Intrusion detection method and device for personnel in high-voltage site operation area, computer equipment, readable storage medium and program product

Abstract

The invention relates to an intrusion detection method and device for personnel in a high-voltage site operation area, computer equipment, a computer readable storage medium and a computer program product, and relates to the technical field of electric power safety. According to the invention, the speed and accuracy of intrusion detection can be improved. The method comprises the following steps: performing combinatorial analysis on image visual information in combination with a target factor to obtain an intrusion feature distribution model of a to-be-detected person; according to the patch primitive subdivision result, obtaining an intrusion feature fusion result and an echelette grating visual image of the to-be-detected person; inputting the echelette grating visual image into the echelette grating visual feature detection model for feature recognition to obtain a pixel component of the echelette grating visual image, and performing three-dimensional visual reconstruction on the to-be-detected person according to the pixel component to obtain a three-dimensional visual reconstruction model; and based on the three-dimensional vision reconstruction model, carrying out ambiguity analysis on the ambiguity of the video image, and carrying out intrusion detection on the to-be-detected person according to the obtained edge surrounding contour feature quantity.

Description

Technical Field

[0001] The present application relates to the field of power safety technology, and in particular to a method, device, computer equipment, computer-readable storage medium, and computer program product for detecting intrusion of personnel in high-voltage site operation areas. Background Art

[0002] High-voltage site operation areas are an essential and important link in industrial production and energy supply. However, due to the particularity and danger of high-voltage power equipment, illegal entry or violations by personnel may lead to serious accidents and losses. Therefore, safety monitoring and management of these areas are of great significance.

[0003] Traditional methods for monitoring high-voltage site operation areas often rely on manual patrols and manual alarms, which have problems such as blind spots and low real-time performance. In addition, the above-mentioned personnel monitoring methods in the operation area are difficult to extract abnormal features due to problems such as blurred edges and changes in personnel posture. They are also easily affected by ambient lighting. In complex operation area environments such as dark nights or low-light environments, it is difficult to continuously monitor intruders around the clock, resulting in low dynamic monitoring accuracy, which in turn affects work safety and production efficiency. Summary of the Invention

[0004] Based on this, it is necessary to provide a method, device, computer equipment, computer-readable storage medium and computer program product for detecting intrusion of personnel in high-voltage site working areas in response to the above technical problems.

[0005] In a first aspect, the present application provides a method for detecting intrusion of personnel in a high-voltage site operation area, comprising:

[0006] The image visual information of the person to be detected is collected using infrared grating technology, and the image visual information is combined and analyzed in combination with target factors to obtain an intrusion feature distribution model of the person to be detected; the person to be detected is a person working in a high-voltage site;

[0007] Performing local grid subdivision on the intrusion feature distribution model to obtain a patch primitive subdivision result of the infrared grating, and obtaining an intrusion feature fusion result of the infrared grating and an infrared grating visual image of the person to be detected based on the patch primitive subdivision result;

[0008] Inputting the infrared grating visual image into an infrared grating visual feature detection model for feature recognition to obtain pixel components of the infrared grating visual image, and performing three-dimensional visual reconstruction of the person to be detected based on the pixel components to obtain a three-dimensional visual reconstruction model;

[0009] Extracting the fuzziness of the video image of the infrared grating based on the three-dimensional visual reconstruction model, performing fuzziness analysis on the fuzziness of the video image, and obtaining edge enclosing contour feature quantities of the infrared grating visual image;

[0010] Intrusion detection is performed on the person to be detected based on the edge enclosing contour feature.

[0011] In one embodiment, before inputting the infrared grating visual image into an infrared grating visual feature detection model for feature recognition to obtain pixel components of the infrared grating visual image, the method further includes:

[0012] Dynamic feature segmentation and reconstruction processing is performed on the image visual information to obtain the edge function of the infrared grating visual image; based on the edge function, the characteristic decomposition instantaneous reconstruction parameters of the infrared grating visual image are obtained; based on the characteristic decomposition instantaneous reconstruction parameters, an intrusion infrared spectrum scale equation of the person to be detected is constructed; the intrusion infrared spectrum scale equation is used to construct the infrared grating visual feature detection model.

[0013] In one embodiment, performing three-dimensional visual reconstruction of the person to be detected based on the pixel components to obtain a three-dimensional visual reconstruction model includes:

[0014] Based on self-supervised learning of the image rotation angle, the adjacent feature points of the infrared grating pixels of the person to be detected are obtained as a target matrix according to the pixel components; a primitive data set is constructed according to the target matrix and the auxiliary distribution area parameters of the image rotation angle; based on the primitive data set, the three-dimensional visual reconstruction of the person to be detected is performed to obtain the three-dimensional visual reconstruction model.

[0015] In one embodiment, performing intrusion detection on the person to be detected based on the edge enclosing contour feature includes:

[0016] Determine whether the edge enclosing contour feature quantity meets the preset discrimination threshold condition; if the edge enclosing contour feature quantity meets the discrimination threshold condition, confirm the presence of the intrusion phenomenon of the person to be detected; generate intrusion warning information based on the time information and location information of the intrusion phenomenon.

[0017] In one embodiment, before performing intrusion detection on the person to be detected based on the edge enclosing contour feature, the method further includes:

[0018] Gradient information is used to extract the key points of the intrusion information of the person to be detected; based on the key points of the intrusion information, the fluctuation characteristic values of the intrusion offset and visual tracking and the infrared grating visual parameter set of the person to be detected are obtained; based on the fluctuation characteristic values and the infrared grating visual parameter set, a resolution threshold with the target point of the intruder is established, and based on the resolution threshold, the resolution threshold condition is obtained.

[0019] In one embodiment, the combined analysis of the image visual information in combination with the target factor to obtain the intrusion feature distribution model of the person to be detected includes:

[0020] Analyze the pixel-level image quality evaluation index parameter system, and based on the analysis results, obtain the numerical visual feature distribution of different reconstructed images; based on the visual feature distribution, combine the three target factors of brightness, contrast and structure to perform a combined analysis of the image visual information to obtain the intrusion feature distribution model of the person to be detected.

[0021] In a second aspect, the present application also provides an intrusion detection device for personnel in a high-voltage site operation area, comprising:

[0022] A combined analysis module is used to collect image visual information of the person to be detected using infrared grating technology, and perform combined analysis on the image visual information in combination with target factors to obtain an intrusion feature distribution model of the person to be detected; the person to be detected is a person working in a high-voltage site;

[0023] a grid subdivision module, configured to perform local grid subdivision on the intrusion feature distribution model to obtain a patch primitive subdivision result of the infrared grating, and obtain an intrusion feature fusion result of the infrared grating and an infrared grating visual image of the person to be detected based on the patch primitive subdivision result;

[0024] a feature recognition module, configured to input the infrared grating visual image into an infrared grating visual feature detection model for feature recognition, obtain pixel components of the infrared grating visual image, and perform three-dimensional visual reconstruction of the person to be detected based on the pixel components to obtain a three-dimensional visual reconstruction model;

[0025] A data analysis module is used to extract the fuzziness of the video image of the infrared grating based on the three-dimensional visual reconstruction model, perform fuzziness analysis on the fuzziness of the video image, and obtain edge encirclement contour feature quantities of the infrared grating visual image;

[0026] The intrusion detection module is used to perform intrusion detection on the person to be detected based on the edge enclosing contour feature.

[0027] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0028] Infrared grating technology is used to collect image visual information of the person to be detected, and the image visual information is combined and analyzed in combination with target factors to obtain an intrusion feature distribution model of the person to be detected; the person to be detected is a person in the high-voltage site working area; the intrusion feature distribution model is locally grid-subdivided to obtain the patch primitive subdivision result of the infrared grating, and based on the patch primitive subdivision result, the intrusion feature fusion result of the infrared grating and the infrared grating visual image of the person to be detected are obtained; the infrared grating visual image is input into the infrared grating visual feature detection model for feature recognition to obtain the pixel components of the infrared grating visual image, and based on the pixel components, the person to be detected is subjected to three-dimensional visual reconstruction to obtain a three-dimensional visual reconstruction model; based on the three-dimensional visual reconstruction model, the fuzziness of the video image of the infrared grating is extracted, and the fuzziness analysis of the video image fuzziness is performed to obtain the edge surrounding contour feature quantity of the infrared grating visual image; based on the edge surrounding contour feature quantity, intrusion detection is performed on the person to be detected.

[0029] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0030] Infrared grating technology is used to collect image visual information of the person to be detected, and the image visual information is combined and analyzed in combination with target factors to obtain an intrusion feature distribution model of the person to be detected; the person to be detected is a person in the high-voltage site working area; the intrusion feature distribution model is locally grid-subdivided to obtain the patch primitive subdivision result of the infrared grating, and based on the patch primitive subdivision result, the intrusion feature fusion result of the infrared grating and the infrared grating visual image of the person to be detected are obtained; the infrared grating visual image is input into the infrared grating visual feature detection model for feature recognition to obtain the pixel components of the infrared grating visual image, and based on the pixel components, the person to be detected is subjected to three-dimensional visual reconstruction to obtain a three-dimensional visual reconstruction model; based on the three-dimensional visual reconstruction model, the fuzziness of the video image of the infrared grating is extracted, and the fuzziness analysis of the video image fuzziness is performed to obtain the edge surrounding contour feature quantity of the infrared grating visual image; based on the edge surrounding contour feature quantity, intrusion detection is performed on the person to be detected.

[0031] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the following steps:

[0032] Infrared grating technology is used to collect image visual information of the person to be detected, and the image visual information is combined and analyzed in combination with target factors to obtain an intrusion feature distribution model of the person to be detected; the person to be detected is a person in the high-voltage site working area; the intrusion feature distribution model is locally grid-subdivided to obtain the patch primitive subdivision result of the infrared grating, and based on the patch primitive subdivision result, the intrusion feature fusion result of the infrared grating and the infrared grating visual image of the person to be detected are obtained; the infrared grating visual image is input into the infrared grating visual feature detection model for feature recognition to obtain the pixel components of the infrared grating visual image, and based on the pixel components, the person to be detected is subjected to three-dimensional visual reconstruction to obtain a three-dimensional visual reconstruction model; based on the three-dimensional visual reconstruction model, the fuzziness of the video image of the infrared grating is extracted, and the fuzziness analysis of the video image fuzziness is performed to obtain the edge surrounding contour feature quantity of the infrared grating visual image; based on the edge surrounding contour feature quantity, intrusion detection is performed on the person to be detected.

[0033] The above-mentioned intrusion detection method, device, computer equipment, computer-readable storage medium and computer program product for personnel in high-voltage site operation areas combine the visual feature distribution of different reconstructed images, and adopt the infrared grating sensor information fusion acquisition method to establish an intrusion feature analysis model for the personnel to be detected. By extracting the infrared grating video image blur of the personnel in the high-voltage site operation area and performing fuzziness analysis on the video image blur, the edge surrounding contour feature of the infrared grating visual image is extracted, and then the intrusion detection of the personnel to be detected is performed based on the edge surrounding contour feature, thereby avoiding inefficient and tedious processes such as manual patrols and manual alarms, realizing infrared grating detection of personnel intrusion in the high-voltage site operation area, improving the speed and accuracy of intrusion detection, and at the same time significantly enhancing the monitoring intensity and dynamic management capabilities of the high-voltage site operation area, so that the work safety and production efficiency of the high-voltage site operation area are effectively guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0035] Figure 1 This is a diagram of an application environment of a method for detecting intrusion by personnel in a high-voltage site operation area according to an embodiment;

[0036] Figure 2Schematic diagram of a flow chart of a method for detecting intrusion of personnel in a high-voltage site operation area according to one embodiment;

[0037] Figure 3 A schematic flow chart of the steps for constructing an intrusion infrared spectrum scale equation in one embodiment;

[0038] Figure 4 is a schematic diagram of intrusion feature distribution in one embodiment;

[0039] Figure 5 A flowchart of a method for detecting intrusion of personnel in a high-voltage site operation area according to a specific embodiment;

[0040] Figure 6 A schematic diagram of an infrared grating acquisition device for personnel in a high-voltage field operation area according to an embodiment;

[0041] Figure 7 A schematic diagram of infrared imaging of a high-voltage field operation area in one embodiment;

[0042] Figure 8 is a schematic diagram of a human intrusion detection result in one embodiment;

[0043] Figure 9 This is a structural block diagram of a device for detecting intrusion of personnel in a high-voltage site operation area according to one embodiment;

[0044] Figure 10 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0046] The intrusion detection method for personnel in high-voltage field operation areas provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, the terminal communicates with the server via a network. The data storage system can store data that the server needs to process. The data storage system can be integrated with the server or placed on the cloud or other network servers.

[0047] Specifically, the method for detecting intrusion of personnel in a high-voltage site operation area provided in the embodiment of the present application can be executed by a server.

[0048] Exemplarily, the server uses infrared grating technology to collect image visual information of the person to be detected, and performs combined analysis on the image visual information in combination with target factors to obtain an intrusion feature distribution model of the person to be detected; the server performs local grid subdivision on the intrusion feature distribution model to obtain the patch primitive subdivision result of the infrared grating, and obtains the intrusion feature fusion result of the infrared grating and the infrared grating visual image of the person to be detected based on the patch primitive subdivision result; the server inputs the infrared grating visual image into the infrared grating visual feature detection model for feature recognition to obtain the pixel components of the infrared grating visual image, and performs three-dimensional visual reconstruction on the person to be detected based on the pixel components to obtain a three-dimensional visual reconstruction model; based on the three-dimensional visual reconstruction model, the server extracts the blur of the video image of the infrared grating, performs blur analysis on the blur of the video image, and obtains the edge surrounding contour feature quantity of the infrared grating visual image; the server performs intrusion detection on the person to be detected based on the edge surrounding contour feature quantity.

[0049] In such Figure 1 In the application environment shown, the terminal can be, but is not limited to, various personal computers, laptops, smart phones, and tablet computers. The server can be implemented as an independent server or a server cluster consisting of multiple servers.

[0050] In one embodiment, Figure 2 As shown, a method for detecting intrusion of personnel in high-voltage site operation areas is provided. Figure 1 The following steps are used as an example to illustrate the server in the example:

[0051] Step S201, using infrared grating technology to collect image visual information of the person to be detected, combining and analyzing the image visual information in combination with target factors to obtain an intrusion feature distribution model of the person to be detected; the person to be detected is a person in the high-voltage site working area.

[0052] Among them, infrared grating technology is mainly used for infrared spectrum analysis and imaging. It uses the optical properties of the grating to decompose infrared light waves into different wavelengths, and analyzes the characteristics of the material by detecting light of different wavelengths.

[0053] Among them, the target factors can be brightness, contrast, structure, etc.

[0054] Specifically, the server uses infrared grating technology to collect image visual information of personnel in the high-voltage site operation area, analyzes the pixel-level image quality evaluation index parameter system, combines the numerical visual feature distribution of different reconstructed images, adopts the infrared grating sensor information fusion collection method, and combines the three factors of brightness, contrast and structure for combined analysis to establish an intrusion feature analysis model for personnel in the high-voltage site operation area.

[0055] Step S202 , performing local grid subdivision on the intrusion feature distribution model to obtain the patch primitive subdivision result of the infrared grating, and obtaining the intrusion feature fusion result of the infrared grating and the infrared grating visual image of the person to be detected based on the patch primitive subdivision result.

[0056] Among them, the intrusion feature fusion result can be , in the above formula, represents the joint eigencomponents of the maximum scale decomposition of the infrared grating, is the mapping diagram of different component factors, is the pixel data of the image in an inverted relationship, is the loss data of the model during the training phase, is the characteristic sampling frequency of personnel intrusion in the high-voltage site operation area, is the phase weight vector, Indicates the visual error compensation coefficient for personnel intrusion in the high-voltage site operation area.

[0057] Specifically, the server uses structured similarity feature analysis to perform local grid subdivision on the intrusion feature distribution model to obtain the patch element subdivision results of the infrared grating. Based on the patch element subdivision results, it is assumed that there are obvious differences in different component factors to obtain the intrusion feature fusion results of the infrared grating and the infrared grating visual image of the person to be detected.

[0058] Step S203: input the infrared grating visual image into the infrared grating visual feature detection model for feature recognition to obtain pixel components of the infrared grating visual image, and perform three-dimensional visual reconstruction of the person to be detected based on the pixel components to obtain a three-dimensional visual reconstruction model.

[0059] Among them, pixel components refer to the components of each pixel in a digital image. These pixel components determine the color and brightness of the pixel and are usually related to the color space of the image.

[0060] Among them, the 3D visual reconstruction model can be , in the above formula, is the estimated value of the missing area of personnel distribution, There is significant discrimination in the feature space, is the error distribution vector.

[0061] Specifically, the server combines the three-dimensional feature grouping detection method, inputs the infrared grating visual image into the infrared grating visual feature detection model for feature recognition, and uses the large-scale virtual scene simulation method to obtain the pixel components of the infrared grating visual image. Based on the pixel components, the three-dimensional visual reconstruction of the person to be detected is performed to obtain a three-dimensional visual reconstruction model.

[0062] Step S204 , based on the three-dimensional visual reconstruction model, extracting the video image blur of the infrared grating, performing blur analysis on the video image blur, and obtaining edge enclosing contour feature quantities of the infrared grating visual image.

[0063] Among them, the edge enclosing contour feature is a physical quantity that describes the characteristics of the object contour in the image and can be used for image analysis and processing.

[0064] Specifically, based on the design of a three-dimensional visual feature reconstruction model of personnel intrusion in the high-voltage site operation area, the server extracts the video image blur of the infrared grating of personnel in the high-voltage site operation area, performs fuzziness analysis on the video image blur, and obtains the edge surrounding contour feature of the infrared grating visual image.

[0065] Step S205: Perform intrusion detection on the person to be detected based on the edge enclosing contour feature.

[0066] Specifically, the server determines whether the edge enclosing contour feature quantity meets the preset resolution threshold condition; if the edge enclosing contour feature quantity meets the resolution threshold condition, it confirms the presence of an intrusion phenomenon by the person to be detected; and generates an intrusion warning message based on the time information and location information of the intrusion phenomenon.

[0067] In the above-mentioned intrusion detection method for personnel in the high-voltage site operation area, the visual feature distribution of different reconstructed images is combined, and the infrared grating sensor information fusion acquisition method is adopted to establish an intrusion feature analysis model for the personnel to be detected. By extracting the infrared grating video image fuzziness of the personnel in the high-voltage site operation area and performing fuzziness analysis on the video image fuzziness, the edge surrounding contour feature quantity of the infrared grating visual image is extracted, and then the intrusion detection of the personnel to be detected is performed based on the edge surrounding contour feature quantity, thereby avoiding inefficient and tedious processes such as manual patrols and manual alarms, realizing infrared grating detection of personnel intrusion in the high-voltage site operation area, improving the speed and accuracy of intrusion detection, and at the same time significantly enhancing the monitoring intensity and dynamic management capabilities of the high-voltage site operation area, so that the work safety and production efficiency of the high-voltage site operation area are effectively guaranteed.

[0068] In one embodiment, Figure 3 As shown, before inputting the infrared grating visual image into the infrared grating visual feature detection model for feature recognition to obtain pixel components of the infrared grating visual image, the method of the present application further includes the following steps:

[0069] Step S301 , performing dynamic feature segmentation and reconstruction processing on the image visual information to obtain the edge function of the infrared grating visual image.

[0070] Step S302: Obtain characteristic decomposition instantaneous reconstruction parameters of the infrared grating visual image according to the edge function.

[0071] Step S303 : constructing an infrared spectrum scale equation of the person to be detected based on the feature decomposition instantaneous reconstruction parameters; the infrared spectrum scale equation of the person to be detected is used to construct an infrared grating visual feature detection model.

[0072] Specifically, the server adopts the backbone feature extraction method to perform dynamic feature segmentation and reconstruction processing on the collected infrared grating visual images of personnel in the high-voltage site operation area, extracts the edge contour feature of the image, and obtains the edge function of the infrared grating visual image fusion of personnel in the high-voltage site operation area; based on the predicted image rotation angle analysis, the instantaneous reconstruction parameters of the infrared grating visual image feature decomposition of personnel in the high-voltage site operation area are obtained; based on the predicted image rotation feature analysis, the infrared spectrum scale equation of personnel intrusion in the high-voltage site operation area is obtained.

[0073] For example, based on the supervised contrastive learning method, the server uses the backbone feature extraction method to perform dynamic feature segmentation and reconstruction on the collected infrared grating visual images of personnel in the high-voltage site operation area, extract the edge contour feature of the image, and obtain the edge function of the infrared grating visual image fusion of personnel in the high-voltage site operation area:

[0074]

[0075] in, is the inter-class similarity parameter, is similar between classes, is the initial sampling rate, is the visual segmentation amplitude. Based on the predicted image rotation angle analysis, the instantaneous reconstruction parameter of the infrared grating visual image feature decomposition of personnel in the high-voltage site operation area is obtained as follows:

[0076]

[0077] calculate Order matrix Through autocorrelation feature matching, infrared grating visual image fusion and dynamic fusion tracking recognition of personnel in the high-voltage site operation area are performed, and the number of steps for infrared grating visual reconstruction of personnel in the high-voltage site operation area is obtained. .

[0078] Based on the analysis of the predicted image rotation characteristics, the infrared spectrum scale equation of human intrusion in the high-voltage site operation area is described as follows:

[0079]

[0080] in, for The visual component of the information of personnel intrusion in the high-voltage site operation area at all times; is the optimized template parameter in the training phase, and are the weight vector and unit weight of infrared grating visual control of personnel intrusion in high-voltage site operation areas, is the one-dimensional identity matrix.

[0081] Based on the above analysis, a visual feature detection model for infrared gratings of personnel in high-voltage site operation areas is designed. Combined with the three-dimensional feature grouping detection method, the feature recognition of infrared gratings of personnel in high-voltage site operation areas is carried out. The pixel components of the infrared grating visual image of personnel in high-voltage site operation areas are obtained by using the large-scale virtual scene simulation method. .

[0082] In one embodiment, in the above step S203, three-dimensional visual reconstruction is performed on the person to be detected based on pixel components to obtain a three-dimensional visual reconstruction model, which specifically includes the following steps:

[0083] Based on self-supervised learning of image rotation angle, the adjacent feature points of the infrared grating pixels of the person to be detected are obtained as the target matrix according to the pixel components; a primitive data set is constructed according to the target matrix and the auxiliary distribution area parameters of the image rotation angle; based on the primitive data set, the three-dimensional visual reconstruction of the person to be detected is performed to obtain a three-dimensional visual reconstruction model.

[0084] Specifically, the server uses a distributed pixel reconstruction method based on the self-supervised learning of the predicted image rotation angle to obtain the adjacent feature points of the infrared grating pixels of the personnel in the high-voltage site operation area as a Order matrix In the process of machine learning, the complex set of infrared grating visual image distribution of personnel in high-voltage site operation areas and are mutually prime, , if the grid feature distribution density of the infrared grating visual image of personnel in the high-voltage site operation area is: , when the auxiliary distribution area parameter of the image rotation angle is , build a metadata dataset , under the network training path, the visual reconstruction output distribution domain of personnel intrusion in the high-voltage site operation area , indicating that there is a solution to the intrusion characteristics of personnel in any high-voltage site operation area. indivual( ), the 3D visual reconstruction model of human intrusion is:

[0085]

[0086] in, is the estimated value of the missing area of personnel distribution, There is significant discrimination in the feature space, is the error distribution vector.

[0087] In one embodiment, based on the edge enclosing contour feature, before performing intrusion detection on the person to be detected, the method of the present application further includes the following steps:

[0088] Gradient information is used to extract the key points of intrusion information of the person to be detected; based on the key points of intrusion information, the fluctuation characteristic values of intrusion offset and visual tracking and the infrared grating visual parameter set of the person to be detected are obtained; based on the fluctuation characteristic values and the infrared grating visual parameter set, a resolution threshold with the target point of the intruder is established, and based on the resolution threshold, the resolution threshold condition is obtained.

[0089] Based on the pixel components, a 3D visual reconstruction of the person to be inspected is performed to obtain a 3D visual reconstruction model, including the following steps:

[0090] Determine whether the edge encirclement contour feature quantity meets the preset discrimination threshold condition; if the edge encirclement contour feature quantity meets the discrimination threshold condition, confirm the presence of an intrusion phenomenon by the person to be detected; generate an intrusion warning message based on the time information and location information of the intrusion phenomenon.

[0091] Specifically, based on the design of a three-dimensional visual feature reconstruction model for personnel intrusion in high-voltage site operation areas, the server extracts the fuzziness of the video image of the infrared grating of personnel in the high-voltage site operation area. Based on the fuzziness analysis, the edge surrounding contour feature of the image is extracted. The dynamic intrusion information extraction and abnormal feature detection of the image are realized through the infrared spectrum feature fusion technology. The view of the same anchor point is regarded as a positive sample pair, and the distribution delay of the high-voltage site operation area is obtained as follows: , the scale of machine vision tracking of personnel intrusion in high-voltage site operation areas is KL, and the expression is:

[0092]

[0093] in, represents minimizing the dimension of negative samples, is the sample pixel intensity, and the fitness function is obtained by maximizing the similarity information between the features of the positive sample pair Initialize the image pixel components of personnel intrusion in the high-voltage site operation area, regard the views of the same anchor point as a positive sample pair, and obtain the i-th view as a positive sample result:

[0094]

[0095] in, represents a positive sample pair, represents the augmented set of different anchor point samples, is the maximum similarity. According to the structural type feature distribution, the multi-level target distribution function is obtained as follows:

[0096] st

[0097] In the above formula, The distance between three-dimensional visualization points i and j representing a single relationship between two intrusion targets.

[0098] Based on the three-dimensional visualization distribution of ground and line surface, it is assumed that The geographic location information of the intruders in the two working areas at the moment is According to the perpendicular angle between the axis and the sun, the geographic information position value of the intruder target is obtained:

[0099]

[0100] In the above formula, Represents the mapping weight.

[0101] Initialize the prior feature quantity of the machine vision distribution of personnel intrusion in the high-voltage site operation area, and Under the constraints, the machine vision fusion group of personnel intrusion in the high-voltage site operation area is obtained , the boundary points of all machine images with personnel intrusion in the high-voltage site operation area are 1 / N, and the normalized weights are:

[0102]

[0103] Gradient information is used to extract the key points of intrusion information of high-voltage site operators. Through the spatial target planning method, the fluctuation characteristic values of intrusion offset and visual tracking are obtained, and the visual parameter set of infrared gratings of personnel in the high-voltage site operation area is obtained. , establish the discrimination threshold between the intruder target point and ,when hour, The visualization balance weight of the infrared grating collection of the intruder is ,Using gradient information fusion, the contour features of personnel in the high-voltage site operation area are analyzed and the visual distribution rules are as follows:

[0104]

[0105] in: is a set of convex feature points, Represents a virtual reference component.

[0106] Thus, the contour features of people intruding into the high-voltage site operation area are extracted, and then the intruder detection is realized.

[0107] In one embodiment, in the above step S201, the image visual information is combined and analyzed in combination with the target factor to obtain the intrusion feature distribution model of the person to be detected, which specifically includes the following steps:

[0108] The pixel-level image quality evaluation index parameter system is analyzed, and based on the analysis results, the numerical visual feature distribution of different reconstructed images is obtained; based on the visual feature distribution, the image visual information is combined and analyzed with the three target factors of brightness, contrast and structure to obtain the intrusion feature distribution model of the person to be detected.

[0109] Specifically, in order to realize the intrusion detection of personnel in the high-voltage site operation area based on infrared grating technology, the server uses infrared grating technology to realize the image visual information collection of personnel in the high-voltage site operation area, analyzes the pixel-level image quality evaluation index parameter system, combines the visual feature distribution of different reconstructed images in numerical terms, adopts the infrared grating sensor information fusion collection method, combines the brightness, contrast and structure factors to analyze the combination of three factors, and establishes the intrusion feature analysis model of personnel in the high-voltage site operation area. The intrusion feature distribution of personnel in the high-voltage site operation area is as follows: Figure 4 As shown. Using structural similarity feature analysis, Figure 4 The mesh model shown performs local mesh subdivision to obtain the patch element subdivision result of the infrared raster for personnel in the high-voltage site operation area:

[0110]

[0111] in, In order to effectively perceive the difference in image structure features, For structural similarity, is brightness, is the grayscale standard function, is the structural similarity index, is the estimated value of the minimized characteristic parameter. Assuming that there are obvious differences in different component factors, the intrusion feature fusion result of the infrared grating is obtained:

[0112]

[0113] In the above formula, represents the joint eigencomponents of the maximum scale decomposition of the infrared grating, is the mapping diagram of different component factors, is the pixel data of the image in an inverted relationship, is the loss data of the model during the training phase, is the characteristic sampling frequency of personnel intrusion in the high-voltage site operation area, is the phase weight vector, Indicates the visual error compensation coefficient for personnel intrusion in the high-voltage site operation area.

[0114] According to the above steps, an infrared grating visual image acquisition and grid segmentation model for personnel in the high-voltage site operation area was constructed, and the dynamic characteristic parameters of the intrusion were analyzed through the fusion of block grid areas.

[0115] In one embodiment, Figure 5 As shown, a method for detecting intrusion of personnel in a high-voltage site operation area in a specific embodiment is provided, which specifically includes the following steps:

[0116] In step S501, infrared grating technology is used to collect image visual information of the person to be detected, and the pixel-level image quality evaluation index parameter system is analyzed. Based on the analysis results, the numerical visual feature distribution of different reconstructed images is obtained; based on the visual feature distribution, the image visual information is combined and analyzed in combination with the three target factors of brightness, contrast and structure to obtain the intrusion feature distribution model of the person to be detected.

[0117] Step S502 , performing local grid subdivision on the intrusion feature distribution model to obtain the patch primitive subdivision result of the infrared grating, and obtaining the intrusion feature fusion result of the infrared grating and the infrared grating visual image of the person to be detected based on the patch primitive subdivision result.

[0118] Step S503, dynamically segment and reconstruct the image visual information to obtain the edge function of the infrared grating visual image; based on the edge function, obtain the characteristic decomposition instantaneous reconstruction parameters of the infrared grating visual image; based on the characteristic decomposition instantaneous reconstruction parameters, construct the intrusion infrared spectrum scale equation of the person to be detected; the intrusion infrared spectrum scale equation is used to construct the infrared grating visual feature detection model.

[0119] In step S504, the infrared grating visual image is input into the infrared grating visual feature detection model for feature recognition to obtain pixel components of the infrared grating visual image. Based on the self-supervised learning of the image rotation angle, the adjacent feature points of the infrared grating pixels of the person to be detected are obtained as the target matrix according to the pixel components; a primitive data set is constructed according to the auxiliary distribution area parameters of the target matrix and the image rotation angle; and based on the primitive data set, three-dimensional visual reconstruction of the person to be detected is performed to obtain a three-dimensional visual reconstruction model.

[0120] Step S505 , based on the three-dimensional visual reconstruction model, extracting the video image blur of the infrared grating, performing blur analysis on the video image blur, and obtaining edge enclosing contour feature quantities of the infrared grating visual image.

[0121] Step S506, using gradient information to extract the key points of the intrusion information of the person to be detected; based on the key points of the intrusion information, obtaining the fluctuation characteristic values of the intrusion offset and visual tracking and the infrared grating visual parameter set of the person to be detected; based on the fluctuation characteristic values and the infrared grating visual parameter set, establishing a resolution threshold with the target point of the intruder, and based on the resolution threshold, obtaining the resolution threshold condition.

[0122] Step S507, determine whether the edge enclosing contour feature quantity meets the preset discrimination threshold condition; if the edge enclosing contour feature quantity meets the discrimination threshold condition, confirm the presence of an intrusion phenomenon of the person to be detected; generate an intrusion warning message based on the time information and location information of the intrusion phenomenon.

[0123] The beneficial effects brought about by the above embodiment are as follows:

[0124] This scheme combines the visual feature distribution of different reconstructed images and adopts the infrared grating sensor information fusion acquisition method to establish an intrusion feature analysis model. By extracting the blurriness of the infrared grating video image of personnel in the high-voltage site operation area and extracting the edge surrounding contour feature of the image based on the blurriness analysis, the infrared grating detection of personnel intrusion in the high-voltage site operation area is realized.

[0125] For example, in order to verify the application performance of the present invention in realizing the infrared grating of personnel in the high-voltage site operation area, a simulation test analysis is carried out, and the visual information sampling pixels of the infrared grating of personnel in the high-voltage site operation area are set to 480*250, the grid distribution area size of the operation area is 10*10m, the carrier fundamental frequency component of the grating infrared spectrum is 0.25, the background area interference intensity is -24dB, the temperature measurement range of the grating equipment is -20℃~55℃, and the accuracy can reach ±2℃ or 2%. According to the above parameter settings, after the algorithm configuration is completed, it is connected to the back-end IVS / NVR equipment or monitoring platform to realize the unified management of the intrusion information of personnel in the high-voltage site operation area. Infrared grating acquisition equipment for personnel in the high-voltage site operation area such as Figure 6 shown.

[0126] according to Figure 6 The infrared grating acquisition results of personnel in the high-voltage site operation area are reconstructed visually to obtain the infrared imaging results as follows: Figure 7 shown.

[0127] from Figure 7 From the perspective of the present invention, this solution can effectively perform infrared grating visual imaging of high-voltage site operation areas, providing a reliable means for real-time monitoring, identification and recording of human intrusion behavior. It is not affected by ambient light and can work under various lighting conditions, including dark nights or low-light environments. This makes it suitable for the complex environments of various high-voltage site operation areas and can continuously monitor intruder detection around the clock, obtaining detection results such as Figure 8 As shown. Figure 8 It can be seen that this solution can effectively detect personnel intrusion into the area and realize abnormal dynamic monitoring of personnel in the operation area.

[0128] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0129] Based on the same inventive concept, the embodiments of the present application also provide a device for detecting intrusion of personnel in high-voltage field operation areas, which is used to implement the above-mentioned method for detecting intrusion of personnel in high-voltage field operation areas. The implementation solution provided by this device is similar to the implementation solution described in the above-mentioned method. Therefore, the specific limitations of one or more embodiments of the device for detecting intrusion of personnel in high-voltage field operation areas provided below can be found in the above-mentioned limitations of the method for detecting intrusion of personnel in high-voltage field operation areas, and will not be repeated here.

[0130] In an exemplary embodiment, Figure 9 As shown, a device for detecting intrusion of personnel in a high-voltage site operation area is provided, comprising:

[0131] The combined analysis module 901 is used to collect image visual information of the person to be detected using infrared grating technology, and perform combined analysis on the image visual information in combination with target factors to obtain an intrusion feature distribution model of the person to be detected; the person to be detected is a person working in a high-voltage site;

[0132] The grid subdivision module 902 is used to perform local grid subdivision on the intrusion feature distribution model to obtain the patch primitive subdivision result of the infrared grating, and obtain the intrusion feature fusion result of the infrared grating and the infrared grating visual image of the person to be detected based on the patch primitive subdivision result;

[0133] The feature recognition module 903 is used to input the infrared grating visual image into the infrared grating visual feature detection model for feature recognition, obtain pixel components of the infrared grating visual image, and perform three-dimensional visual reconstruction of the person to be detected based on the pixel components to obtain a three-dimensional visual reconstruction model;

[0134] The data analysis module 904 is used to extract the fuzziness of the infrared grating video image based on the three-dimensional visual reconstruction model, perform fuzziness analysis on the fuzziness of the video image, and obtain the edge enclosing contour feature of the infrared grating visual image;

[0135] The intrusion detection module 905 is used to perform intrusion detection on the person to be detected based on the edge enclosing contour feature.

[0136] In one embodiment, the intrusion detection device for personnel in high-voltage site operation areas also includes an equation construction module, which is used to perform dynamic feature segmentation and reconstruction processing on image visual information to obtain the edge function of the infrared grating visual image; based on the edge function, the characteristic decomposition instantaneous reconstruction parameters of the infrared grating visual image are obtained; based on the characteristic decomposition instantaneous reconstruction parameters, the intrusion infrared spectrum scale equation of the personnel to be detected is constructed; the intrusion infrared spectrum scale equation is used to construct an infrared grating visual feature detection model.

[0137] In one embodiment, the feature recognition module 903 is also used for self-supervised learning based on the image rotation angle, and obtains the adjacent feature points of the infrared grating pixels of the person to be detected as a target matrix according to the pixel components; constructs a primitive data set according to the target matrix and the auxiliary distribution area parameters of the image rotation angle; and performs three-dimensional visual reconstruction of the person to be detected based on the primitive data set to obtain a three-dimensional visual reconstruction model.

[0138] In one embodiment, the intrusion detection module 905 is also used to determine whether the edge enclosing contour feature quantity meets the preset resolution threshold condition; when the edge enclosing contour feature quantity meets the resolution threshold condition, it is confirmed that there is an intrusion phenomenon of the person to be detected; and an intrusion warning message is generated based on the time information and location information of the intrusion phenomenon.

[0139] In one embodiment, the intrusion detection device for personnel in the high-voltage site operation area also includes a condition setting module, which is used to use gradient information to extract the key points of the intrusion information of the person to be detected; based on the key points of the intrusion information, the fluctuation characteristic values of the intrusion offset and visual tracking and the infrared grating visual parameter set of the person to be detected are obtained; based on the fluctuation characteristic values and the infrared grating visual parameter set, a resolution threshold with the target point of the intruder is established, and based on the resolution threshold, a resolution threshold condition is obtained.

[0140] In one embodiment, the combined analysis module 901 is also used to analyze the pixel-level image quality evaluation index parameter system, and obtain the numerical visual feature distribution of different reconstructed images based on the analysis results; based on the visual feature distribution, the image visual information is combined with the three target factors of brightness, contrast and structure to obtain the intrusion feature distribution model of the person to be detected.

[0141] Each module in the aforementioned intrusion detection device for personnel in high-voltage field operation areas may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor within a computer device in hardware form, or may be stored in a computer device memory in software form, allowing the processor to call and execute the corresponding operations of each module.

[0142] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 10 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, memory and input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for detecting intrusion of personnel in a high-voltage site operation area is implemented.

[0143] Those skilled in the art will understand that Figure 10 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0144] In one embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0145] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0146] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0147] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0148] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), quantum computing-based data processing logic devices, artificial intelligence (AI) processors, and the like.

[0149] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0150] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A method for detecting intrusion of personnel in a high-voltage site operation area, characterized in that: The method comprises: The image visual information of the person to be detected is collected using infrared grating technology, and the image visual information is combined and analyzed in combination with target factors to obtain an intrusion feature distribution model of the person to be detected; the person to be detected is a person working in a high-voltage site; Performing local grid subdivision on the intrusion feature distribution model to obtain a patch primitive subdivision result of the infrared grating, and obtaining an intrusion feature fusion result of the infrared grating and an infrared grating visual image of the person to be detected based on the patch primitive subdivision result; Inputting the infrared grating visual image into an infrared grating visual feature detection model for feature recognition to obtain pixel components of the infrared grating visual image, and performing three-dimensional visual reconstruction of the person to be detected based on the pixel components to obtain a three-dimensional visual reconstruction model; Extracting the fuzziness of the video image of the infrared grating based on the three-dimensional visual reconstruction model, performing fuzziness analysis on the fuzziness of the video image, and obtaining edge enclosing contour feature quantities of the infrared grating visual image; Intrusion detection is performed on the person to be detected based on the edge enclosing contour feature.

2. The method according to claim 1, characterized in that Before inputting the infrared grating visual image into the infrared grating visual feature detection model for feature recognition to obtain pixel components of the infrared grating visual image, the method further includes: Performing dynamic feature segmentation and reconstruction processing on the image visual information to obtain the edge function of the infrared grating visual image; According to the edge function, characteristic decomposition instantaneous reconstruction parameters of the infrared grating visual image are obtained; According to the characteristic decomposition instantaneous reconstruction parameters, an intrusion infrared spectrum scale equation of the person to be detected is constructed; the intrusion infrared spectrum scale equation is used to construct the infrared grating visual feature detection model.

3. The method according to claim 1, characterized in that The step of performing three-dimensional visual reconstruction on the person to be detected based on the pixel components to obtain a three-dimensional visual reconstruction model includes: Based on the self-supervised learning of the image rotation angle, the adjacent feature points of the infrared grating pixels of the person to be detected are obtained as a target matrix according to the pixel components; Constructing a primitive data set according to the target matrix and the auxiliary distribution area parameters of the image rotation angle; The person to be detected is subjected to three-dimensional visual reconstruction according to the primitive data set to obtain the three-dimensional visual reconstruction model.

4. The method according to claim 1, wherein The step of performing intrusion detection on the person to be detected based on the edge enclosing contour feature comprises: Determining whether the edge enclosing contour feature quantity meets a preset resolution threshold condition; In the case where the edge enclosing contour feature quantity satisfies the discrimination threshold condition, confirming the presence of the intrusion phenomenon of the person to be detected; Intrusion warning information is generated according to the time information and location information of the intrusion phenomenon.

5. The method according to claim 4, characterized in that Before performing intrusion detection on the person to be detected based on the edge enclosing contour feature, the method further includes: Using gradient information to extract key points of intrusion information of the person to be detected; According to the intrusion information key points, the intrusion offset and the fluctuation characteristic value of the visual tracking and the infrared grating visual parameter set of the person to be detected are obtained; A discrimination threshold for the intruder target point is established based on the fluctuation characteristic value and the infrared grating visual parameter set, and the discrimination threshold condition is obtained based on the discrimination threshold.

6. The method according to any one of claims 1 to 5, characterized in that The combined analysis of the image visual information in combination with the target factor to obtain the intrusion feature distribution model of the person to be detected includes: Analyze the pixel-level image quality evaluation index parameter system, and based on the analysis results, obtain the numerical visual feature distribution of different reconstructed images; According to the visual feature distribution, the image visual information is combined and analyzed in combination with the three target factors of brightness, contrast and structure to obtain the intrusion feature distribution model of the person to be detected.

7. A device for detecting intrusion of personnel in high-voltage field operation areas, characterized in that: The device comprises: A combined analysis module is used to collect image visual information of the person to be detected using infrared grating technology, and perform combined analysis on the image visual information in combination with target factors to obtain an intrusion feature distribution model of the person to be detected; the person to be detected is a person working in a high-voltage site; a grid subdivision module, configured to perform local grid subdivision on the intrusion feature distribution model to obtain a patch primitive subdivision result of the infrared grating, and obtain an intrusion feature fusion result of the infrared grating and an infrared grating visual image of the person to be detected based on the patch primitive subdivision result; a feature recognition module, configured to input the infrared grating visual image into an infrared grating visual feature detection model for feature recognition, obtain pixel components of the infrared grating visual image, and perform three-dimensional visual reconstruction of the person to be detected based on the pixel components to obtain a three-dimensional visual reconstruction model; A data analysis module is used to extract the fuzziness of the video image of the infrared grating based on the three-dimensional visual reconstruction model, perform fuzziness analysis on the fuzziness of the video image, and obtain edge encirclement contour feature quantities of the infrared grating visual image; The intrusion detection module is used to perform intrusion detection on the person to be detected based on the edge enclosing contour feature.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images