Method, system and readable storage medium for identifying risks of near-electrical operations

Abstract

The present invention provides a method, system, and readable storage medium for identifying risks associated with near-electrical operations. The method includes: inputting a panoramic image into a first target recognition model and a second target recognition model, respectively; determining whether there is an overlapping area between the first target and the second target based on the position information of the first target and the second target; if there is an overlapping area between the first target and the second target, segmenting the overlapping area from the panoramic image and inputting the overlapping area into a pre-trained regression model; performing pixel separation on the overlapping area based on a pixel separation coefficient; obtaining first spatial coordinates of each pixel point in the first target based on first pixel information and first coordinate information, obtaining second spatial coordinates of each pixel point in the second target based on second pixel information and second coordinate information, and determining whether there is a risk associated with near-electrical operations based on the first spatial coordinates and the second spatial coordinates. The present invention can improve the accuracy of identifying the current risk associated with near-electrical operations.

Description

Technical Field

[0001] The present invention relates to the technical field of risk identification for near-electrical operations, and in particular to a method, system, and readable storage medium for risk identification for near-electrical operations. Background Art

[0002] In near-electrical work scenarios, accurately identifying the relative positional relationship between workers and live equipment and promptly identifying potential safety risks are crucial to ensuring workers' safety. However, traditional methods for identifying risks near-electrical work often rely on manual inspections or simple image processing techniques, which have numerous limitations. On the one hand, manual inspections are inefficient and susceptible to human factors, making it difficult to achieve all-weather, all-around monitoring. On the other hand, simple image processing techniques struggle to accurately identify target objects and their spatial relationships in complex scenarios, resulting in low risk identification accuracy.

[0003] With the rapid development of computer vision and deep learning technologies, image recognition-based risk identification methods for near-electrical work have gradually become a research hotspot. However, existing image recognition-based risk identification methods still face significant challenges when dealing with complex scenarios such as overlapping and occluded objects. Specifically, when workers and live equipment overlap in an image, traditional methods struggle to accurately segment the overlapping area, making it impossible to accurately calculate the spatial distance between the two, thus affecting the accuracy of risk identification. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, system and readable storage medium for identifying risks of near-electrical operations, aiming to solve the problem that traditional technologies are unable to accurately divide overlapping areas and accurately calculate the spatial distance between the two, resulting in low accuracy in identifying risks of near-electrical operations.

[0005] In a first aspect, the present invention provides a method for identifying risks of near-electrical operations, the method comprising:

[0006] Acquire a panoramic image of an operator performing near-electrical work, and input the panoramic image into a first target recognition model and a second target recognition model, respectively, to obtain position information about the first target and the second target;

[0007] Determining whether there is an overlapping area between the first target and the second target based on the position information of the first target and the position information of the second target;

[0008] If there is an overlapping area between the first target and the second target, segmenting the overlapping area from the panoramic image, and inputting the overlapping area into a pre-trained regression model to obtain a pixel separation coefficient;

[0009] Performing pixel separation on the overlapping area according to the pixel separation coefficient, and obtaining first pixel information of each pixel point in the area covered by the first target and first coordinate information corresponding to the first pixel information, and second pixel information of each pixel point in the area covered by the second target and second coordinate information corresponding to the second pixel information according to the pixel separation result;

[0010] The first spatial coordinates of each pixel point in the first target are obtained based on the first pixel information and the first coordinate information, the second spatial coordinates of each pixel point in the second target are obtained based on the second pixel information and the second coordinate information, and whether there is a risk of near-electrical operation is determined based on the first spatial coordinates and the second spatial coordinates.

[0011] In summary, according to the above-mentioned method for identifying risks of near-electrical operations, the first and second targets in the overlapping area are accurately segmented to restore the pixel information originally belonging to the first and second targets, and then the spatial distance between the two is accurately calculated, thereby greatly improving the accuracy of identifying the current risk of near-electrical operations. Specifically, the first and second targets are first identified in the panoramic image separately, and then the presence of an overlapping area is determined based on the position information of the two targets. The pixel separation coefficient of the overlapping area is then accurately identified, and the actual pixel information of the first target (first pixel information) and the actual pixel information of the second target (second pixel information) are obtained under the pixel separation coefficient. The spatial coordinates of each pixel point in the two targets are then accurately calculated based on the actual pixel information of the two targets, and then it is accurately determined whether the two targets are at risk of near-electrical operations.

[0012] Furthermore, the step of determining whether there is an overlapping area between the first target and the second target based on the position information of the first target and the position information of the second target includes:

[0013] respectively obtaining a first area and a second area of ​​a region where the first target is located and a second area where the second target is located, and obtaining an intersection-over-union ratio of the first target and the second target in the panoramic image based on the first area and the second area;

[0014] Determining whether the intersection-over-union ratio is greater than a first preset threshold;

[0015] If the intersection-over-union ratio is greater than a first preset threshold, it is determined that there is an overlapping area between the first target and the second target.

[0016] Furthermore, if there is an overlapping area between the first target and the second target, the step of segmenting the overlapping area from the panoramic image and inputting the overlapping area into a pre-trained regression model to obtain a pixel separation coefficient includes:

[0017] Acquire a first image containing a single first object and a second image containing a single second object, extract third pixel information and third pixel coordinates of each pixel point of the first object in the first image, and extract fourth pixel information and fourth pixel coordinates of each pixel point of the second object in the second image;

[0018] randomly fusing pixel points in the first image with pixel points in the second image according to the third pixel coordinates and the fourth pixel coordinates to obtain a fused image, wherein the size of the fused image is equal to that of the first image and the second image;

[0019] Extracting fused pixel information of each pixel point in the fused image, and calculating a pixel separation coefficient based on the fused pixel information and third pixel information and fourth pixel information corresponding to the fused pixel information;

[0020] The fused pixel information of each pixel in the fused image and the pixel separation coefficient corresponding to each fused pixel information are input into the regression model to be trained to obtain a pre-trained regression model.

[0021] Furthermore, the step of performing pixel separation on the overlapping area according to the pixel separation coefficient, and obtaining first pixel information of each pixel point in the area covered by the first target and first coordinate information corresponding to the first pixel information, and second pixel information of each pixel point in the area covered by the second target and second coordinate information corresponding to the second pixel information according to the pixel separation result includes:

[0022] Retrieving first average pixel information and second average pixel information from a preset database according to pixel information of each overlapping pixel point in the overlapping area;

[0023] The first pixel information and the second pixel information are obtained according to the pixel information of each overlapping pixel point, the pixel separation coefficient corresponding to each overlapping pixel point, the first average pixel information, and the second average pixel information.

[0024] Furthermore, the steps of building a preset database include:

[0025] Extracting fused pixel information of each fused point in the fused image, defining any fused point as a target fused point, obtaining a first set of non-fused pixels excluding the target fused point in an area covered by the first object in the first image, and a second set of non-fused pixels excluding the target fused point in an area covered by the second object in the second image;

[0026] Obtaining third average pixel information based on third pixel information of all pixels included in the first unfused pixel set, and obtaining fourth average pixel information based on fourth pixel information of all pixels included in the second unfused pixel set;

[0027] The fused pixel information is associated with the corresponding third average pixel information and fourth pixel information one by one, and a preset database is obtained according to the association results.

[0028] Furthermore, the step of obtaining the first pixel information and the second pixel information according to the pixel information of each overlapping pixel point, the pixel separation coefficient corresponding to each overlapping pixel point, the first average pixel information, and the second average pixel information includes:

[0029] Get the first pixel information or the second pixel information according to the following formula:

[0030] ;

[0031] in, Represents the pixel information of the i-th overlapping pixel point, Respectively represent the coordinate information of the i-th overlapping pixel point, represents the pixel separation coefficient corresponding to the i-th overlapping pixel point, represents the first pixel information or the second pixel information corresponding to the i-th overlapping pixel point, Represents the first average pixel information or the second average pixel information corresponding to the i-th overlapping pixel point.

[0032] Furthermore, the step of obtaining the first spatial coordinates of each pixel point in the first target according to the first pixel information and the first coordinate information, and obtaining the second spatial coordinates of each pixel point in the second target according to the second pixel information and the second coordinate information includes:

[0033] Acquire a first depth value of each pixel point in the first target and a second depth value of each pixel point in the second target respectively according to the first pixel information and the second pixel information;

[0034] The first space coordinate or the second space coordinate is calculated according to the following formula:

[0035] ;

[0036] in, Respectively represent the x-coordinate, y-coordinate, and z-coordinate of the i-th pixel point in the first or second target in three-dimensional space, Respectively represent the x-coordinate and y-coordinate of the i-th pixel point in the first or second target in the panoramic image, Indicates the x-coordinate and y-coordinate of the principal point of the camera used to capture the panoramic image. Respectively represent the focal lengths of the camera used to capture the panoramic image in the x-axis and y-axis directions, Indicates the depth value of the i-th pixel in the first or second object in the panoramic image.

[0037] Furthermore, the step of determining whether there is a risk of near-electrical operation based on the first spatial coordinates and the second spatial coordinates includes:

[0038] Traversing the spatial distance between the pixel point in the first object and the pixel point in the second object according to the first spatial coordinate and the second spatial coordinate;

[0039] Determining whether the spatial distances are all greater than a second preset threshold;

[0040] If there is at least one spatial distance less than or equal to the second preset threshold, it is determined that the current near-electrical operation is risky;

[0041] If the spatial distances are all greater than the second preset threshold, it is determined that the current near-electrical operation is safe and compliant.

[0042] In a second aspect, the present invention provides a system for identifying risks of near-electrical operations, the system comprising:

[0043] a target recognition module, configured to obtain a panoramic image of an operator performing near-electrical work, and input the panoramic image into a first target recognition model and a second target recognition model, respectively, to obtain position information of the first target and the second target;

[0044] An overlapping area detection module, configured to determine whether there is an overlapping area between the first target and the second target based on the position information of the first target and the position information of the second target;

[0045] a pixel separation coefficient identification module, configured to segment the overlapping area from the panoramic image if there is an overlapping area between the first target and the second target, and input the overlapping area into a pre-trained regression model to obtain a pixel separation coefficient;

[0046] a pixel separation module, configured to perform pixel separation on the overlapping area according to the pixel separation coefficient, and obtain, based on the pixel separation result, first pixel information of each pixel point in the area covered by the first target and first coordinate information corresponding to the first pixel information, and second pixel information of each pixel point in the area covered by the second target and second coordinate information corresponding to the second pixel information;

[0047] A risk detection module is used to obtain the first spatial coordinates of each pixel point in the first target based on the first pixel information and the first coordinate information, obtain the second spatial coordinates of each pixel point in the second target based on the second pixel information and the second coordinate information, and determine whether there is a risk of near-electrical operation based on the first spatial coordinates and the second spatial coordinates.

[0048] In a third aspect, the present invention provides a readable storage medium, wherein the readable storage medium stores one or more programs, which, when executed by a processor, implement the above-mentioned method for identifying risks of near-electrical operations.

[0049] In a fourth aspect, the present invention provides an electronic device, comprising a memory and a processor, wherein:

[0050] The memory is used to store computer programs;

[0051] When the processor is used to execute the computer program stored in the memory, the above-mentioned method for identifying risks of near-electrical operations is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a flow chart of a method for identifying risks of near-electrical operations proposed in one embodiment of the present invention;

[0053] Figure 2 This is a schematic diagram of the structure of a risk identification system for near-electrical operations proposed in one embodiment of the present invention.

[0054] The following specific embodiments will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein should be the common meanings understood by people with ordinary skills in the field to which the invention belongs. The words "including" and similar words used in this article mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0056] like Figure 1 As shown, an embodiment of the present invention provides a method for identifying risks of near-electrical operations, the method comprising steps S101 to S105, wherein:

[0057] Step S101: Acquire a panoramic image of an operator performing near-electrical work, and input the panoramic image into a first target recognition model and a second target recognition model, respectively, to obtain position information of the first target and position information of the second target;

[0058] It should be noted that the goal of acquiring panoramic images is to include the complete first and second targets. The first and second target recognition models are trained using a large number of historical images that include the first and second targets. Before training, the first and second targets in the historical images are annotated with the complete first and second targets to create a training set. This is then input into the corresponding target recognition model for training, thereby obtaining the first and second target recognition models. Therefore, the position information of the identified first and second targets is complete.

[0059] It should also be pointed out that all position information in the present invention refers to coordinates. If it is the position information of a pixel point, it refers to the coordinates of the pixel point. If it is the position information of a target, it refers to the coordinates of all pixel points contained in the target. In addition, the first target and the second target are one of the operating personnel or the point-carrying equipment, and the first target and the second target are different.

[0060] Step S102: determining whether there is an overlapping area between the first target and the second target based on the location information of the first target and the location information of the second target;

[0061] It should be pointed out that in order to accurately identify whether there is an overlapping area between the first target and the second target in the panoramic image, it is necessary to first obtain the first area and the second area of ​​the area where the first target is located and the second area where the second target is located, respectively, and obtain the intersection-and-union ratio of the first target and the second target in the panoramic image based on the first area and the second area; determine whether the intersection-and-union ratio is greater than a first preset threshold; if the intersection-and-union ratio is greater than the first preset threshold, it is determined that there is an overlapping area between the first target and the second target. Specifically, the intersection-and-union ratio is obtained by dividing the intersection of the first area and the second area by the union of the first area and the second area. The value of the first preset threshold is set according to the specific usage scenario and is not limited in detail in this embodiment. For example, it can be 0.1, 0.2, etc.

[0062] In addition, if the intersection-over-union ratio is less than or equal to the first preset threshold, it means that there is no overlapping area or very little overlapping area between the first target and the second target. In this case, pixel separation is not required. In this case, the pixel information of the first target and the second target extracted are all their actual pixel values, and the depth values ​​of the pixel points of the first target and the second target can be directly obtained. The subsequent steps are the same as the processing steps of the first target and the second target after pixel separation is achieved, and will not be repeated in this embodiment.

[0063] Step S103: If there is an overlapping area between the first target and the second target, segment the overlapping area from the panoramic image, and input the overlapping area into a pre-trained regression model to obtain a pixel separation coefficient;

[0064] It should be pointed out that in the process of constructing a pre-trained regression model, a first image containing a single first target and a second image containing a single second target are first obtained, and the third pixel information and third pixel coordinates of each pixel point of the first target in the first image, and the fourth pixel information and fourth pixel coordinates of each pixel point of the second target in the second image are extracted; then, according to the third pixel coordinates and the fourth pixel coordinates, the pixels in the first image are randomly fused with the pixels in the second image to obtain a fused image, and the size of the fused image is equal to that of the first image and the second image; the fused pixel information of each pixel point in the fused image is extracted, and the pixel separation coefficient is calculated based on the fused pixel information and the third pixel information and the fourth pixel information corresponding to the fused pixel information; the fused pixel information of each pixel point in the fused image and the pixel separation coefficient corresponding to each fused pixel information are input into the regression model to be trained to obtain a pre-trained regression model.

[0065] It should be pointed out that in this step, theoretically, during the pixel fusion process, the formula should be satisfied: In the training process of the regression model, since the fused image is obtained by fusing the first image and the second image with known relevant pixel information, the pixel separation coefficient can be accurately calculated based on this. In addition, since the pixels in the overlapping area are derived from the overlap of different parts of the first target and the second target, in the actual operation process, the appearance color of the first target and the appearance color of the second target are generally very different, and the shape of the overlapping area may be extremely irregular, which may further amplify the difference between the pixel part belonging to the first target and the pixel part belonging to the second target in the overlapping area. Based on this, in order to further improve the pixel separation accuracy of the overlapping area, in the training of the regression model, the actual pixel separation coefficient is calculated for each pixel in the fused image separately, instead of using the conventional method of using a universal pixel separation coefficient to adapt to the entire fused area or the overlapping area.

[0066] Step S104: performing pixel separation on the overlapping area according to the pixel separation coefficient, and obtaining first pixel information of each pixel point in the area covered by the first target and first coordinate information corresponding to the first pixel information, and second pixel information of each pixel point in the area covered by the second target and second coordinate information corresponding to the second pixel information according to the pixel separation result;

[0067] It should be noted that in this step, the first average pixel information and the second average pixel information are retrieved from a preset database based on the pixel information of each overlapping pixel in the overlapping area; the first pixel information and the second pixel information are obtained based on the pixel information of each overlapping pixel, the pixel separation coefficient corresponding to each overlapping pixel, the first average pixel information, and the second average pixel information. As for the first coordinate information and the second coordinate information, for non-overlapping areas, they can be obtained from the respective position information; for overlapping areas, they can be obtained from the pixel coordinates of each pixel in the overlapping area.

[0068] Furthermore, it is important to note that during the pixel separation process, while the pixel separation coefficient for each pixel in the overlapping area is accurately identified, the actual pixel information of the first and second targets in the overlapping area is unknown. Therefore, this step introduces the concepts of non-overlapping areas and prior pixels. Due to the uncertainty of overlapping areas in panoramic images during actual shooting, if the non-overlapping areas in the panoramic image are used as the source of average pixel information, on the one hand, assuming that the non-overlapping areas are small, resulting in a small number of covered pixels, the representativeness of the average pixel information will inevitably be affected. On the other hand, the non-overlapping areas in the panoramic image are unique, while the overlapping areas require a large number of pixels to be pixel-separated. Using the average pixel information obtained from the unique non-overlapping area as the prior pixel for all pixels in the overlapping area will be affected by pixel position and the actual working environment, further reducing the representativeness of the average pixel information, resulting in a very large error in the actual pixel information of the first and second targets finally restored.

[0069] Based on this, in some embodiments, the fusion pixel information of each fusion point in the fusion image is extracted, any fusion point is defined as a target fusion point, and a first non-fusion pixel set excluding the target fusion point in the area covered by the first target in the first image and a second non-fusion pixel set excluding the target fusion point in the area covered by the second target in the second image are obtained; the third average pixel information is obtained based on the third pixel information of all pixel points contained in the first non-fusion pixel set, and the fourth average pixel information is obtained based on the fourth pixel information of all pixel points contained in the second non-fusion pixel set; the fusion pixel information is respectively associated one by one with the corresponding third average pixel information and fourth pixel information, and a preset database is obtained based on the association results.

[0070] In some embodiments, the first pixel information or the second pixel information is obtained specifically according to the following formula:

[0071] ;

[0072] in, Represents the pixel information of the i-th overlapping pixel point, Respectively represent the coordinate information of the i-th overlapping pixel point, represents the pixel separation coefficient corresponding to the i-th overlapping pixel point, represents the first pixel information or the second pixel information corresponding to the i-th overlapping pixel point, Represents the first average pixel information or the second average pixel information corresponding to the i-th overlapping pixel point.

[0073] Step S105: Obtain the first spatial coordinates of each pixel point in the first target based on the first pixel information and the first coordinate information, obtain the second spatial coordinates of each pixel point in the second target based on the second pixel information and the second coordinate information, and determine whether there is a risk of near-electrical operation based on the first spatial coordinates and the second spatial coordinates.

[0074] It should be noted that, in order to obtain the relevant spatial coordinates, it is first necessary to obtain the first depth value of each pixel point in the first target and the second depth value of each pixel point in the second target respectively according to the first pixel information and the second pixel information;

[0075] The first space coordinate or the second space coordinate is calculated according to the following formula:

[0076] ;

[0077] in, Respectively represent the x-coordinate, y-coordinate, and z-coordinate of the i-th pixel point in the first or second target in three-dimensional space, Respectively represent the x-coordinate and y-coordinate of the i-th pixel point in the first or second target in the panoramic image, Indicates the x-coordinate and y-coordinate of the principal point of the camera used to capture the panoramic image. Respectively represent the focal lengths of the camera used to capture the panoramic image in the x-axis and y-axis directions, Indicates the depth value of the i-th pixel in the first or second object in the panoramic image.

[0078] In addition, the depth value is obtained by identifying a first monocular depth model and a second monocular depth model. The first monocular depth model is trained by an image containing known pixel information and corresponding depth values ​​of the first target, and the second monocular depth model is trained by an image containing known pixel information and corresponding depth values ​​of the second target.

[0079] In addition, the spatial distance between the pixel points in the first target and the pixel points in the second target is traversed based on the first spatial coordinate and the second spatial coordinate; it is determined whether the spatial distances are all greater than the second preset threshold; if there is at least one spatial distance less than or equal to the second preset threshold, it is determined that the current near-electrical operation is risky, and an alarm message will be sent to the cloud service platform or the operation management platform within the first preset time to remind the operator that there is an operation risk. The platform pre-stores relevant information of the operator, such as telephone number, etc., and can then remotely issue an alarm reminder to the operator; if the spatial distances are all greater than the second preset threshold, it is determined that the current near-electrical operation is safe and compliant. The second preset threshold and the first preset time are related to specific usage requirements or application scenarios and are not specified in detail in this embodiment.

[0080] In summary, according to the above-mentioned method for identifying risks of near-electrical operations, the first and second targets in the overlapping area are accurately segmented to restore the pixel information originally belonging to the first and second targets, and then the spatial distance between the two is accurately calculated, thereby greatly improving the accuracy of identifying the current risk of near-electrical operations. Specifically, the first and second targets are first identified in the panoramic image separately, and then the presence of an overlapping area is determined based on the position information of the two targets. The pixel separation coefficient of the overlapping area is then accurately identified, and the actual pixel information of the first target (first pixel information) and the actual pixel information of the second target (second pixel information) are obtained under the pixel separation coefficient. The spatial coordinates of each pixel point in the two targets are then accurately calculated based on the actual pixel information of the two targets, and then it is accurately determined whether the two targets are at risk of near-electrical operations.

[0081] like Figure 2 One embodiment of the present invention further provides a risk identification system for near-electrical operations, the system comprising:

[0082] The target recognition module 10 is used to obtain a panoramic image of the operator when performing near-electrical work, and input the panoramic image into the first target recognition model and the second target recognition model respectively to obtain position information about the first target and the second target;

[0083] An overlapping area detection module 20 is configured to determine whether there is an overlapping area between the first target and the second target based on the position information of the first target and the position information of the second target;

[0084] a pixel separation coefficient identification module 30 for segmenting the overlapping area from the panoramic image if there is an overlapping area between the first target and the second target, and inputting the overlapping area into a pre-trained regression model to obtain a pixel separation coefficient;

[0085] a pixel separation module 40 configured to perform pixel separation on the overlapping area according to the pixel separation coefficient, and obtain, based on the pixel separation result, first pixel information of each pixel point in the area covered by the first target and first coordinate information corresponding to the first pixel information, and second pixel information of each pixel point in the area covered by the second target and second coordinate information corresponding to the second pixel information;

[0086] The risk detection module 50 is used to obtain the first spatial coordinates of each pixel point in the first target based on the first pixel information and the first coordinate information, obtain the second spatial coordinates of each pixel point in the second target based on the second pixel information and the second coordinate information, and determine whether there is a risk of near-electrical operation based on the first spatial coordinates and the second spatial coordinates.

[0087] On the other hand, the present invention further provides a readable storage medium having one or more programs stored thereon, which, when executed by a processor, implements the above-mentioned method for identifying risks of near-electrical operations.

[0088] On the other hand, the present invention further proposes an electronic device, including a memory and a processor, wherein the memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory to implement the above-mentioned method for identifying risks of near-electrical operations.

[0089] Those skilled in the art will appreciate that the logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device), or in conjunction with such instruction execution system, apparatus, or device. For purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transmit a program for use by an instruction execution system, apparatus, or device, or in conjunction with such instruction execution system, apparatus, or device.

[0090] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting, or processing it in another suitable manner as necessary, and then storing it in a computer memory.

[0091] It should be understood that various components of the present invention may be implemented using hardware, software, firmware, or a combination thereof. In the aforementioned embodiments, multiple steps or methods may be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one or a combination of the following technologies known in the art may be used: a discrete logic circuit having logic gate circuits for implementing logic functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0092] While the embodiments of the present invention have been described in detail above, it will be apparent to those skilled in the art that various modifications and variations of these embodiments are possible. However, it should be understood that such modifications and variations are within the scope and spirit of the present invention as set forth in the claims. Furthermore, the invention described herein is susceptible to other embodiments and may be practiced or implemented in a variety of ways.

Claims

1. A method for identifying risks of working near electricity, characterized in that: The method comprises: Acquire a panoramic image of an operator performing near-electrical work, and input the panoramic image into a first target recognition model and a second target recognition model, respectively, to obtain position information about the first target and the second target; Determining whether there is an overlapping area between the first target and the second target based on the position information of the first target and the position information of the second target; If there is an overlapping area between the first target and the second target, segmenting the overlapping area from the panoramic image, and inputting the overlapping area into a pre-trained regression model to obtain a pixel separation coefficient; Performing pixel separation on the overlapping area according to the pixel separation coefficient, and obtaining first pixel information of each pixel point in the area covered by the first target and first coordinate information corresponding to the first pixel information, and second pixel information of each pixel point in the area covered by the second target and second coordinate information corresponding to the second pixel information according to the pixel separation result; The first spatial coordinates of each pixel point in the first target are obtained based on the first pixel information and the first coordinate information, the second spatial coordinates of each pixel point in the second target are obtained based on the second pixel information and the second coordinate information, and whether there is a risk of near-electrical operation is determined based on the first spatial coordinates and the second spatial coordinates.

2. The method for identifying risks of near-electrical operations according to claim 1, characterized in that: The step of determining whether there is an overlapping area between the first target and the second target based on the position information of the first target and the position information of the second target includes: respectively obtaining a first area and a second area of ​​a region where the first target is located and a second area where the second target is located, and obtaining an intersection-over-union ratio of the first target and the second target in the panoramic image based on the first area and the second area; Determining whether the intersection-over-union ratio is greater than a first preset threshold; If the intersection-over-union ratio is greater than a first preset threshold, it is determined that there is an overlapping area between the first target and the second target.

3. The method for identifying risks of near-electrical work according to claim 2, characterized in that: If there is an overlapping area between the first target and the second target, the step of segmenting the overlapping area from the panoramic image and inputting the overlapping area into a pre-trained regression model to obtain a pixel separation coefficient includes: Acquire a first image containing a single first object and a second image containing a single second object, extract third pixel information and third pixel coordinates of each pixel point of the first object in the first image, and extract fourth pixel information and fourth pixel coordinates of each pixel point of the second object in the second image; randomly fusing pixel points in the first image with pixel points in the second image according to the third pixel coordinates and the fourth pixel coordinates to obtain a fused image, wherein the size of the fused image is equal to that of the first image and the second image; Extracting fused pixel information of each pixel point in the fused image, and calculating a pixel separation coefficient based on the fused pixel information and third pixel information and fourth pixel information corresponding to the fused pixel information; The fused pixel information of each pixel in the fused image and the pixel separation coefficient corresponding to each fused pixel information are input into the regression model to be trained to obtain a pre-trained regression model.

4. The method for identifying risks of near-electrical work according to claim 3, characterized in that: The step of performing pixel separation on the overlapping area according to the pixel separation coefficient, and obtaining first pixel information of each pixel point in the area covered by the first target and first coordinate information corresponding to the first pixel information, and second pixel information of each pixel point in the area covered by the second target and second coordinate information corresponding to the second pixel information according to the pixel separation result includes: Retrieving first average pixel information and second average pixel information from a preset database according to pixel information of each overlapping pixel point in the overlapping area; The first pixel information and the second pixel information are obtained according to the pixel information of each overlapping pixel point, the pixel separation coefficient corresponding to each overlapping pixel point, the first average pixel information, and the second average pixel information.

5. The method for identifying risks of near-electrical work according to claim 4, characterized in that: The steps to build a preset database include: Extracting fused pixel information of each fused point in the fused image, defining any fused point as a target fused point, obtaining a first set of non-fused pixels excluding the target fused point in an area covered by the first object in the first image, and a second set of non-fused pixels excluding the target fused point in an area covered by the second object in the second image; Obtaining third average pixel information based on third pixel information of all pixels included in the first unfused pixel set, and obtaining fourth average pixel information based on fourth pixel information of all pixels included in the second unfused pixel set; The fused pixel information is associated with the corresponding third average pixel information and fourth pixel information one by one, and a preset database is obtained according to the association results.

6. The method for identifying risks of near-electrical work according to claim 4, characterized in that: The step of obtaining the first pixel information and the second pixel information according to the pixel information of each overlapping pixel point, the pixel separation coefficient corresponding to each overlapping pixel point, the first average pixel information, and the second average pixel information comprises: Get the first pixel information or the second pixel information according to the following formula: ； in, Represents the pixel information of the i-th overlapping pixel point, Respectively represent the coordinate information of the i-th overlapping pixel point, represents the pixel separation coefficient corresponding to the i-th overlapping pixel point, represents the first pixel information or the second pixel information corresponding to the i-th overlapping pixel point, Represents the first average pixel information or the second average pixel information corresponding to the i-th overlapping pixel point.

7. The method for identifying risks of near-electrical work according to claim 6, characterized in that: The step of obtaining the first spatial coordinates of each pixel point in the first target according to the first pixel information and the first coordinate information, and obtaining the second spatial coordinates of each pixel point in the second target according to the second pixel information and the second coordinate information comprises: Acquire a first depth value of each pixel point in the first target and a second depth value of each pixel point in the second target respectively according to the first pixel information and the second pixel information; The first space coordinate or the second space coordinate is calculated according to the following formula: ； in, Respectively represent the x-coordinate, y-coordinate, and z-coordinate of the i-th pixel point in the first or second target in three-dimensional space, Respectively represent the x-coordinate and y-coordinate of the i-th pixel point in the first or second target in the panoramic image, Indicates the x-coordinate and y-coordinate of the principal point of the camera used to capture the panoramic image. Respectively represent the focal lengths of the camera used to capture the panoramic image in the x-axis and y-axis directions, Indicates the depth value of the i-th pixel in the first or second object in the panoramic image.

8. The method for identifying risks of near-electrical work according to claim 6, characterized in that: The step of determining whether there is a risk of near-electrical operation based on the first spatial coordinates and the second spatial coordinates includes: Traversing the spatial distance between the pixel point in the first object and the pixel point in the second object according to the first spatial coordinate and the second spatial coordinate; Determining whether the spatial distances are all greater than a second preset threshold; If there is at least one spatial distance less than or equal to the second preset threshold, it is determined that the current near-electrical operation is risky; If the spatial distances are all greater than the second preset threshold, it is determined that the current near-electrical operation is safe and compliant.

9. A risk identification system for near-electrical operations, characterized in that: The system comprises: a target recognition module, configured to obtain a panoramic image of an operator performing near-electrical work, and input the panoramic image into a first target recognition model and a second target recognition model, respectively, to obtain position information of the first target and the second target; An overlapping area detection module, configured to determine whether there is an overlapping area between the first target and the second target based on the position information of the first target and the position information of the second target; a pixel separation coefficient identification module, configured to segment the overlapping area from the panoramic image if there is an overlapping area between the first target and the second target, and input the overlapping area into a pre-trained regression model to obtain a pixel separation coefficient; a pixel separation module, configured to perform pixel separation on the overlapping area according to the pixel separation coefficient, and obtain, based on the pixel separation result, first pixel information of each pixel point in the area covered by the first target and first coordinate information corresponding to the first pixel information, and second pixel information of each pixel point in the area covered by the second target and second coordinate information corresponding to the second pixel information; A risk detection module is used to obtain the first spatial coordinates of each pixel point in the first target based on the first pixel information and the first coordinate information, obtain the second spatial coordinates of each pixel point in the second target based on the second pixel information and the second coordinate information, and determine whether there is a risk of near-electrical operation based on the first spatial coordinates and the second spatial coordinates.

10. A readable storage medium, characterized in that: The readable storage medium stores one or more programs, which, when executed by a processor, implement the method for identifying risks of near-electrical operations according to any one of claims 1 to 8.

Images