data name according to the data time to obtain the original data set.

[0072] Optionally, the video signal loop is HDMI.

[0073] Preferably, the industrial production data situation feature set includes: the upper limit of individual data and the accuracy of individual data.

[0074] Specifically, through the correlation evaluation of the industrial production data situation feature set and the original data set, obtain abnormal data and normal data, and correct the abnormal data to obtain a corrected data set, including:

[0075] Construct a data recognition model;

[0076] Input the original data set and the industrial production data situation feature set into the data recognition model for threshold and accuracy comparison to obtain a recognition result;

[0077] Determine the data in the original data set whose recognition result satisfies any one of the judgment criteria as the normal data;

[0078] Determine the data in the original data set whose recognition result does not satisfy any one of the judgment criteria as the abnormal data; the judgment criteria are the upper limit and accuracy in the industrial production data situation feature set;

[0079] Calculate the deviation coefficient of the abnormal data according to the industrial production data situation feature set;

[0080] Use the deviation coefficient to calculate and correct the abnormal data to obtain the corrected data set.

[0081] Specifically, the absolute value of the deviation coefficient is the difference between the number of digits of the abnormal data and the precision of the individual data.

[0082] Reference Figure 2 , a system for industrial video data acquisition and recognition, comprising:

[0083] A video acquisition module, configured to perform industrial video acquisition on a target factory through a video acquisition device to obtain a set of sequential static data images;

[0084] An image recognition module, configured to recognize the set of sequential static data images through an image recognition device to obtain an original data set;

[0085] A data situation feature module, configured to perform big data analysis on all data feature situations in a pre-recorded video with industrial production information as a search tag to obtain an industrial production data situation feature set;

[0086] A data correction module, configured to obtain abnormal data and normal data through an associated evaluation of the industrial production data situation feature set and the original data set, and correct the abnormal data to obtain a corrected data set;

[0087] A data transmission module, configured to store the normal data and the corrected data set in a database.

[0088] Specifically, the method provided in this embodiment includes:

[0089] Step S10: Perform industrial video acquisition of a target factory through a video acquisition device to obtain industrial static data images;

[0090] Further, step S10 further includes:

[0091] Step S11: Connect the video signal of the video acquisition device to the industrial system of the target factory to obtain the industrial video of the target factory;

[0092] Specifically, the video acquisition device is connected in series to the video signal loop of the industrial system of the target factory. Preferably, the video acquisition device can complete the acquisition of the industrial production video picture through video communication via an HDMI video signal loop.

[0093] Step S12: Based on the video forwarding function of the video acquisition device, use the video output port of the video acquisition device to send the obtained industrial real-time video of the target factory to the display screen of the target factory;

[0094] Step S13: Based on the video-to-graphic conversion function of the video acquisition device, perform digital processing on the real-time copied and transmitted video picture within the video acquisition device to obtain industrial static data images.

[0095] Further, step S13 includes:

[0096] Step S131: By establishing a plane rectangular coordinate system, place the industrial static image in the first quadrant of the plane rectangular coordinate system;

[0097] Step S132: Through the plane rectangular coordinate system, determine the central coordinates of each single-item data image in the industrial static image, and obtain the central coordinates and sizes of each single-item data image in the industrial static image;

[0098] Step S133: Name each single-item data image in the industrial static image according to the central coordinates of the single-item data image, and obtain the name identifiers of each single-item data image;

[0099] Step S134: Intercept the image of each single-item data image in the industrial static image according to the size of the single-item data image in the industrial static image, and obtain each single-item data image;

[0100] Step S135: By setting the interception frequency of each single-item data image in the industrial static image, obtain the time identifiers of each single-item data image in the industrial static image;

[0101] Step S136: Classify each single-item data image in the industrial static image based on the name identifier to obtain a set of industrial static data images with the same name;

[0102] Step S137: Sort the industrial static data images with the same name according to the time identifier to obtain a set of sequential industrial static data images.

[0103] Specifically, classify the data images according to different names based on the data image identifier to obtain a set of industrial static data images with the same name. Among them, the set of data images with the same name refers to the result of intercepting data images at the same screen coordinate position. Furthermore, sort the set of data images with the same name in chronological order according to the time identifier, so as to obtain a set of sequential industrial static data images.

[0104] Step S20: Identify the industrial static data image through an image recognition device to obtain the data result in the industrial static image;

[0105] Further, step S20 further includes:

[0106] Step S21: Based on the name identifier of the industrial static data image, assign the same data name to the data identified in the industrial static data image according to the name identifier of the original industrial static data image;

[0107] Step S22: Based on the time identifier of the industrial static data image, assign the same data time to the data identified in the industrial static data image according to the time identifier of the original industrial static data image;

[0108] Step S23: Sort the data in the same name according to the data time to obtain a data set.

[0109] Step S30: Construct a set of industrial production data situation characteristics through big data and industrial production information;

[0110] Specifically, the target factory is any factory that needs to collect and identify video data during the production and manufacturing process. Analyze the industrial production data based on industrial production information and big data technology to obtain a set of industrial production data situation characteristics (industrial production information is obtained by recording the industrial data of the target factory, including: the range of each individual industrial data of the target factory, the type of each individual industrial data, etc. Furthermore, using industrial production information as a search tag, collect all data feature situations in the industrial video to be identified based on big data technology to obtain a set of production data situation characteristics. Among them, the set of production data situation characteristics refers to the set obtained by summarizing the characteristics of each individual data in the industrial video to be identified in the target factory, including: the upper limit of each individual data, the precision of each individual data, etc. By establishing a set of industrial production data situation characteristics, it provides an identification basis for subsequent identification of abnormal data situations in the target factory, provides a correction basis for correcting abnormal data, and improves the extraction efficiency and accuracy of industrial video data.).

[0111] Step S40: Conduct an association evaluation between the set of industrial production data situation characteristics and the data results in the industrial static data image, correct abnormal data, and obtain a data set.

[0112] Step S41: Classify the set of industrial production data situation characteristics according to the data name to generate thresholds for each individual data. The data threshold includes the upper limit of the individual data;

[0113] Step S42: Classify the set of industrial production data situation characteristics according to the data name to generate the precision of each individual data. The data precision is the number of digits of the individual data;

[0114] Step S43: Construct an identification model for each individual data, and add the thresholds for each individual data and the precision of each individual data to the corresponding identification models for each individual data respectively;

[0115] Step S44: Classify and input the data set in the industrial static data image into the data identification model, and sequentially judge within the model: whether the data is consistent with the specified number of digits of the data identification model, and whether the data is less than or equal to the specified upper limit of the data model, to obtain an output result;

[0116] Step S45: If all are satisfied, directly store the data in the database;

[0117] Step S46: If any of them is not satisfied, the abnormal data is associated and evaluated with the industrial production data situation feature set. After correcting the abnormal data, it is then stored in the database.

[0118] Further, step S46 includes:

[0119] Step S461: Name the abnormal data according to the data name;

[0120] Step S462: Based on the name of the abnormal data, associate the corresponding industrial production data situation feature set, and calculate the deviation coefficient of the abnormal data;

[0121] Specifically, the industrial production data situation feature set corresponding to the abnormal data has a data upper limit and a number of digits. If the abnormal data is a i , the number of digits of the abnormal data is b, the data upper limit within the corresponding industrial production data situation feature set is a h , the number of digits is c, and the previous normal data under the abnormal data name is a i-1 , when a i > a h , a i > a i-1 , b > c, the deviation coefficient of the abnormal data: b - c; when a i < a h , a i < a i-1 , b < c, the deviation coefficient of the abnormal data: -(b - c);

[0122] Step S463: Based on the deviation coefficient of the data, calculate and correct the abnormal data to obtain the corrected result of the abnormal data, and store it in the database. For the abnormal data a i , when a i> a h , a i > a i-1 , b > c, the abnormal data correction: a i / 10 (b-c) ; when a i < a h , a i < a i-1 , b < c, the abnormal data correction: a i / 10 -(b-c) ;

[0123] Further, this embodiment also provides a system for industrial video data acquisition and recognition. Among them, the system includes:

[0124] A video acquisition module, which is used to acquire the industrial video of the target factory to obtain industrial static data images;

[0125] The image recognition module is used to recognize industrial static data images and obtain the data results in the industrial static images;

[0126] The data situation feature module is used to integrate big data and industrial production information to construct an industrial production data situation feature set;

[0127] The data correction module is used to conduct an association evaluation between the industrial production data situation feature set and the data results in the industrial static data images, correct abnormal data, and obtain a data set;

[0128] The data transmission module is used to store the data set after industrial video acquisition and recognition in a database.

[0129] Preferably, the system further includes:

[0130] The video acquisition unit is used to acquire the industrial system video of the target factory and obtain the industrial video of the target factory;

[0131] The view - image conversion unit is used to digitally process the acquired industrial video of the target factory to obtain industrial static data images.

[0132] Furthermore, the system further includes:

[0133] The name classification unit is used to classify the individual data images in the industrial static images with the same name based on the name identifier, and obtain a set of industrial static data images with the same name;

[0134] The time - sequence sorting unit is used to sort the industrial static data images with the same name in sequence based on the time identifier, and obtain a set of sequential industrial static data images;

[0135] The image - data translation unit is used to recognize the set of industrial static data images and obtain the data set in the industrial static data images.

[0136] Preferably, the system further includes:

[0137] The industrial production data situation feature set unit constructs an industrial production data situation feature set by integrating big data and industrial production information;

[0138] The abnormal analysis unit conducts data comparison between the industrial production data situation feature set and the data results recognized in the industrial static images to find abnormal data.

[0139] Furthermore, the system further includes:

[0140] The abnormal correction unit conducts association calculations based on the abnormal data and the industrial production data situation feature set to obtain the abnormal data correction result;

[0141] A data storage unit stores all data sets in a database.

[0142] The beneficial effects of the present invention are as follows:

[0143] By constructing a set of industrial production data situation characteristics, the present invention provides an identification basis for identifying abnormal data situations, provides a correction basis for correcting abnormal data, and improves the extraction efficiency and accuracy of industrial video data; through the sequential static data image set, the communication isolation between data collection and the industrial production system is realized, and the security of the system is improved.

[0144] In this specification, each embodiment is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. For the same or similar parts among the embodiments, reference can be made to each other. For the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and reference can be made to the description in the method part for the relevant parts.

[0145] In this article, specific examples are used to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A method for industrial video data collection and recognition, characterized in that: include: The target factory is subjected to industrial video acquisition through a video acquisition device to obtain a set of sequential static data images; Recognize the sequential static data image set by an image recognition device to obtain an original data set; Using industrial production information as the search tag, we conduct big data analysis on all data features in the pre-recorded video to obtain a set of industrial production data situation features; By performing correlation evaluation on the industrial production data situation feature set and the original data set, abnormal data and normal data are obtained, and the abnormal data is corrected to obtain a corrected data set; The normal data and the correction data are stored in a database.

2. A method for industrial video data acquisition and recognition according to claim 1, characterized in that: The industrial video of the target factory is collected by the video acquisition device to obtain a set of sequential static data images, including: Connecting the video acquisition device to the video signal loop of the target factory to obtain industrial video; Using the video output port of the video acquisition device to send the industrial video to a display screen of a target factory; The video capture device is used to copy the video images transmitted to the display screen in real time, and the video images are digitally processed to obtain industrial static data images; Establishing a plane rectangular coordinate system, and placing the industrial static data image in the first quadrant of the plane rectangular coordinate system; Determine the center coordinates and size of each single data image in the industrial static data image by using the plane rectangular coordinate system; Naming the individual data images according to the central coordinates to obtain name identifiers of the individual data images; Performing image interception on the industrial static data image according to the size to obtain each of the single data images; By setting the interception frequency of each of the single data images, a time mark of each of the single data images is obtained; Classifying the name identifiers to obtain a set of static data images of the same type; The single data images in the same type of static data image set are sequentially sorted according to the time identifier to obtain a sequential static data image set.

3. A method for industrial video data acquisition and recognition according to claim 2, characterized in that: The sequential static data image set is recognized by an image recognition device to obtain an original data set, including: Recognize the sequential static data image set to obtain recognition data; Setting a data name and a data time for the identification data according to the name identifier and the time identifier of the sequential static data image set; The identification data with the same data name are sorted in sequence according to the data time to obtain the original data set.

4. The method for industrial video data acquisition and recognition according to claim 1, characterized in that: The video signal loop is HDMI.

5. The method for industrial video data acquisition and recognition according to claim 1, characterized in that: The industrial production data situation feature set includes: a single data upper limit and a single data accuracy.

6. A method for industrial video data acquisition and recognition according to claim 5, characterized in that: By performing correlation evaluation on the industrial production data situation feature set and the original data set, abnormal data and normal data are obtained, and the abnormal data is corrected to obtain a corrected data set, including: Build data recognition models; Inputting the original data set and the industrial production data situation feature set into the data recognition model to perform threshold and accuracy comparison to obtain a recognition result; Determine the data in the original data set whose recognition result satisfies any one of the judgment criteria as the normal data; Determine the data in the original data set whose identification result does not meet any of the judgment criteria as the abnormal data; the judgment criteria are the upper limit and accuracy of the industrial production data situation feature set; Calculating the deviation coefficient of the abnormal data according to the industrial production data situation feature set; The abnormal data is calculated and corrected using the deviation coefficient to obtain the corrected data set.

7. A method for industrial video data acquisition and recognition according to claim 6, characterized in that: The absolute value of the deviation coefficient is the difference between the number of digits of the abnormal data and the precision of the single item of data.

8. A system for industrial video data acquisition and recognition, characterized in that: The method for industrial video data acquisition and recognition as described in claim 1, wherein the system comprises: A video acquisition module is used to acquire industrial video of a target factory through a video acquisition device to obtain a set of sequential static data images; An image recognition module, used to recognize the sequential static data image set through an image recognition device to obtain an original data set; The data situation feature module is used to perform big data analysis on all data features in the pre-recorded video using industrial production information as a search tag to obtain a set of industrial production data situation features; A data correction module, used for obtaining abnormal data and normal data by performing correlation evaluation on the industrial production data situation feature set and the original data set, and correcting the abnormal data to obtain a corrected data set; The data transmission module is used to store the normal data and the corrected data set in a database.

Images