Screening and evaluation method and device of image processing technology, electronic equipment and storage medium

Abstract

The invention provides a screening and evaluation method and device for an image processing technology, electronic equipment and a storage medium, and belongs to the field of devices for big data retrieval and data processing. According to the screening and evaluation method adopting the image processing technology, aiming at the obtained ultraviolet images, the processing methods of screening out the ultraviolet images meeting the requirements from the data source can be pre-classified according to keywords and high-frequency words in literatures or patents and the like, and the subsequent workload is reduced. Strong feature extraction and classification capabilities of machine learning are utilized to carry out classification marking on ultraviolet image validity evaluation methods in literatures, patents and the like. Then, the search screening data volume is gradually reduced, a data filtering structure similar to an hourglass is formed, meanwhile, a method library is established, the retrieval time is greatly shortened, and the retrieval efficiency is improved. Under the condition that the selection accuracy of the ultraviolet image processing method is ensured, the screening workload is greatly reduced, and the troubleshooting analysis efficiency is improved.

Description

Technical Field

[0001] The present invention relates to devices in the field of big data retrieval and data processing, and in particular to a screening method, device, electronic equipment and storage medium of image processing technology. Background Art

[0002] As an important part of the transmission line, the operational reliability of power cables directly affects the safety of the power grid. When the line insulators, hardware and other equipment on the cable line fail, they will discharge and radiate ultraviolet light at the surface fault site. The discharge corona characteristic quantity of the fault can be effectively detected by the sensing equipment. Based on the discharge corona characteristic quantity, the staff can troubleshoot and analyze the equipment operation and fault conditions on the cable line.

[0003] In related technologies, ultraviolet feature extraction usually uses an unmanned aerial vehicle system equipped with ultraviolet optical detection equipment to capture images along the cable line. After the staff processes and screens the captured images, they judge the fault of the cable line based on the ultraviolet images.

[0004] However, the feature extraction method used in related technologies obtains a large number of images, and is affected by the external environment and shooting angle. The extraction of feature quantities in the obtained images often needs to be evaluated by different standards. However, there are many types of image processing methods at present, and there is a lack of screening of massive methods for image processing technology and evaluation of the effects of these methods on ultraviolet image processing. For specific ultraviolet images, staff often need to spend a lot of time searching and selecting image processing methods that meet the requirements for corresponding processing, which increases the workload and affects the efficiency of cable line inspection. Summary of the invention

[0005] The embodiment of the present invention provides a screening method, device, electronic device and computer storage medium of image processing technology, which can significantly reduce the screening workload and improve the efficiency of investigation and analysis while ensuring the accuracy of the selection of ultraviolet image processing methods. The technical solution is as follows:

[0006] In a first aspect, an embodiment of the present invention provides a method for screening and evaluating image processing technology, comprising: screening out a set of processing methods for ultraviolet images that meet requirements from a data source, wherein the data source includes literature and patents in an Internet database;

[0007] Using the processing methods in the processing method set in sequence to verify the method of the ultraviolet image, normalizing the image size and color space of the evaluation results generated by different methods to obtain a normalized result set;

[0008] A set of evaluation methods for ultraviolet image effectiveness evaluation methods that meet the requirements are screened out from the data sources;

[0009] The evaluation results in the normalized result set are evaluated in sequence using the methods in the evaluation method set to obtain the scores of the processing methods of each ultraviolet image in the processing method set, and the optimal processing method for the ultraviolet image is obtained based on the scores.

[0010] Optionally, the processing method set for selecting ultraviolet images meeting the requirements from the data source includes:

[0011] Integrate and pre-classify the processing methods that meet the conditions in the data source based on the initial search conditions;

[0012] The integrated and pre-classified processing methods are classified and labeled using machine learning methods, and the labeled methods are added to the method library.

[0013] Optionally, the integrating and pre-classifying the processing methods that meet the conditions in the data source based on the initial search conditions includes:

[0014] Searching the data source for the initial search condition and analyzing each of them;

[0015] When analyzing each item, make a preliminary judgment on the field based on the high-frequency words and keywords in the literature and patents and remove useless data;

[0016] The results of the data source search and analysis are exported to the same format, and the document management tool is used to automatically remove duplicates and integrate the search results.

[0017] Optionally, the using of a machine learning method to classify and label the integrated and pre-classified processing methods, and adding the labeled methods to a method library, comprises:

[0018] Selecting some documents and patents from the integrated search results, marking them in detail and making them into a usable data set, and extracting meaningful features from the texts of the documents and patents, substituting them into the model and training them;

[0019] Based on the trained model, all documents and patents in the search results are classified and marked and added to the method library to form a processing method set.

[0020] Optionally, the score is an average score obtained by evaluating a single evaluation result using multiple evaluation methods.

[0021] In a second aspect, the present invention provides a device, comprising:

[0022] A method integration screening module is used to screen out a set of ultraviolet image processing methods that meet the requirements from a data source, wherein the data source includes documents and patents in an Internet database, and is also used to screen out a set of evaluation methods for ultraviolet image effectiveness evaluation methods that meet the requirements from the data source;

[0023] A verification processing module, used to sequentially use the processing methods in the processing method set to perform method verification on the ultraviolet image, and perform normalization processing on the image size and color space of the evaluation results generated by different methods to obtain a normalized result set;

[0024] An evaluation module is used to evaluate the evaluation results in the normalized result set using the methods in the evaluation method set in turn to obtain the scores of the processing methods of each ultraviolet image in the processing method set, and obtain the optimal processing method for the ultraviolet image based on the scores.

[0025] Optionally, the method integrated screening module is further used to search the initial search conditions in the data source and analyze them one by one;

[0026] When analyzing each item, make a preliminary judgment on the field based on the high-frequency words and keywords in the literature and patents and remove useless data;

[0027] The results of the data source search and analysis are exported to the same format, and the document management tool is used to automatically remove duplicates and integrate the search results.

[0028] Optionally, the method integration screening module is further used to select some documents and patents from the integrated search results, mark them in detail and make them into a usable data set, and extract meaningful features from the texts of the documents and patents, substitute them into the model and train them;

[0029] Based on the trained model, all documents and patents in the search results are classified and marked and added to the method library to form a processing method set.

[0030] In a third aspect, the present invention further provides an electronic device, comprising:

[0031] processor;

[0032] a memory for storing instructions executable by a processor;

[0033] Wherein, the processor is configured to execute the screening method of the image processing technology described in the first aspect above.

[0034] In a fourth aspect, the present invention further provides a storage medium having computer instructions stored thereon, wherein the computer instructions, when executed by a processor, implement the screening method of the image processing technology described in the first aspect.

[0035] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0036] The screening and evaluation method of the image processing technology provided by the embodiment of the present invention can pre-classify the processing methods according to the keywords and high-frequency words in the literature or patents, etc., for the acquired ultraviolet images by screening out the processing methods of ultraviolet images that meet the requirements from the data source, thereby reducing the subsequent workload. The strong feature extraction and classification capabilities of machine learning are used to classify and mark the ultraviolet image effectiveness evaluation methods in the literature, patents, etc. Then, the amount of search and screening data is gradually reduced to form a data filtering structure similar to an hourglass, and a method library is established at the same time, which greatly shortens the retrieval time and improves the retrieval efficiency. The processing methods can be further screened in the process of verification according to the screened processing methods, and the data such as the processing time generated in the verification process can assist in the subsequent method result evaluation. The normalization operation of the total score of the evaluation method can eliminate the differences between different evaluation methods, so that in the final comprehensive score, each evaluation method has the same weight. The comprehensive score can make the final method score closer to reality, and the final method score will not be affected by the excessively high or low score under a certain evaluation method. It can greatly reduce the screening workload and improve the efficiency of investigation and analysis while ensuring the accuracy of the selection of ultraviolet image processing methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 creative work.

[0038] Figure 1 It is a flowchart of a screening method of an image processing technology provided by an embodiment of the present invention;

[0039] Figure 2 It is a schematic diagram of a screening method flow of an image processing technology provided by an embodiment of the present invention;

[0040] Figure 3 is a structural block diagram of a screening device of an image processing technology provided by an embodiment of the present invention;

[0041] Figure 4 It is a schematic diagram of a control structure of an electronic device provided by an embodiment of the present invention.

[0042] In the figure:

[0043] 1-Method integration and screening module; 2-Verification and processing module; 3-Evaluation module. DETAILED DESCRIPTION

[0044] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0045] Embodiment 1:

[0046] Figure 1 It is a flowchart of a screening method of an image processing technology provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of a screening method flow of an image processing technology provided by an embodiment of the present invention; Figure 3 1 is a structural block diagram of a screening device for image processing technology provided by an embodiment of the present invention. Figures 1 to 3 As shown, an embodiment of the present invention provides a screening method of image processing technology, and the screening method of image processing technology includes:

[0047] S1. A set of UV image processing methods F1 that meet the requirements is selected from data sources, where the data sources include literature and patents in Internet databases.

[0048] Specifically, this step includes:

[0049] S11. Based on the initial search conditions, the processing methods that meet the conditions in the data sources are integrated and pre-classified. The data sources include literature and patents in databases such as ScienceDirect, SpringerLink, IEEE, CNKI and Wanfang.

[0050] Further, integration and pre-classification include the following processes:

[0051] S111. Search the data source for the initial search conditions and analyze them one by one;

[0052] S112. When analyzing each item, make a preliminary judgment on the field based on the high-frequency words and keywords in the literature and patents and remove useless data;

[0053] S113. Export the results of the data source search and analysis into the same format, and use the document management tool to automatically remove duplicates and integrate the search results. The integration method of the unified format facilitates subsequent operations.

[0054] It should be noted that the pre-classification can also use the Delphi method to modify the patent and literature keywords and high-frequency phrases in each technical field classification, and delete the patent and literature keywords and high-frequency phrases that do not belong to the set. According to the keywords and high-frequency words in the source literature, patents, and search engine results of the method, the method is first briefly divided into fields and irrelevant methods are eliminated to achieve the purpose of reducing the amount of data.

[0055] S12. Use machine learning methods to classify and label the processing methods after integration and pre-classification, and add the labeled methods to the method library.

[0056] Further,

[0057] S121. Select some documents and patents from the integrated search results, mark them in detail and make them into a usable data set, extract meaningful features from the texts of the documents and patents, substitute them into the model and train them.

[0058] S122. Classify and label all documents and patents in the search results based on the trained model and add them to the method library to form a processing method set.

[0059] It should be noted that classification tagging includes: different classification methods based on the advantages and disadvantages, types, applicable fields, etc. of the method and labeling them accordingly. The method library is used to store labeled methods, which can effectively save subsequent tagging time, avoid repeated tagging, and ultimately form a usable data set. Extracting meaningful features from the text of literature and patents refers to keywords, phrases, syntactic structures, topic model vectors, etc. Natural language processing (NLP) tools can be used for preprocessing (such as word segmentation, stop word removal, stemming, etc.).

[0060] The selected model needs to be selected according to the characteristics of the task, such as naive Bayes, support vector machine, decision tree, random forest, or deep learning models such as convolutional neural network (CNN), recurrent neural network (RNN), Transformer and other machine learning models.

[0061] S2. Use the processing methods in the processing method set in turn to verify the ultraviolet image, normalize the image size and color space of the evaluation results generated by different methods, and obtain a normalized result set.

[0062] It should be noted that method verification is to use the methods in method set F1 in sequence to process the same ultraviolet image data set, and integrate the processing results and usage methods into one data set.

[0063] Furthermore, the evaluation results of different ultraviolet image processing methods may ultimately result in different processing results in output size and color space. Therefore, the present invention selects a unified image size and color space to normalize the processing results to facilitate subsequent evaluation operations.

[0064] S3. Filter out an evaluation method set F2 of ultraviolet image effectiveness evaluation methods that meet the requirements from the data source.

[0065] It should be noted that a prerequisite that the evaluation method in F2 must meet is that the normalized image processed by the current ultraviolet image processing method can be evaluated.

[0066] It should be noted that being able to be evaluated means that the normalized image meets the constraints of the input data of the evaluation method, and the evaluation method can effectively distinguish the advantages and disadvantages of this type of image processing method.

[0067] S4. Use the methods in the evaluation method set to evaluate the evaluation results in the normalized result set in turn to obtain the score of each ultraviolet image processing method in the processing method set, and obtain the optimal ultraviolet image processing method based on the score.

[0068] Furthermore, the number of methods that can be evaluated for each ultraviolet image processing method is not the same, that is, the number of methods in the evaluation method set F2 is different, so the comprehensive score is: the total score of N evaluation methods / N.

[0069] The screening and evaluation method of the image processing technology provided by the embodiment of the present invention can pre-classify the processing methods according to the keywords and high-frequency words in the literature or patents, etc., for the acquired ultraviolet images by screening out the processing methods of ultraviolet images that meet the requirements from the data source, thereby reducing the subsequent workload. The strong feature extraction and classification capabilities of machine learning are used to classify and mark the ultraviolet image effectiveness evaluation methods in the literature, patents, etc. Then, the amount of search and screening data is gradually reduced to form a data filtering structure similar to an hourglass, and a method library is established at the same time, which greatly shortens the retrieval time and improves the retrieval efficiency. The processing methods can be further screened in the process of verification according to the screened processing methods, and the data such as the processing time generated in the verification process can assist in the subsequent method result evaluation. The normalization operation of the total score of the evaluation method can eliminate the differences between different evaluation methods, so that in the final comprehensive score, each evaluation method has the same weight. The comprehensive score can make the final method score closer to reality, and the final method score will not be affected by the excessively high or low score under a certain evaluation method. It can greatly reduce the screening workload and improve the efficiency of investigation and analysis while ensuring the accuracy of the selection of ultraviolet image processing methods.

[0070] Embodiment 2:

[0071] Figure 3 1 is a structural block diagram of a screening device for image processing technology provided by an embodiment of the present invention. Figure 3As shown, an embodiment of the present invention also provides a screening device for image processing technology, which has the function of implementing the screening method of the above-mentioned image processing technology. It includes a method integration screening module 1, a verification processing module 2 and an evaluation module 3. Among them, the method integration screening module 1 is used to screen out a set of processing methods for ultraviolet images that meet the requirements from the data source, and the data source includes documents and patents in the Internet database, and is also used to screen out a set of evaluation methods for ultraviolet image effectiveness evaluation methods that meet the requirements from the data source. The verification processing module 2 is used to sequentially use the processing methods in the processing method set to verify the method of the ultraviolet image, and normalize the image size and color space of the evaluation results generated by different methods to obtain a normalized result set. The evaluation module 3 is used to sequentially use the methods in the evaluation method set to evaluate the evaluation results in the normalized result set, so as to obtain the score of each ultraviolet image processing method in the processing method set, and obtain the optimal ultraviolet image processing method based on the score.

[0072] Furthermore, the method integration and screening module 1 is also used to search the initial search conditions in the data source and analyze them one by one. When analyzing them one by one, the field is preliminarily determined based on the high-frequency words and keywords in the literature and patents and useless data is removed. The results of the data source search and analysis are exported to the same format, and the document management tool is used to automatically remove duplicates and integrate the search results.

[0073] Furthermore, the method integration and screening module 1 is also used to select some documents and patents in the integrated search results for detailed marking and make them into a usable data set, and extract meaningful features from the texts of the documents and patents, substitute them into the model and train them. Based on the trained model, all documents and patents in the search results are classified and marked and added to the method library to form a processing method set.

[0074] Embodiment 3:

[0075] Figure 4 Schematic diagram of a control structure of an electronic device provided by an embodiment of the present invention. Figure 4 As shown, the electronic device includes a processor and a memory for storing instructions executable by the processor. The electronic device 4100 may be a computer device, and the electronic device 4100 may include one or more of the following components: a processor 4101 , a memory 4102 , a communication interface 4103 , and a bus 4104 .

[0076] The processor 4101 includes one or more processing cores, and the processor 4101 executes various functional applications and information processing by running software programs and modules. The memory 4102 and the communication interface 4103 are connected to the processor 4101 through the bus 4104. The memory 4102 can be used to store at least one instruction, and the processor 4101 is used to execute the at least one instruction to implement each step in the above method embodiment.

[0077] In addition, memory 4102 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, and volatile or non-volatile storage devices include but are not limited to: magnetic disks or optical disks, electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), static random access memory (SRAM), read-only memory (ROM), magnetic memory, flash memory, and programmable read-only memory (PROM).

[0078] Embodiment 4:

[0079] Exemplarily, in an embodiment of the present invention, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory including instructions, and the instructions can be used by a processor to execute the aforementioned image processing technology screening method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0080] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0081] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (apparatus), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0082] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0083] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0084] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit its protection scope. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that after reading the present invention, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the pending claims of the invention.

[0085] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit its protection scope. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that after reading the present invention, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the pending claims of the invention.

Claims

1. A screening method of image processing technology, characterized in that: include: A set of ultraviolet image processing methods that meet the requirements are screened out from data sources, wherein the data sources include literature and patents in Internet databases; Using the processing methods in the processing method set in turn to verify the method of the ultraviolet image, normalizing the image size and color space of the evaluation results generated by different methods to obtain a normalized result set; A set of evaluation methods for ultraviolet image effectiveness evaluation methods that meet the requirements are screened out from the data sources; The evaluation results in the normalized result set are evaluated in sequence using the methods in the evaluation method set to obtain the scores of the processing methods of each ultraviolet image in the processing method set, and the optimal processing method for the ultraviolet image is obtained based on the scores.

2. The screening method of image processing technology according to claim 1, characterized in that: The processing method set for selecting ultraviolet images meeting the requirements from the data source includes: Integrate and pre-classify the processing methods that meet the conditions in the data source based on the initial search conditions; The integrated and pre-classified processing methods are classified and labeled using machine learning methods, and the labeled methods are added to the method library.

3. The screening method of image processing technology according to claim 2, characterized in that: The step of integrating and pre-classifying the processing methods that meet the conditions in the data source based on the initial search conditions includes: Searching the data source for the initial search condition and analyzing each of them; When analyzing each item, make a preliminary judgment on the field based on the high-frequency words and keywords in the literature and patents and remove useless data; The results of the data source search and analysis are exported to the same format, and the document management tool is used to automatically remove duplicates and integrate the search results.

4. The screening method of image processing technology according to claim 3, characterized in that: The method of using a machine learning method to classify and label the processing methods after integration and pre-classification, and adding the labeled methods to a method library, comprises: Selecting some documents and patents from the integrated search results, marking them in detail and making them into a usable data set, and extracting meaningful features from the texts of the documents and patents, substituting them into the model and training them; Based on the trained model, all documents and patents in the search results are classified and marked and added to the method library to form a processing method set.

5. The screening method of image processing technology according to claim 1, characterized in that: The score is an average score obtained by performing evaluations using a plurality of evaluation methods on a single evaluation result.

6. A screening device using image processing technology, characterized in that: include: A method integration screening module (1) is used to screen out a set of ultraviolet image processing methods that meet the requirements from a data source, wherein the data source includes documents and patents in an Internet database, and is also used to screen out a set of evaluation methods for ultraviolet image effectiveness evaluation methods that meet the requirements from the data source; A verification processing module (2) is used to sequentially use the processing methods in the processing method set to perform method verification on the ultraviolet image, and perform normalization processing on the image size and color space of the evaluation results generated by different methods to obtain a normalized result set; An evaluation module (3) is used to evaluate the evaluation results in the normalized result set using the methods in the evaluation method set in turn to obtain the scores of the processing methods of each ultraviolet image in the processing method set, and to obtain the optimal processing method for the ultraviolet image based on the scores.

7. The device according to claim 6, characterized in that The method integrated screening module (1) is also used to search the initial search conditions in the data source and analyze them one by one; When analyzing each item, make a preliminary judgment on the field based on the high-frequency words and keywords in the literature and patents and remove useless data; The results of the data source search and analysis are exported to the same format, and the document management tool is used to automatically remove duplicates and integrate the search results.

8. The device according to claim 7, characterized in that The method integration screening module (1) is also used to select some documents and patents from the integrated search results, mark them in detail and make them into a usable data set, and extract meaningful features from the texts of the documents and patents, substitute them into the model and train them; Based on the trained model, all documents and patents in the search results are classified and marked and added to the method library to form a processing method set.

9. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by a processor; Wherein, the processor is configured to execute the screening method of the image processing technology described in any one of claims 1 to 4.

10. A storage medium having computer instructions stored thereon, characterized in that: When the computer instructions are executed by a processor, the screening method of the image processing technology described in any one of claims 1 to 5 is implemented.

Images