A method for monitoring cross-boundary violations during on-site operations at a substation

Abstract

The present invention discloses a method for monitoring cross-border violations in on-site operations of a substation, comprising: acquiring image data of the substation operation site through a mobile terminal camera; identifying the image data according to a pre-built mobile terminal detection model, and acquiring detection frames of a triangular cone, an operator and a safety helmet in the image data; performing perimeter cross-border monitoring according to the triangular cone and the detection frame of the operator; acquiring the real height of the safety helmet based on RTK positioning technology; performing elevation cross-border monitoring according to the real height and detection frame of the safety helmet and the detection frame of the operator; the present invention can not only effectively identify ground violations, but also identify elevation cross-border violations, and has flexible deployment, simple operation, strong real-time performance and good reusability.

Description

Technical Field

[0001] The invention relates to a method for monitoring cross-boundary violations in on-site operations of a transformer substation, and belongs to the technical field of intelligent transformer substations. Background Art

[0002] The substation site environment is complex and there are many factors that threaten human life. During the construction of the substation, it is necessary to quickly and accurately detect and warn the workers who are about to enter the dangerous area in time to reduce the loss of life and property. However, most of the existing dangerous area detection methods are based on sensors, which require the deployment of a large number of sensors. The actual operation is complicated, and the sensors attached to the workers will interfere with the workers' construction operations. Manual observation is time-consuming and laborious, and is easily affected by the observer's mental state. With the development of artificial intelligence, many researchers have begun to combine deep learning technology with the problem of dangerous area detection. However, the existing technology mainly relies on a large amount of labeled data for model training, and then performs detection and identification at the back end. It does not make full use of the computing resources of the front end, consumes too much data transmission, and reduces applicability.

[0003] Traditional video surveillance systems are mainly based on manual monitoring and post-event video observation for evidence collection. In addition, the architecture design of the video surveillance system is relatively simple, and it cannot perform real-time intelligent analysis of video images, cannot achieve remote monitoring, and cannot promptly alarm for abnormal events. Summary of the invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a method for monitoring cross-border violations in substation field operations, which can not only effectively identify ground violations, but also identify elevation cross-border violations, and has flexible deployment, simple operation, strong real-time performance and good reusability.

[0005] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0006] The present invention provides a method for monitoring transboundary violations in substation field operations, comprising:

[0007] Obtain image data of the substation operation site through mobile cameras;

[0008] The image data is identified according to the pre-built mobile detection model to obtain the detection frames of the pyramid, the operator, and the helmet in the image data;

[0009] Perimeter crossing monitoring is performed based on the detection frame of the triangular cone and the operator;

[0010] Get the real height of the helmet based on RTK positioning technology;

[0011] Elevation out-of-bounds monitoring is performed based on the actual height of the safety helmet and the detection frame as well as the detection frame of the operator.

[0012] Optionally, the construction of the mobile terminal monitoring model includes:

[0013] Convert the YOLOv5 model trained in PyTorch to an ONNX model;

[0014] Convert the ONNX model to an NCNN model and use it as a mobile monitoring model;

[0015] The training of the YOLOv5 model includes:

[0016] Obtain image information of the substation operation site and add detection frame labels of triangular cones and operators;

[0017] Generate sample images and construct a sample set based on the image information with detection frame labels added;

[0018] Divide the sample set into training set and validation set;

[0019] Initialize the YOLOv5 model and input the sample images in the training set into the YOLOv5 model for training;

[0020] The sample images in the validation set are input into the trained YOLOv5 model for verification. If the preset requirements are met, the training is completed.

[0021] Optionally, the perimeter crossing monitoring based on the detection frame of the triangular cone and the operator includes:

[0022] The polygonal area enclosed by the inner end points of the bottom of the triangular cone detection frame is regarded as the safe area, and the area outside the polygonal area is the dangerous area;

[0023] Take the coordinates of the lower left and lower right corners of the operator detection frame as detection points and obtain the projection points of the detection points on the ground;

[0024] Based on the PNpoly algorithm, it is determined whether the two projection points are located in a safe area or a polygonal area. If both projection points are located in a safe area, the staff has not crossed the boundary on the ground; if both projection points are located in a dangerous area, the staff has crossed the boundary on the ground; if the two projection points are located in a safe area and a dangerous area respectively, the staff is crossing the boundary on the ground.

[0025] Optionally, there are four triangular cones, which are respectively denoted as A, B, C, and D; a rectangular area ABCD surrounded by lines connecting the inner end points of the bottom of the detection frames of the triangular cones A, B, C, and D is used as the safety area.

[0026] Optionally, obtaining a projection point of the detection point on the ground includes:

[0027] Obtain the detection frame of the triangular cone and the operator in the image data of the AC connection side;

[0028] Draw a projection line and an extension line of the projection line from the bottom midpoint of the operator p detection frame to the ground. The intersection point of the projection line and the AC line is marked as M, and the intersection point of the extension line and the BD line is marked as N.

[0029] The position ratio of the points M and N in the AC line side image data on the AC line and the BD line is calculated based on the triangular cone in the AC line side image data and the detection frame of the operator;

[0030] Obtain the detection frame of the triangular cone and the operator in the image data on the AB line side;

[0031] According to the position ratio of the points M and N on the AC line and the BD line in the image data on the AC line side, the positions of the points M and N on the AC line and the BD line in the image data on the AB line side are obtained;

[0032] Set virtual pyramids E and F based on the positions of points M and N on the AC and BD lines in the image data on the AB line side, and the lower right corner horizontal coordinate of the virtual pyramid E detection frame coincides with point M, and the lower left corner horizontal coordinate of the virtual pyramid F detection frame coincides with point N;

[0033] Calculate the coordinates of points M and N in the image data on the AB line side according to the triangular cone in the image data on the AB line side and the detection frame of the operator;

[0034] The projection point of the bottom midpoint of the operator p detection frame in the AB line side image data on the ground is recorded as point H;

[0035] Calculate the coordinates of point H in the image data on the AB line side according to the coordinates of points M and N in the image data on the AB line side;

[0036] The coordinates of the projection points on the ground of the lower left and lower right corners of the detection frame of the operator p in the AB line side image data are calculated according to the coordinates of point H in the AB line side image data.

[0037] Optionally, calculating the position ratio of points M and N on the AC line and the BD line in the AC line side image data includes:

[0038] Get the horizontal coordinates of the upper left corner and lower right corner of the worker p detection frame in the AC connection side image data Determine the horizontal coordinates x of points M and N in the image data on the AC connection side M 、x N :

[0039]

[0040] Get the horizontal coordinate of the upper left corner of the detection box of the triangular cone A and B in the AC line side image data The horizontal coordinate of the lower right corner of the detection box of the triangular cone C and D Determine the position ratios K1 and K2 of points M and N on the AC line and the BD line in the AC line side image data:

[0041]

[0042] Optionally, calculating the coordinates of points M and N in the image data on the side of the AB line includes:

[0043] Get the ordinate of the upper left corner of the detection frame of the triangular cone A and C in the image data on the AB line side The ordinate of the lower right corner of the detection box of the triangular cone A and C Determine the height h of the detection box of the triangular cone E E :

[0044]

[0045]

[0046] Get the ordinate of the upper left corner of the detection frame of the triangular pyramid B and D in the image data on the AB line side The ordinate of the lower right corner of the detection box of the triangular cone B and D Determine the height h of the detection frame of the triangular cone F F :

[0047]

[0048]

[0049] Determine the position ratio K of point M on the AC line in the image data on the AB line side M :

[0050]

[0051] Determine the position ratio K of point N on the BD line in the image data on the AB line side N :

[0052]

[0053] Get the horizontal coordinate of the lower right corner of the detection box of the triangular cone A and C in the image data on the AB line side Determine the coordinates (x M ,y M ):

[0054]

[0055]

[0056] Get the horizontal coordinate of the upper left corner of the detection box of the triangular pyramid B and D in the image data on the AB line side Determine the coordinates (x N ,y N ):

[0057]

[0058]

[0059] Optionally, calculating the coordinates of point H in the image data on the side of the AB line includes:

[0060] Get the upper left and lower right coordinates of the detection box of worker p in the image data on the AB line side The ordinate of the lower right corner of the detection box of operator p Determine the horizontal coordinate x of point H in the image data on the AB line side H :

[0061]

[0062] Get the horizontal coordinates x of points M and N in the image data on the AB line side M 、x N , determine the position ratio K of point H on the MN line in the image data on the AB line side H :

[0063]

[0064] Get the y coordinates of points M and N in the image data on the AB line side M ,y N , determine the coordinates of point H in the image data on the AB line side (x H ,y H ):

[0065]

[0066] Optionally, the coordinates of the projection point on the ground of the lower left corner of the detection frame of the operator p in the image data on the AB line side are: The coordinates of the projection point of the lower right corner on the ground are

[0067] Optionally, the performing elevation crossing monitoring according to the real height and the detection frame and the detection frame of the operator includes:

[0068] In the image data on the AB line side, the center point coordinates of the helmet detection frame are used as the reference point to obtain the projection point of the reference point on the ground;

[0069] According to the ordinates of the reference point and the projection point of the reference point on the ground in the image data on the AB line side, the distance from the reference point to the projection point of the reference point on the ground is obtained, which is recorded as h Q ;

[0070] Combined with the actual height H of the helmet R , determine the proportional coefficient K between the image data on the AB side and the real data:

[0071]

[0072] Get the vertical coordinate of the top midpoint of the detection frame of operator p in the image data on the AB line side The vertical coordinate y of point H in the image data on the AB line side H , determine the distance h from the top midpoint of the detection frame of operator p to the ground p :

[0073]

[0074] Combine the ratio coefficient K between the image data on the AB side and the real data to determine the real height H from the top midpoint of the detection frame of the operator p to the ground. p :

[0075]

[0076] If the true height H p If the actual height H is greater than or equal to the preset out-of-bounds height, the staff member's altitude is out of bounds; p If the height is less than the preset crossing height, the staff member has not crossed the boundary.

[0077] Compared with the prior art, the present invention has the following beneficial effects:

[0078] The present invention provides a method for monitoring cross-border violations in substation field operations, which recognizes image data through a pre-built mobile detection model, and has flexible deployment and simple and convenient operation. Different from traditional physical fences and other types of virtual electronic fence technologies, it can not only effectively identify ground violations, but also identify high-altitude cross-border violations, and has strong real-time performance and good reusability. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 It is a flow chart of a method for monitoring transboundary violations in substation field operations provided by an embodiment of the present invention;

[0080] Figure 2 is a schematic diagram of substation field operation provided by an embodiment of the present invention;

[0081] Figure 3It is a schematic diagram of image data of AC line side of substation field operation provided by an embodiment of the present invention;

[0082] Figure 4 It is a schematic diagram of image data on the AB connection side of a substation on-site operation provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0083] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0084] like Figure 1 As shown, an embodiment of the present invention provides a method for monitoring cross-boundary violations in substation field operations, comprising the following steps:

[0085] 1. Obtain image data of the substation operation site through the mobile camera.

[0086] 2. Identify the image data based on the pre-built mobile detection model to obtain the detection frames of the pyramid, worker, and helmet in the image data.

[0087] 2.1、The construction of the mobile monitoring model includes:

[0088] Convert the YOLOv5 model trained in PyTorch to an ONNX model;

[0089] Convert the ONNX model to an NCNN model and use it as a mobile monitoring model;

[0090] 2.2. The training of the YOLOv5 model includes:

[0091] Obtain image information of the substation operation site and add detection frame labels of triangular cones and operators;

[0092] Generate sample images and construct a sample set based on the image information with detection frame labels added;

[0093] Divide the sample set into training set and validation set;

[0094] Initialize the YOLOv5 model and input the sample images in the training set into the YOLOv5 model for training;

[0095] The sample images in the validation set are input into the trained YOLOv5 model for verification. If the preset requirements are met, the training is completed.

[0096] 3. Conduct perimeter crossing monitoring based on the detection frame of the triangular cone and the operator; specifically include:

[0097] 3.1. The polygonal area enclosed by the inner end points of the bottom of the triangular cone detection frame is regarded as the safe area, and the area outside the polygonal area is the dangerous area;

[0098] 3.2. Take the coordinates of the lower left and lower right corners of the operator detection frame as detection points and obtain the projection points of the detection points on the ground;

[0099] 3.3. Based on the PNpoly algorithm, it is determined whether the two projection points are located in the safe area or the polygonal area. If both projection points are located in the safe area, the staff has not crossed the boundary on the ground; if both projection points are located in the dangerous area, the staff has crossed the boundary on the ground; if the two projection points are located in the safe area and the dangerous area respectively, the staff is crossing the boundary on the ground.

[0100] like Figure 2 As shown, during on-site operations at a substation, workers usually set up four triangular cones. This implementation takes four triangular cones as an example:

[0101] like Figure 3-4 As shown, there are four triangular cones, denoted as A, B, C, and D respectively; the rectangular area ABCD formed by the lines connecting the inner end points of the bottom of the detection frame of the triangular cones A, B, C, and D is used as the safe area. In this case, obtaining the projection point of the detection point on the ground includes:

[0102] S1. Obtain the detection frame of the triangular cone and the operator in the image data of the AC connection side (collected by camera S);

[0103] S2. Draw a projection line from the bottom midpoint of the operator p detection frame to the ground and the extension line of the projection line. The intersection point of the projection line and the AC line is marked as M, and the intersection point of the extension line and the BD line is marked as N;

[0104] S3, calculating the position ratio of points M and N on the AC line and the BD line in the AC line side image data according to the triangular cone in the AC line side image data and the detection frame of the operator; specifically including:

[0105] S3.1. Obtain the horizontal coordinates of the upper left corner and lower right corner of the detection frame of the operator p in the AC connection side image data Determine the horizontal coordinates x of points M and N in the image data on the AC connection side M 、x N :

[0106]

[0107] S3.2. Obtain the horizontal coordinate of the upper left corner of the detection frame of the triangular cone A and B in the AC line side image data The horizontal coordinate of the lower right corner of the detection box of the triangular cone C and D Determine the position ratios K1 and K2 of points M and N on the AC line and the BD line in the AC line side image data:

[0108]

[0109] S4, obtaining the triangular cone and the detection frame of the operator in the image data on the AB connection side (collected by camera F);

[0110] S5, obtaining the positions of the points M and N on the AC line and the BD line in the image data on the AB line side according to the position ratio of the points M and N on the AC line and the BD line in the image data on the AC line side;

[0111] S6, setting virtual pyramids E and F based on the positions of points M and N on the AC and BD lines in the image data on the AB line side, and the abscissa of the lower right corner of the virtual pyramid E detection frame coincides with point M, and the abscissa of the lower left corner of the virtual pyramid F detection frame coincides with point N;

[0112] S7, calculating the coordinates of points M and N in the image data on the AB line side according to the triangular cone in the image data on the AB line side and the detection frame of the operator; specifically including:

[0113] S7.1. Obtain the ordinate of the upper left corner of the detection frame of the triangular pyramid A and C in the image data on the AB line side The ordinate of the lower right corner of the detection box of the triangular cone A and C Determine the height h of the detection box of the triangular cone E E :

[0114]

[0115]

[0116] S7.2. Obtain the ordinate of the upper left corner of the detection frame of the triangular pyramid B and D in the image data on the AB line side The ordinate of the lower right corner of the detection box of the triangular cone B and D Determine the height h of the detection frame of the triangular cone F F :

[0117]

[0118]

[0119] S7.3. Determine the position ratio K of point M on line AC in the image data on the AB side of the line M :

[0120]

[0121] S7.4. Determine the position ratio K of point N on the BD line in the image data on the AB line side N :

[0122]

[0123] S7.5. Obtain the horizontal coordinate of the lower right corner of the detection frame of the triangular cone A and C in the image data on the AB line side Determine the coordinates (x M ,y M ):

[0124]

[0125]

[0126] S7.6. Obtain the horizontal coordinate of the upper left corner of the detection frame of the triangular pyramid B and D in the image data on the AB line side Determine the coordinates (x N ,y N ):

[0127]

[0128]

[0129] S8, recording the projection point of the bottom midpoint of the operator p detection frame in the AB line side image data on the ground as point H;

[0130] S9, calculating the coordinates of point H in the image data on the AB line side according to the coordinates of points M and N in the image data on the AB line side;

[0131] S9.1. Obtain the upper left corner and lower right corner horizontal coordinates of the detection frame of the operator p in the image data on the AB line side The ordinate of the lower right corner of the detection box of operator p Determine the horizontal coordinate x of point H in the image data on the AB line side H :

[0132]

[0133] S9.2. Obtain the horizontal coordinates x of points M and N in the image data on the AB line side M 、x N , determine the position ratio K of point H on the MN line in the image data on the AB line side H :

[0134]

[0135] S9.3. Obtain the ordinate y of points M and N in the image data on the AB line side M ,y N , determine the coordinates of point H in the image data on the AB line side (x H ,y H ):

[0136]

[0137] S10. Calculate the projection point coordinates of the lower left and lower right corners of the detection frame of the operator p in the AB line side image data on the ground according to the coordinates of point H in the AB line side image data.

[0138] The coordinates of the projection point on the ground of the lower left corner of the detection frame of worker p in the image data on the AB line are The coordinates of the projection point of the lower right corner on the ground are

[0139] 4. Get the real height of the helmet based on RTK positioning technology. RTK positioning technology mainly includes:

[0140] 1) Install the base station at the substation renovation and expansion construction site. Place the base station in a location with high terrain, open sky, and surrounding obstacles with an elevation angle of no more than 10°, which is conducive to satellite signal reception and data link transmission, and there are no signal reflectors around, and the radio antenna is set up at a high place;

[0141] 2) Place a mobile station, i.e., a user receiver, on the worker’s helmet;

[0142] 3) Use the 4G transparent transmission module to connect to the RTK mobile station to transmit the data calculated by the RTK mobile station to the cloud server through the 4G transparent transmission module;

[0143] 4) The acquired height information is transmitted to the height crossing detection module for height calibration.

[0144] 5. Perform elevation out-of-bounds monitoring based on the actual height of the helmet and the detection frame as well as the detection frame of the operator; specifically including:

[0145] In the image data on the AB line side, the center point coordinates of the helmet detection frame are used as the reference point to obtain the projection point of the reference point on the ground;

[0146] According to the ordinates of the reference point and the projection point of the reference point on the ground in the image data on the AB line side, the distance from the reference point to the projection point of the reference point on the ground is obtained, which is recorded as h Q ;

[0147] Combined with the actual height H of the helmet R , determine the proportional coefficient K between the image data on the AB side and the real data:

[0148]

[0149] Get the vertical coordinate of the top midpoint of the detection frame of operator p in the image data on the AB line side The vertical coordinate y of point H in the image data on the AB line side H , determine the distance h from the top midpoint of the detection frame of operator p to the ground p :

[0150]

[0151] Combine the ratio coefficient K between the image data on the AB side and the real data to determine the real height H from the top midpoint of the detection frame of the operator p to the ground. p :

[0152]

[0153] If the true height H p If the actual height H is greater than or equal to the preset out-of-bounds height, the staff member’s height is out of bounds; p If the height is less than the preset out-of-bounds height, the staff member has not crossed the boundary.

[0154] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0155] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for monitoring violations of substation field operations, characterized in that: include: Obtain image data of the substation operation site through mobile cameras; The image data is identified according to the pre-built mobile detection model to obtain the detection frames of the pyramid, the operator, and the helmet in the image data; Perimeter crossing monitoring is performed based on the detection frame of the triangular cone and the operator; Get the real height of the helmet based on RTK positioning technology; Perform elevation out-of-bounds monitoring based on the actual height of the helmet and the detection frame as well as the detection frame of the operator; The perimeter crossing monitoring based on the detection frame of the triangular cone and the operator includes: The polygonal area enclosed by the line connecting the inner end points of the bottom of the triangular cone detection frame is used as the safe area, and the area outside the polygonal area is the dangerous area; there are four triangular cones, which are respectively recorded as A, B, C, and D; the rectangular area ABCD enclosed by the line connecting the inner end points of the bottom of the detection frame of the triangular cones A, B, C, and D is used as the safe area; The coordinates of the lower left corner and the lower right corner of the operator detection frame are used as detection points to obtain the projection points of the detection points on the ground; the obtaining of the projection points of the detection points on the ground includes: Obtain the detection frame of the triangular cone and the operator in the image data of the AC connection side; Draw a projection line and an extension line of the projection line from the bottom midpoint of the operator p detection frame to the ground. The intersection point of the projection line and the AC line is marked as M, and the intersection point of the extension line and the BD line is marked as N. The position ratio of the points M and N in the AC line side image data on the AC line and the BD line is calculated based on the triangular cone in the AC line side image data and the detection frame of the operator; Obtain the detection frame of the triangular cone and the operator in the image data on the AB line side; According to the position ratio of the points M and N on the AC line and the BD line in the image data on the AC line side, the positions of the points M and N on the AC line and the BD line in the image data on the AB line side are obtained; Set virtual pyramids E and F based on the positions of points M and N on the AC and BD lines in the image data on the AB line side, and the lower right corner horizontal coordinate of the virtual pyramid E detection frame coincides with point M, and the lower left corner horizontal coordinate of the virtual pyramid F detection frame coincides with point N; Calculate the coordinates of points M and N in the image data on the AB line side according to the triangular cone in the image data on the AB line side and the detection frame of the operator; The projection point of the bottom midpoint of the operator p detection frame in the AB line side image data on the ground is recorded as point H; Calculate the coordinates of point H in the image data on the AB line side according to the coordinates of points M and N in the image data on the AB line side; Calculate the projection point coordinates of the lower left corner and the lower right corner of the detection frame of the operator p in the image data on the AB line side on the ground according to the coordinates of the point H in the image data on the AB line side; Based on the PNpoly algorithm, it is determined whether the two projection points are located in a safe area or a polygonal area. If both projection points are located in a safe area, the staff has not crossed the boundary on the ground; if both projection points are located in a dangerous area, the staff has crossed the boundary on the ground; if the two projection points are located in a safe area and a dangerous area respectively, the staff is crossing the boundary on the ground.

2. A method for monitoring violations of substation field operations according to claim 1, characterized in that: The construction of the mobile terminal monitoring model includes: Convert the YOLOv5 model trained in PyTorch to an ONNX model; Convert the ONNX model to an NCNN model and use it as a mobile monitoring model; The training of the YOLOv5 model includes: Obtain image information of the substation operation site and add detection frame labels of triangular cones and operators; Generate sample images and construct sample sets based on the image information with detection frame labels added; Divide the sample set into training set and validation set; Initialize the YOLOv5 model and input the sample images in the training set into the YOLOv5 model for training; The sample images in the validation set are input into the trained YOLOv5 model for verification. If the preset requirements are met, the training is completed.

3. A method for monitoring transboundary violations in substation field operations according to claim 1, characterized in that: The calculating of the position ratio of the points M and N on the AC line and the BD line in the AC line side image data comprises: Get the horizontal coordinates of the upper left corner and lower right corner of the worker p detection frame in the AC connection side image data Determine the horizontal coordinates x of points M and N in the image data on the AC connection side M 、x N : Get the horizontal coordinate of the upper left corner of the detection box of the triangular cone A and B in the AC line side image data The horizontal coordinate of the lower right corner of the detection box of the triangular cone C and D Determine the position ratios K1 and K2 of points M and N on the AC line and the BD line in the AC line side image data:

4. A method for monitoring transboundary violations in substation field operations according to claim 3, characterized in that: The calculation of the coordinates of points M and N in the image data on the side of the AB line includes: Get the ordinate of the upper left corner of the detection frame of the triangular cone A and C in the image data on the AB line side The ordinate of the lower right corner of the detection box of the triangular cone A and C Determine the height h of the detection box of the triangular cone E E : Get the ordinate of the upper left corner of the detection frame of the triangular pyramid B and D in the image data on the AB line side The ordinate of the lower right corner of the detection box of the triangular cone B and D Determine the height h of the detection frame of the triangular cone F F : Determine the position ratio K of point M on the AC line in the image data on the AB line side M : Determine the position ratio K of point N on the BD line in the image data on the AB line side N : Get the horizontal coordinate of the lower right corner of the detection box of the triangular pyramid A and C in the image data on the AB line side Determine the coordinates (x M ,y M ): Get the horizontal coordinate of the upper left corner of the detection box of the triangular pyramid B and D in the image data on the AB line side Determine the coordinates (x N ,y N ):

5. A method for monitoring transboundary violations in substation field operations according to claim 4, characterized in that: The calculation of the coordinates of point H in the image data on the side of the AB line includes: Get the upper left and lower right coordinates of the detection box of worker p in the image data on the AB line side The ordinate of the lower right corner of the detection box of operator p Determine the horizontal coordinate x of point H in the image data on the AB line side H : Get the horizontal coordinates x of points M and N in the image data on the AB line side M 、x N , determine the position ratio K of point H on the MN line in the image data on the AB line side H : Get the y coordinates of points M and N in the image data on the AB line side M ,y N , determine the coordinates of point H in the image data on the AB line side (x H ,y H ):

6. A method for monitoring transboundary violations during on-site substation operations according to claim 5, characterized in that: The coordinates of the projection point on the ground of the lower left corner of the detection frame of the operator p in the AB line side image data are The coordinates of the projection point of the lower right corner on the ground are 7. A method for monitoring transboundary violations during on-site substation operations according to claim 6, characterized in that: The monitoring of elevation crossing according to the real height, the detection frame and the detection frame of the operator includes: In the image data on the AB line side, the center point coordinates of the helmet detection frame are used as the reference point to obtain the projection point of the reference point on the ground; According to the ordinates of the reference point and the projection point of the reference point on the ground in the image data on the AB line side, the distance from the reference point to the projection point of the reference point on the ground is obtained, which is recorded as h Q ; Combined with the actual height H of the helmet R , determine the proportional coefficient K between the image data on the AB side and the real data: Get the vertical coordinate of the top midpoint of the detection frame of operator p in the image data on the AB line side The vertical coordinate y of point H in the image data on the AB line side H , determine the distance h from the top midpoint of the detection frame of operator p to the ground p : Combine the ratio coefficient K between the image data on the AB side and the real data to determine the real height H from the top midpoint of the detection frame of the operator p to the ground. p : If the true height H p If the actual height H is greater than or equal to the preset out-of-bounds height, the staff member's altitude is out of bounds; p If the height is less than the preset crossing height, the staff member has not crossed the boundary.

Images