Mobile hazard event awareness method and system

Abstract

The application provides a mobile hidden danger event sensing method and system, relates to the technical field of data processing, and comprises the following steps: determining a conductor adjacent danger area based on conductor projection position information of a power transmission conductor in a monitoring field of view and safety interval requirements of the power transmission conductor, and generating a candidate receiving point along the conductor adjacent danger area; generating operation side leading change information of a mobile hidden danger object according to an event stream output by an event camera, and determining an operation side receiving point according to a receiving relationship of the operation side leading change information and the candidate receiving point in the conductor adjacent danger area; generating a mobile hidden danger invasion window based on the operation side receiving point, generating evidence data and mobile hidden danger event reporting information. The application takes the part of the mobile hidden danger object that changes towards the conductor adjacent danger area first as a dynamic leading end, and determines an operation side receiving point corresponding to the change of the dynamic leading end in the conductor adjacent danger area, thereby avoiding the problems of a large number of alarms and a non-concentrated evidence range.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to methods and systems for sensing mobile potential incidents. Background Technology

[0002] With the increase in low-altitude operations around power transmission lines, drones are also approaching power lines more frequently. For drones performing sling operations, it's not just the drone itself that's near the power line; the suspended object also sways as the drone moves laterally, hovers, lifts, or lowers. In other words, when the suspended object sways, the potential risk location near the power line changes depending on the state of the suspended object.

[0003] Currently, power transmission line monitoring typically employs micro-capture devices or video monitoring equipment to collect images of the area surrounding the transmission lines, and then makes judgments based on the positional relationship between the drone itself or the overall aerial moving target in the images and the transmission lines. Some monitoring methods also establish fixed monitoring areas around the transmission lines; when the drone itself or the overall aerial moving target approaches the transmission lines or enters the fixed monitoring area, alarm information and image records are generated.

[0004] However, when a drone used for hoisting operations approaches a power transmission line, the first part to approach the line may not be the drone itself, but rather a localized part such as the hook, the edge of the suspended object, or the lower end of the hoisting rope, which may swing into place during the operation. If the alarm is still based on the drone itself, the entire moving target in the air, or a fixed monitoring area, it is easy to confuse ordinary approach with signs of impending wire contact, resulting in alarms being triggered as soon as the target approaches the line, leading to many false alarms. Furthermore, the actual hoisting part that first approaches the line is not considered as a key evidence point, resulting in inaccurate alarm evidence. Summary of the Invention

[0005] This application provides a method and system for sensing mobile potential hazards, in order to solve the problem that existing technologies usually use whether the drone itself or the entire mobile target is close to the wire as the basis for alarms, which easily mixes ordinary proximity status with actual wire tangling warning signs, resulting in a large number of alarms and a lack of focus in evidence collection.

[0006] The first aspect of this application provides a method for detecting mobile hazard events, including: Based on the projection position information of the transmission line in the monitoring field of view and the safety interval requirements of the transmission line, the dangerous area adjacent to the transmission line is determined, and candidate receiving points are generated along the dangerous area adjacent to the transmission line. Based on the event stream output by the event camera, operational-side leader change information of the moving hazard object is generated. The operational-side leader change information is used to characterize the operational-side change state of the dynamic leader end of the moving hazard object relative to the operational-side change state of the adjacent danger zone of the conductor. Based on the connection between the pilot change information on the work side and the candidate receiving points in the dangerous area adjacent to the conductor, the receiving points on the work side are determined. Based on the work-side receiving point, a mobile hazard encroachment window is generated, and evidence collection data and mobile hazard event reporting information are generated according to the mobile hazard encroachment window.

[0007] Optionally, in one possible implementation of the first aspect, determining the adjacent danger zone of the transmission line based on the transmission line projection position information in the monitoring field of view and the safety spacing requirements of the transmission line includes: Based on the endpoint position, direction change position, and preset segmentation interval of the current projection of the conductor in the conductor projection position information, multiple conductor projection points on the current projection of the conductor are determined; Connecting adjacent traverse projection points yields the traverse route, and the line connecting two adjacent traverse projection points is defined as a local segment of the traverse. Based on the safety spacing requirements of transmission lines, local danger zones are set up on the side of local sections of each line. The adjacent danger zone of the conductor is generated based on each of the local danger zones.

[0008] Optionally, in one possible implementation of the first aspect, the provision of local danger zones on the side of each local section of the transmission line based on safety spacing requirements includes: Any local segment of the conductor is identified as the target local segment, and the positions of the first and second ends of the target local segment are determined. The two opposite sides of the local segment of the target conductor in the monitoring field of view are respectively defined as the first local generation side and the second local generation side; The acceptance width is determined based on the aforementioned safety interval requirements; The target local generation side is determined from the first local generation side and the second local generation side, and the direction from the target conductor local segment to the target local generation side is taken as the target lateral direction; In the target lateral direction, the positions that are respectively distanced from the first segment end position and the second segment end position from the bearing width are determined as the first outer end point and the second outer end point; Connect the first outer endpoint and the second outer endpoint to obtain the local outer boundary; Enclosing the local segment of the target conductor, the first lateral boundary from the end position of the first segment to the first outer endpoint, the second lateral boundary from the end position of the second segment to the second outer endpoint, and the local outer boundary, a local danger zone corresponding to the local generation side of the target is obtained.

[0009] Optionally, in one possible implementation of the first aspect, generating candidate acceptance points along the conductor adjacent to the danger zone includes: The local segment of the conductor that shares a side with any of the aforementioned local danger zones is identified as the target local segment of the conductor, and the boundary in any of the aforementioned local danger zones that is opposite to the target local segment of the conductor is identified as the receiving side boundary. Starting from the midpoint of the local segment of the target conductor, a receiving lateral line is generated along the direction from the local segment of the target conductor toward the receiving lateral boundary; The intersection of the receiving side line and the receiving side boundary is determined as the boundary receiving position, and the boundary receiving position is used as the candidate receiving point.

[0010] Optionally, in one possible implementation of the first aspect, generating the operational-side precursor change information of the moving hazard object based on the event stream output by the event camera includes: Extract the change region of the main body event of the suspended UAV in the event stream; Based on the position of the changed region of the ontology event in the event flow, determine the candidate follow-up event regions in the event flow that are adjacent to the changed region of the ontology event; Identify candidate follower event regions in which there are continuous event portions between the candidate follower event regions and the main event change regions, and use the continuous event portions as hanging connection belts; Based on the relative changes of the suspension connection belt, the main body event change region, and the candidate follower event region in a continuous event stream segment, the suspension event change region in the candidate follower event region is determined, and the near-line boundary of the suspension event change region toward the adjacent danger area of ​​the conductor is determined. The dynamic leader is determined based on the near-line boundary, and the operational-side leader change information is generated based on the dynamic leader.

[0011] Optionally, in one possible implementation of the first aspect, determining the dynamic leader based on the nearline boundary includes: The relative swing direction of the nearline boundary with respect to the body event change region is determined based on the boundary position sequence of the nearline boundary in the continuous event flow segment and the body position sequence of the body event change region in the continuous event flow segment. Based on the relative swing direction and the side of the adjacent danger zone of the conductor relative to the near-line boundary, a local boundary segment is determined from the near-line boundary; The end portion of the local boundary segment located at the front end of the relative swing direction is identified as a candidate leader end; Among the candidate leader terminals, the candidate leader terminal whose extension line along the corresponding relative swing direction intersects the adjacent danger zone of the conductor and whose intersection point is the closest is selected as the dynamic leader terminal.

[0012] Optionally, in one possible implementation of the first aspect, generating the operation-side leader change information based on the dynamic leader includes: Based on the displacement change of the main event change region in the continuous event stream segment and the offset change of the suspended object event change region relative to the main event change region, the operation action change node is determined; Determine the position of the dynamic pilot end before the operation action change node and the position of the second pilot end after the operation action change node; Connect the first leader position and the second leader position to obtain a leader change segment, and take the direction from the first leader position to the second leader position as the leader direction; Starting from the position of the second leader end, a leader pointing line is generated along the leader direction; Based on the directional relationship between the pilot line and the adjacent danger zone of the conductor, the operational-side pilot change information is generated.

[0013] Optionally, in one possible implementation of the first aspect, generating the work-side leader change information based on the directional relationship of the leader line relative to the adjacent danger zone of the conductor includes: Based on the extension of the leading pointing line along the leading direction and the change in the distance between the leading pointing line and the adjacent danger zone of the conductor, the regional pointing state of the leading pointing line relative to the adjacent danger zone of the conductor is determined; the regional pointing state includes a pointing state and a moving away state. Based on the state transition relationship between the pointing state and the far-away state in the continuous event stream segment, and the persistence relationship of the pointing state, the operational side change state of the dynamic leader is determined. The operation-side leader change information is generated based on the operation-side change status of the dynamic leader terminal.

[0014] Optionally, in one possible implementation of the first aspect, determining the work-side receiving point based on the receiving relationship between the work-side pilot change information and the candidate receiving point in the adjacent danger zone of the conductor includes: Based on the change information of the working side leader, the working side of the dynamic leader relative to the dangerous area adjacent to the conductor is determined; The position where the pilot directional line enters the dangerous area adjacent to the conductor is determined as the entry position, and the directional receiving area is determined in the dangerous area adjacent to the conductor based on the entry position; Select the candidate receiving points located on the working side and corresponding to the pointing receiving area from the candidate receiving points to obtain the receiving candidate points; Based on the connection relationship between the candidate receiving points and the entry position, the receiving point on the work side is determined.

[0015] Optionally, in one possible implementation of the first aspect, determining the work-side acceptance point based on the acceptance relationship between the candidate acceptance point and the entry position includes: When the number of candidate sites for acceptance is determined to be a single point, the single candidate site for acceptance is determined as the acceptance site on the work side. When the number of the receiving candidate points is determined to be multiple, the multiple receiving candidate points are sorted according to the receiving distance between each receiving candidate point and the entry position along the leading direction to obtain a receiving point sequence; The candidate receiving point with the smallest receiving distance in the receiving point sequence is determined as the receiving point on the working side.

[0016] Optionally, in one possible implementation of the first aspect, generating the mobile hazard encroachment window based on the work-side acceptance point includes: The conductor-side receiving area is determined based on the direction of the receiving area corresponding to the receiving point on the working side. Based on the continuous positions of the dynamic leader in the continuous event stream segment, the region traversed by the dynamic leader is determined; The pilot pointing area is determined based on the line segment between the second pilot end position and the entry position of the pilot pointing line; The moving hazard intrusion window is generated based on the conductor side receiving area, the area traversed by the dynamic leader end, and the area traversed by the leader direction.

[0017] Optionally, in one possible implementation of the first aspect, generating evidence data and mobile hazard event reporting information based on the mobile hazard intrusion window includes: Based on the monitoring field of view correspondence between the event camera and the RGB camera, the moving hazard intrusion window is mapped to the evidence collection area in the RGB image; An RGB evidence image is generated based on the evidence collection area, which includes a moving hazard encroachment window marker, a work-side acceptance point marker, and a pilot directional line marker. The evidence collection event stream segment corresponding to the moving hazard encroachment window is then extracted. The evidence data is generated based on the RGB evidence image and the evidence event stream fragment; Based on the position of the mobile hazard encroachment window in the monitoring field of view, the position of the work-side acceptance point, the correspondence between the pilot directional line and the directional acceptance area, the work-side pilot change information, and the evidence collection data, the mobile hazard event reporting information is generated.

[0018] A second aspect of this application provides a mobile hazard event detection system, comprising: The conductor module, based on the conductor projection position information in the monitoring field of view and the safety interval requirements of the transmission conductor, determines the dangerous area adjacent to the conductor and generates candidate receiving points along the dangerous area adjacent to the conductor. The pilot module is used to generate operational pilot change information of the moving hazard object based on the event stream output by the event camera. The operational pilot change information is used to characterize the operational pilot change state of the moving hazard object relative to the operational side of the adjacent danger zone of the conductor. The receiving module is used to determine the receiving point on the working side based on the receiving relationship between the pilot change information on the working side and the candidate receiving point in the dangerous area adjacent to the conductor. The window module generates a mobile hazard encroachment window based on the work-side receiving point, and generates evidence collection data and mobile hazard event reporting information based on the mobile hazard encroachment window.

[0019] The mobile hazard event sensing method and system provided in this application have the following beneficial effects: This application no longer relies solely on the UAV itself, the entire aerial moving target, or a fixed monitoring area as the basis for alarms. Instead, it focuses on the dynamic leading end of the suspended object that moves into the dangerous area near the conductor during operation as the key point of risk perception. The actual risk location is determined by the relationship between the changes in the leading end on the operation side and the candidate receiving points. This distinguishes between ordinary proximity to the conductor and the precursory signs of the suspended part moving into the dangerous area near the conductor, reducing false alarms caused by the entire target approaching the conductor. Simultaneously, the moving hazard intrusion window can be generated around the risk-indicating area of ​​the receiving point on the operation side and the dynamic leading end, allowing the evidence data to more concentratedly cover the hooks, edges of suspended objects, or lower ends of ropes that are truly likely to approach the conductor first, thereby improving the accuracy of moving hazard event alarms and evidence collection. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the mobile hazard event perception method provided in an embodiment of this application; Figure 2 This is an application scenario diagram of the mobile hidden danger event perception method provided in the embodiments of this application; Figure 3 This is a schematic diagram of the structure of the mobile hazard event perception system provided in the embodiments of this application; Figure 4 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0022] The technical solutions of this application will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0023] See Figure 1 This is a flowchart illustrating the mobile hazard event perception method provided in an embodiment of this application. Figure 1 The execution entity of the method shown can be a software and / or hardware device. The execution entity of this application can include, but is not limited to, at least one of the following: user equipment, network equipment, etc. User equipment can include, but is not limited to, computers, smartphones, personal digital assistants (PDAs), and the aforementioned electronic devices. Network equipment can include, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers. Cloud computing is a type of distributed computing, consisting of a super virtual computer composed of a group of loosely coupled computers. This embodiment does not impose any limitations on this.

[0024] See Figure 2 This is an application scenario diagram of this application. The event camera 102 is positioned facing the power transmission line 101 and its surrounding low-altitude work area to collect event streams generated when a mobile hazard object 103 approaches the power transmission line 101. The mobile hazard object 103 can be a suspended drone; when its suspended part swings towards the adjacent danger zone 104 of the power line, the mobile hazard intrusion window 105 covers the leading change area of ​​the suspended part and the risk-bearing area on the power line side, representing the corresponding area for evidence collection and reporting determined by the method of this application.

[0025] The method for detecting mobile potential hazards includes steps S1 to S4, as detailed below: Step S1: Based on the information of the projection position of the transmission line in the monitoring field of view and the safety interval requirements of the transmission line, determine the dangerous area adjacent to the transmission line and generate candidate receiving points along the dangerous area adjacent to the transmission line.

[0026] It should be noted that in low-altitude suspended operations on power transmission lines, conventional monitoring methods typically involve setting up a fixed monitoring area near the transmission line or determining whether to issue an alarm based on the distance between the drone and the line. When the drone is near the line, the hook, the edge of the suspended object, or the lower end of the sling may swing towards the line before the drone itself, and the fixed area or the drone's position may not accurately reflect this local swing position. Therefore, this step first uses the actual projected position of the transmission line in the monitoring field of view as a benchmark, and combines this with the safety clearance requirements of the transmission line to determine the adjacent danger zone. Then, candidate receiving points are generated within this zone, ensuring that subsequent dynamic changes on the operational side of the pilot end can be received at specific areas and points on the line side.

[0027] The monitoring field of view refers to the image field of view that the event camera or RGB camera can cover the transmission line and its surrounding low-altitude work area. The conductor projection position information is used to characterize the image position of the transmission line within the monitoring field of view. The safety distance requirement for the transmission line can be the safety distance between the conductor and external moving targets specified in the line operation and maintenance specifications, or a safety distance pre-configured based on the line voltage level, on-site work type, and monitoring ratio. Candidate acceptance points are candidate locations generated along the conductor's adjacent hazardous area, used to accept subsequent pilot directional lines or dynamic pilot end placement trends.

[0028] In some embodiments, step S1 includes steps S11 and S12: Step S11: Based on the information of the conductor projection position in the monitoring field of view and the safety interval requirements of the transmission conductor, determine the dangerous area adjacent to the conductor.

[0029] Understandably, step S11 is used to convert the actual safety clearance requirements near the transmission line into an area range within the monitoring field of view. The inputs to this step include the projected position information of the transmission line and the safety clearance requirements, and the output is the hazardous area adjacent to the transmission line. This hazardous area is used in subsequent steps to determine whether the dynamic pilot end of the suspended object is swaying towards the hazardous space on the transmission line side, and is also used to generate candidate locations.

[0030] In some embodiments, the conductor projection position information in step S11 includes the conductor reference projection and the current conductor projection; obtaining the conductor projection position information includes steps A1 and A2: Step A1: Determine the reference projection of the transmission line in the monitoring field of view, and extract the current projection of the line based on the current acquired image.

[0031] The conductor reference projection can be determined during the installation and commissioning phase of the monitoring device. After the monitoring device is fixed in the transmission line, reference images can be acquired during periods when there are no suspended drones, no significant obstructions, and the conductor position is stable. Based on the linear edges of the transmission conductor in the reference images, the reference position of the conductor in the monitoring field of view is determined, and this reference position is used as the conductor reference projection. The conductor reference projection can be recorded in the form of several reference pixels, line segments, or fitted curves.

[0032] When extracting the current projection of a traverse from the currently captured image, a traverse search zone can be established centered on the traverse reference projection. The width of the traverse search zone can be determined based on the possible wind deflection range of the traverse, the installation error of the monitoring device, and the range of image jitter; for example, it can be extended outward by 30 to 80 pixels on both sides of the traverse reference projection. Then, linear edges are extracted within the traverse search zone. Linear edges that are continuously distributed, have a width within the width range of the traverse image, and whose direction is close to that of the traverse reference projection are identified as the current projection of the traverse. If the traverse in the currently captured image is partially obscured by hanging objects, the drone itself, or other objects, the traverse edges on both sides of the obscured area can be extended and connected to obtain the complete current projection of the traverse.

[0033] Step A2: Obtain the traverse projection position information based on the traverse reference projection and the current traverse projection.

[0034] The traverse projection position information can include the traverse reference projection, the current traverse projection, the endpoint positions of the current traverse projection, the changes in the direction of the current traverse projection, and the offset information of the current traverse projection relative to the traverse reference projection. The offset information can be determined by the shortest distance from the sampling point on the current traverse projection to the traverse reference projection, or by the average offset direction and average offset amount of the current traverse projection relative to the traverse reference projection. In this way, the traverse projection position information retains both the reference position of the traverse and reflects its actual position in the current image, providing input for subsequent determination of traverse projection points and local danger zones.

[0035] It should be noted that steps A1 to A2 form the conductor projection position information by combining the conductor reference projection and the conductor current projection, so that the subsequent adjacent danger zone of the conductor can be generated based on the actual position of the conductor in the current picture, rather than being fixed in the historical position, thereby providing a spatial basis for judging the local placement of the suspended object into the conductor side area.

[0036] In some embodiments, step S11 includes steps S111 to S114: Step S111: Based on the endpoint position, direction change position, and preset segmentation interval of the current projection of the conductor in the conductor projection position information, determine multiple conductor projection points on the current projection of the conductor.

[0037] The endpoints of the current traverse projection are the starting and ending points of the traverse projection within the monitoring field of view. The change-of-direction location is the point where the traverse's direction changes significantly within the current projection. To determine the change-of-direction location, sampling points can be extracted along the current traverse projection at fixed step sizes. Several adjacent sampling points are fitted into local line segments, and the angle between adjacent local line segments is calculated. If the angle is greater than a preset angle, such as 8 to 15 degrees, the corresponding location is determined as the change-of-direction location. The preset segment interval controls the distance between adjacent traverse projection points and can be determined based on the current traverse projection length. For example, when the current traverse projection length is L pixels, the preset segment interval can be L / 20 to L / 10; alternatively, it can be 30 to 80 pixels as a fixed segment interval. The endpoints, change-of-direction locations, and points supplemented according to the preset segment interval together constitute multiple traverse projection points.

[0038] Step S111 transforms the current projection of the conductor from a continuous linear object into multiple ordered conductor projection points, giving subsequent conductor traverse lines and local segments a clear structural basis.

[0039] Step S112: Connect adjacent traverse projection points to obtain the traverse route line, and determine the line segment between two adjacent traverse projection points as a local segment of the traverse.

[0040] Multiple traverse projection points are sorted according to the current extension order of the traverse projections, and adjacent traverse projection points are connected sequentially to obtain the traverse route line. The traverse route line consists of multiple connecting parts, each corresponding to a traverse local segment. Each traverse local segment has a first segment end position and a second segment end position, which are the traverse projection points at the two ends of the traverse local segment, respectively. The traverse local segments serve as the construction units for subsequent local hazard zones, enabling the formation of segmented hazard zones adjacent to the traverse along the traverse route.

[0041] Through step S112, the subsequent processing no longer treats the entire conductor as a single object, but rather a local segment of the conductor as a local processing object, which makes it easier to connect the placement trend of the suspended object to a specific local area of ​​the conductor.

[0042] Step S113: Based on the safety interval requirements of the transmission lines, local danger zones are set up on the side of each local section of the transmission line.

[0043] Understandably, step S113 is used to implement the safety interval requirements to the side of each conductor segment. This step processes conductor segments, is based on the safety interval requirements of the transmission line, and outputs multiple local danger zones.

[0044] In some embodiments, step S113 includes steps S1131 to S1137: Step S1131: Determine any local segment of the conductor as the target local segment, and determine the first end position and the second end position of the target local segment.

[0045] Specifically, the local segments of the traverse line can be traversed sequentially according to their arrangement. The currently processed local segment is identified as the target traverse segment. The two ends of the target traverse segment are the first segment end position and the second segment end position, respectively. The first and second segment end positions are recorded in the monitoring field of view coordinates and used for subsequent generation of lateral boundaries and outer endpoints.

[0046] Step S1132: The two opposite sides of the local segment of the target conductor in the monitoring field of view are respectively determined as the first local generation side and the second local generation side.

[0047] A direction vector is formed from the first end position to the second end position to create a local segment of the target traverse. This direction vector is then rotated by 90 degrees and -90 degrees respectively to obtain two lateral directions of the local segment. These two lateral directions correspond to two opposite sides of the local segment in the monitoring field of view and are designated as the first local generation side and the second local generation side, respectively. By considering both sides simultaneously, the system can accommodate situations where the suspended UAV approaches from either side of the traverse.

[0048] Step S1133: Determine the receiving width based on the safety interval requirements.

[0049] The safety clearance requirement can be expressed as the actual safe distance D between the transmission line and an external moving target. After the monitoring device is installed, the pixel ratio P at the depth of the conductor can be obtained through on-site calibration, where P represents the number of pixels per meter. The bearing width W can be obtained by multiplying D and P, i.e., W = D × P. If different voltage levels correspond to different safety distances, the corresponding safety distance D can be selected based on the current line voltage level. If there are perspective changes in the monitoring field of view, the local pixel ratio can be determined for different local sections of the conductor, and the corresponding bearing width can be calculated accordingly.

[0050] Through step S1133, the safety interval requirement is converted into the pixel width in the monitoring field of view, so that the lateral extent of the local danger zone has a clear source.

[0051] Step S1134: Determine the target local generation side from the first local generation side and the second local generation side, and take the direction from the local segment of the target conductor to the target local generation side as the target lateral direction.

[0052] Specifically, the construction of local hazard zones can be performed using the first local generation side and the second local generation side as the target local generation side, respectively. When the first local generation side is the target local generation side, the direction from the target traverse segment to the first local generation side is taken as the target lateral direction; when the second local generation side is the target local generation side, the direction from the target traverse segment to the second local generation side is taken as the target lateral direction. The target lateral direction is used to determine the direction in which the local hazard zone expands outward from the traverse segment.

[0053] Step S1135: In the lateral direction of the target, determine the positions that are respectively a distance of the bearing width from the first end position and the second end position as the first outer endpoint and the second outer endpoint.

[0054] The first outer endpoint is obtained by extending the bearing width along the lateral direction of the target from the first segment end position; the second outer endpoint is obtained by extending the bearing width along the lateral direction of the target from the second segment end position. The first and second outer endpoints are used to define the location of the local danger zone away from the local section of the conductor.

[0055] Step S1136: Connect the first outer endpoint and the second outer endpoint to obtain the local outer boundary.

[0056] Connecting the first and second outer endpoints yields a local outer boundary positioned opposite the local segment of the target conductor. The lateral distance between the local outer boundary and the local segment of the target conductor is defined by the bearing width, forming the outer boundary of the local danger zone.

[0057] Step S1137: Enclose the local segment of the target conductor, the first lateral boundary from the first segment end position to the first outer endpoint, the second lateral boundary from the second segment end position to the second outer endpoint, and the local outer boundary to obtain the local danger zone corresponding to the local generation side of the target.

[0058] The target traverse segment, the first lateral boundary, the second lateral boundary, and the local outer boundary together form a closed region, which serves as the local danger zone corresponding to the local generation side of the target. If the closed region is not a regular rectangle due to image perspective or changes in traverse direction, the polygon enclosed by the four boundaries is still used as the local danger zone. This local danger zone records its corresponding traverse segment, local generation side, connection width, and boundary coordinates.

[0059] It should be noted that steps S1131 to S1137 construct local danger zones by defining the segment end position, target lateral direction, and connection width, ensuring that each conductor segment has a local area beside it corresponding to the safety clearance requirement. These multiple local danger zones provide local spatial units for the generation of adjacent danger zones and candidate connection points for subsequent conductors.

[0060] Step S114: Generate the adjacent danger zone of the conductor based on each local danger zone.

[0061] Multiple local danger zones arranged along the conductor's direction and located on the same side are combined to obtain the adjacent danger zone of the conductor on the corresponding side. If there is a gap between adjacent local danger zones smaller than a preset gap threshold, the gap can be filled; if adjacent local danger zones overlap, the overlapping part is retained. The preset gap threshold can be 5% to 10% of the acceptance width. The adjacent danger zones of the conductors on both sides together form the adjacent danger zone of the transmission conductor in the monitoring field of view.

[0062] It should be noted that step S11 determines the dangerous area near the conductor by the current projection of the conductor and the safety interval requirements, so that the space range that needs to be monitored near the conductor can be generated with the current projection of the conductor, and provides a spatial reference for the subsequent dynamic leader end of the suspended object to be pointed.

[0063] Step S12: Generate candidate acceptance points along the traverse near the danger zone.

[0064] Understandably, step S12 is used to generate candidate locations in the hazard area adjacent to the conductor that can accommodate the tendency of suspended objects to be placed. The input to this step is the hazard area adjacent to the conductor and the local hazard area, and the output is a set of candidate locations.

[0065] In some embodiments, step S12 includes steps S121 to S123: Step S121: Determine the local segment of the conductor that shares a side with any local danger zone as the target local segment of the conductor, and determine the boundary of any local danger zone that is opposite to the target local segment of the conductor as the receiving side boundary.

[0066] For any local danger zone, the corresponding target conductor segment from step S1137 is read. This target conductor segment shares an edge with the local danger zone. The boundary of the local danger zone opposite to the target conductor segment is the local outer boundary, which is determined as the receiving side boundary. The receiving side boundary is used to indicate the initial outer receiving position when the suspended object is placed from the outside of the conductor into the adjacent danger zone.

[0067] Step S122: Starting from the midpoint of the local segment of the target traverse, generate a receiving lateral line along the direction from the local segment of the target traverse to the receiving lateral boundary.

[0068] The midpoint of a local segment of the target traverse can be determined by the midpoint between the first and second segment ends. Starting from the midpoint, a receiving lateral line is generated along the direction from the local segment of the target traverse towards the receiving side boundary. The receiving lateral line is used to establish the correspondence between the local traverse segment and the receiving side boundary within the local danger zone.

[0069] Step S123: Determine the intersection of the receiving lateral line and the receiving lateral boundary as the boundary receiving position, and use the boundary receiving position as a candidate receiving point.

[0070] Calculate the intersection point of the lateral line and the boundary of the receiving side. If there is one intersection point, that intersection point is the boundary receiving point; if there are multiple intersection points, select the intersection point closest to the location in the segment as the boundary receiving point; if the lateral line does not intersect with the boundary of the receiving side, select the boundary point on the boundary of the receiving side that is closest to the location in the segment as the boundary receiving point. The boundary receiving points corresponding to each local danger zone constitute a set of candidate receiving points.

[0071] It should be noted that step S12 generates candidate receiving points by using the local danger zone and the receiving side boundary, so that the danger zone adjacent to the conductor is not just a surface area, but further forms a set of points that can be received by the subsequent dynamic leader pointing trend.

[0072] Preferably, step S1 transforms the safety concern area near the transmission line into regions and locations that can be accepted by the event flow analysis results through the continuous construction between the current projection of the conductor, safety interval requirements, local danger zones, and candidate acceptance points. In this way, subsequent steps can map the local swaying trend of the suspended object to the specific acceptance location on the conductor side, rather than simply determining whether the UAV body or the target as a whole is close to the conductor.

[0073] Step S2: Generate the operation-side leader change information of the moving hazard object based on the event stream output by the event camera. The operation-side leader change information is used to characterize the operation-side change state of the dynamic leader end of the moving hazard object relative to the adjacent danger area of ​​the conductor.

[0074] It should be noted that conventional image monitoring methods typically use the UAV itself or the entire aerial moving target as the identification object. When a hoisting UAV performs lateral movement, hovering correction, hoisting, or lowering, local parts such as the hook, the edge of the hoisted object, or the lower end of the hoisting rope may move into the danger zone near the guide wire before the UAV itself. If only the position of the UAV itself is processed, it is difficult to determine the local part that actually approaches the guide wire first. Therefore, this step uses the event stream output by the event camera to extract the event change area of ​​the hoisting UAV itself, the candidate follow-up event area, the event change area of ​​the hoisted object, and the near-line boundary, and further determines the dynamic leader end and its operational side change status. The operational side leader change information output in step S2 is used in step S3 to determine the connection relationship between the dynamic leader end and the candidate receiving point.

[0075] The event stream consists of multiple event points, each of which may include an event occurrence timestamp, the event point's pixel coordinates in the monitoring field of view coordinate system, and the polarity of the brightness change. To ensure temporal continuity of motion changes, the event stream can be divided into continuous event stream segments according to the timestamps. Continuous event stream segments can be determined using a sliding time window, with a window length ranging from 0.5 seconds to 2 seconds and an overlap ratio between adjacent windows ranging from 40% to 70%. In one embodiment, the window length is 1 second, and adjacent windows overlap by 0.5 seconds. Continuous event stream segments are used to form the body position sequence, boundary position sequence, and continuous position of the dynamic leader.

[0076] In some embodiments, step S2 includes steps S21 to S26: Step S21: Extract the change region of the main body event of the suspended UAV in the event stream.

[0077] Event density maps are generated according to time sub-windows in continuous event stream segments. The time sub-window can be 20 to 50 milliseconds. Event points within each time sub-window are counted by pixel position to obtain the event density map. Adjacent pixels with event densities greater than a first density threshold are grouped into event connected regions. The first density threshold can be determined based on background event noise, for example, 3 to 5 times the average background event density. Area and bounding box constraints are applied to the event connected regions, retaining regions whose area falls within the UAV body area and whose bounding box aspect ratio falls within the body aspect ratio range as candidate body regions. The UAV body area range and aspect ratio range can be obtained by calibrating typical sling-mounted UAV samples; for example, the area range is 100 to 5000 pixels, and the aspect ratio range is 0.5 to 3. If the center point displacement of candidate body regions in multiple consecutive time sub-windows is continuous, the area change rate is less than the area change threshold, and the distance between the center points of adjacent time sub-windows is less than the upper limit of body displacement, then these candidate body regions are merged into a single body event change region.

[0078] Step S21 can determine the event change area corresponding to the sling-operated UAV body from the event stream. This area serves as a spatial reference for subsequent searching of candidate follow-up event areas and sling connection belts.

[0079] Step S22: Based on the position of the change region of the ontology event in the event flow, determine the candidate follow-up event regions in the event flow that are adjacent to the change region of the ontology event.

[0080] Specifically, candidate search areas are established below and to the sides of the bounding box of the event change region. The upper boundary of the candidate search area can connect with the lower boundary of the bounding box of the event change region, the horizontal width can be 1.5 to 3 times the width of the bounding box of the event change region, and the vertical height can be 2 to 6 times the height of the bounding box of the event change region. If the hanging direction may be tilted, the width of the main body can be expanded by 1 times on both sides of the event change region. Connectivity is extracted from the event density map in the candidate search area, and event connectivity regions with a minimum distance to the event change region less than the proximity distance threshold are retained as candidate follow-up event regions. The proximity distance threshold can be 0.2 to 1 times the height of the bounding box of the event change region, or 20 to 100 pixels. If a candidate follow-up event region appears only in a single time sub-window and does not appear again, the occasional event region is discarded; if a candidate follow-up event region maintains a relative positional association with the event change region in at least two consecutive time sub-windows, it is retained as a candidate follow-up event region.

[0081] Step S22 can extract candidate regions that may have a following relationship with the drone's hanging structure from the vicinity of the change area of ​​the main event, thus avoiding treating all moving areas in the event flow as hanging objects.

[0082] Step S23: Determine the candidate follower event region in which there is a continuous event portion between the candidate follower event region and the main event change region, and use the continuous event portion as the hanging connection belt.

[0083] A connection search band is established between the center point of the ontological event change region and the center point of the candidate follower event region. The width of the connection search band can be 5 to 20 pixels. If there is a connected path of event points within the connection search band that extends from the boundary of the ontological event change region to the boundary of the candidate follower event region, and the maximum interval between adjacent event points in the connected path is less than the break distance threshold, then it is determined that there is a continuous event portion between the candidate follower event region and the ontological event change region. The break distance threshold can be 3 to 10 pixels. If the connected path appears in two or more consecutive time sub-windows, then the set of event points corresponding to the connected path is determined as the hanging connection band. If there are multiple connected paths, the connected path with the highest average event density and the shortest path length is selected as the hanging connection band.

[0084] Step S23 can explicitly express the connection between the change region of the main event and the candidate follower event region, so that the determination of the change region of the subsequent hanging object event not only depends on spatial proximity, but also on the event connection relationship.

[0085] Step S24: Based on the relative changes of the hanging connection belt, the main body event change area, and the candidate follower event area in the continuous event flow segment, determine the hanging object event change area in the candidate follower event area, and determine the near-line boundary of the hanging object event change area toward the adjacent danger area of ​​the conductor.

[0086] Understandably, step S24 is used to filter out the actual hoisting event change areas related to the hoisting operation from the candidate follow-up event areas, and further determine the boundary towards the near-line danger zone. The inputs to this step are the hoisting connection strip, the body event change area, the candidate follow-up event areas, and the near-line danger zone, and the outputs are the hoisting event change area and the near-line boundary.

[0087] In some embodiments, step S24, based on the relative changes of the hanging connection strip, the body event change region, and the candidate follower event region in a continuous event stream segment, determines the hanging object event change region within the candidate follower event region and determines the near-line boundary of the hanging object event change region toward the conductor's adjacent hazardous area, including steps B1 to B3: Step B1: Determine the direction of the suspension connection belt from the main body event change area to the connected candidate follower event area as the suspension direction.

[0088] Specifically, the end of the suspension connecting belt closer to the area of ​​change in the main event is designated as the first connecting end, and the end closer to the area of ​​the candidate follower event is designated as the second connecting end. The direction from the first connecting end to the second connecting end is taken as the suspension direction. If the suspension connecting belt consists of multiple event points, a principal direction fitting can be performed on the suspension connecting belt, and the direction of the fitted line from the area of ​​change in the main event to the area of ​​the candidate follower event is taken as the suspension direction.

[0089] Step B2: In a continuous event stream segment, the candidate follower event region located in the suspension direction, connected to the suspension connecting belt, and laterally offset relative to the main body event change region is determined as the suspension event change region.

[0090] The relative position between the center point of the candidate follower event region and the center point of the main event change region is decomposed to obtain the component along the hanging direction and the lateral component perpendicular to the hanging direction. If the candidate follower event region is located in the hanging direction, and the time it maintains connection with the hanging connecting strip within a continuous event flow segment is not less than 50% of the length of the continuous event flow segment, and the change in the lateral component relative to the initial lateral component exceeds the lateral offset threshold, then the candidate follower event region is determined as the hanging object event change region. The lateral offset threshold can be taken as 10% to 30% of the width of the bounding box of the main event change region, or 10 to 40 pixels. If multiple candidate follower event regions meet the above conditions, the region with the longest connection duration with the hanging connecting strip is selected as the hanging object event change region; if the connection duration is the same, the region with the largest lateral offset change is selected as the hanging object event change region.

[0091] Step B3: Determine the near-line boundary of the area in the area of ​​change of the suspended object event that faces the adjacent danger zone of the conductor.

[0092] The outer contour of the area affected by the hanging object event is extracted, resulting in a set of boundary points consisting of multiple boundary points. The minimum distance from each boundary point to the adjacent hazardous area of ​​the conductor is calculated, and the set of consecutive boundary points whose minimum distance is less than the near-line distance threshold is selected as near-line boundary candidates. The near-line distance threshold can be 1 to 2 times the bearing width. If multiple near-line boundary candidates exist, the set of boundary points with the closest average distance to the adjacent hazardous area of ​​the conductor is selected as the near-line boundary. If the distance from all boundary points to the adjacent hazardous area of ​​the conductor is greater than the near-line distance threshold, the set of consecutive boundary points closest to the adjacent hazardous area of ​​the conductor is selected as the near-line boundary, and the corresponding state is marked as the observation state.

[0093] It should be noted that steps B1 to B3 determine the event change area of ​​the suspended object by the suspension connection belt, suspension direction, connection holding state and lateral offset relationship, and further determine the near line boundary, so that the subsequent dynamic leader can be generated from the side boundary of the suspended object facing the dangerous area of ​​the conductor, rather than from the entire suspended object or the UAV body.

[0094] Step S25: Determine the dynamic leader based on the nearline boundary.

[0095] In some embodiments, step S25 includes steps S251 to S254: Step S251: Determine the relative swing direction of the nearline boundary with respect to the change region of the body event based on the boundary position sequence of the nearline boundary in the continuous event flow segment and the body position sequence of the change region of the body event in the continuous event flow segment.

[0096] The boundary position sequence can be composed of the positions of the near-line boundary centroids within each time sub-window, and the body position sequence can be composed of the positions of the center point of the body event change region within each time sub-window. For two adjacent time sub-windows, the near-line boundary centroid displacement vector and the body center displacement vector are calculated. Subtracting the body center displacement vector from the near-line boundary centroid displacement vector yields the relative displacement vector of the near-line boundary relative to the body event change region. The average direction of multiple relative displacement vectors in a continuous event flow segment is calculated to obtain the relative oscillation direction. If the relative displacement vector direction fluctuation exceeds a direction fluctuation threshold, a weighted average is performed using the three to five time sub-windows closest to the current time.

[0097] Step S252: Based on the relative swing direction and the location of the adjacent danger zone of the conductor relative to the side of the near-line boundary, determine the local boundary segment from the near-line boundary.

[0098] In the set of boundary points near the conductor boundary, select the boundary points located ahead of the relative swing direction and whose distance to the adjacent danger zone is less than the average distance of the near-line boundary. Select the spatially continuous set of these boundary points as candidates for local boundary segments. If multiple local boundary segment candidates exist, select the set of points with the largest projected length along the relative swing direction as the local boundary segment; if the projected lengths are the same, select the set of points with the smaller average distance to the adjacent danger zone as the local boundary segment.

[0099] Step S253: Determine the end portion of the local boundary segment located at the front end of the relative swing direction as the candidate leader end.

[0100] The boundary point with the largest projection value along the relative swing direction in the local boundary segment is taken as the end point, and the end segment consisting of this end point and several adjacent boundary points is taken as the end part. The position of the candidate leader can be taken as the centroid of the end segment. If the local boundary segment contains only a few boundary points, the boundary point at the foremost point along the relative swing direction is directly taken as the candidate leader.

[0101] Step S254: Among the candidate leader ends, select the candidate leader end whose extension line along the corresponding relative swing direction intersects the adjacent danger area of ​​the conductor and whose intersection point is the closest, and use it as the dynamic leader end.

[0102] Starting from each candidate leader, an extension line is generated along the corresponding relative swing direction, and the intersection of the extension line with the adjacent danger zone of the conductor is calculated. If there are multiple intersections between the extension line and the adjacent danger zone of the conductor, the intersection point that first enters the adjacent danger zone of the conductor along the relative swing direction is selected as the valid intersection point. The intersection distance is the distance from the candidate leader to the valid intersection point. The candidate leader with the smallest intersection distance is selected as the dynamic leader. If multiple candidate leaders have the same intersection distance, the candidate leader with the higher event density in its local boundary segment is selected; if the event density is still the same, the candidate leader closer to the current projection of the conductor is selected. If the extension lines of all candidate leaders do not intersect with the adjacent danger zone of the conductor, no intrusion alarm type dynamic leader is generated. Instead, the candidate leader closest to the adjacent danger zone of the conductor is selected as the observation leader, and the judgment is updated in the next continuous event stream segment.

[0103] It should be noted that steps S251 to S254 determine the dynamic leader point that first points towards the adjacent danger zone of the conductor along the swing direction based on the swing relationship between the near-line boundary and the event change area of ​​the main body. This dynamic leader point is the basis for the subsequent generation of leader change information on the work side.

[0104] Step S26: Generate the operation-side leader change information based on the dynamic leader terminal.

[0105] In some embodiments, step S26 includes steps S261 to S265: Step S261: Determine the operation action change node based on the displacement change of the main event change area in the continuous event flow segment and the offset change of the suspended object event change area relative to the main event change area.

[0106] Calculate the displacement direction and amount of the center point of the main event change area between adjacent time sub-windows, and simultaneously calculate the offset direction and amount of the center point of the suspended object event change area relative to the center point of the main event change area. If the change angle of the displacement direction of the main event change area between adjacent time sub-windows exceeds a first change angle threshold, or the change angle of the offset direction of the suspended object event change area relative to the main event change area exceeds a second change angle threshold, or the change in the offset of the suspended object event change area relative to the main event change area exceeds an offset change threshold, then the midpoint of the corresponding time sub-window is determined as the operation action change node. The first and second change angle thresholds can be taken from 15 degrees to 30 degrees, and the offset change threshold can be taken from 10% to 25% of the width of the outer frame of the suspended object event change area. If multiple time points meet the conditions, the earliest time point that meets the conditions and whose subsequent two consecutive time sub-windows still maintain the change trend is selected as the operation action change node.

[0107] Step S262: Determine the position of the first pilot end before the operation action change node and the position of the second pilot end after the operation action change node.

[0108] The first leader position can be taken as the average position of the dynamic leader within the adjacent time sub-window before the change node of the operation action, and the second leader position can be taken as the average position of the dynamic leader within the adjacent time sub-window after the change node of the operation action. If the dynamic leader position is missing in a certain time sub-window, linear interpolation is performed using the positions of the previous and next time sub-windows; if more than two consecutive time sub-windows are missing, no valid leader direction is generated for the current continuous event stream segment, and the segment is marked as a segment with insufficient evidence.

[0109] Step S263: Connect the first leader position and the second leader position to obtain the leader change segment, and take the direction from the first leader position to the second leader position as the leader direction.

[0110] Connecting the first and second leader positions yields a leader change segment. The direction from the first leader position to the second leader position is used as the leader direction. If the distance between the first and second leader positions is less than the minimum displacement threshold, the current dynamic leader change is considered insufficient to form a valid leader direction; in this case, the leader direction from the previous continuous event stream segment can be used, or the current segment can be marked as a segment with no valid change. The minimum displacement threshold can be between 5 and 15 pixels.

[0111] Step S264: Starting from the position of the second leader end, generate a leader pointing line along the leader direction.

[0112] Starting from the second leader position, extend along the leader direction to obtain the leader pointing line. The extension length of the leader pointing line shall at least cover the distance between the second leader position and the nearest boundary of the adjacent danger zone of the conductor; if the leader pointing line does not intersect with the adjacent danger zone of the conductor, it shall extend to the boundary of the monitoring field of view or to a preset maximum length. The preset maximum length can be 3 to 5 times the coverage width.

[0113] Step S265: Generate pilot change information on the work side based on the pointing relationship between the pilot pointing line and the adjacent danger zone of the conductor.

[0114] In some embodiments, step S265 includes steps S2651 to S2653: Step S2651: Based on the extension result of the leader pointing line along the leader direction and the change in the distance between the leader pointing line and the adjacent danger area of ​​the conductor, determine the regional pointing state of the leader pointing line relative to the adjacent danger area of ​​the conductor; the regional pointing state includes the pointing state and the moving away state.

[0115] In some embodiments, step S2651, determining the regional pointing state of the pilot pointing line relative to the adjacent danger zone of the conductor based on the extension result of the pilot pointing line along the pilot direction and the change in the distance between the pilot pointing line and the adjacent danger zone of the conductor, includes steps C1 and C2: Step C1: Determine that the leader pointing line extends along the leader direction and points towards or approaches the danger zone adjacent to the conductor, and determine the area pointing state as the pointing state.

[0116] If the leader pointing line intersects with the adjacent danger zone of the conductor, then the leader pointing line is determined to point towards the adjacent danger zone. If the leader pointing line does not intersect with the adjacent danger zone, but the minimum distance between the leader pointing line and the adjacent danger zone decreases within a continuous time sub-window after extending along the leader direction, and the angle between the leader direction and the direction from the dynamic leader tip to the nearest point of the adjacent danger zone is less than the pointing angle threshold, then the leader pointing line is determined to approach the adjacent danger zone. The pointing angle threshold can be between 30 degrees and 45 degrees. When any of the above conditions are met, the area pointing state is determined to be the pointing state.

[0117] Step C2: If the leader pointing line deviates from the adjacent danger zone of the conductor after extending along the leader direction, or if the distance between the leader pointing line and the adjacent danger zone of the conductor increases, the area pointing state is determined to be a "far away" state.

[0118] If the leader pointing line does not intersect with the adjacent danger zone of the conductor, and the minimum distance between the leader pointing line and the adjacent danger zone of the conductor increases within a continuous time sub-window, then the leader pointing line is determined to be moving away from the adjacent danger zone of the conductor. If the angle between the leader direction and the direction from the dynamic leader tip to the nearest point of the adjacent danger zone of the conductor is greater than the distance-away angle threshold, the leader pointing line is also determined to be moving away from the adjacent danger zone of the conductor. The distance-away angle threshold can be between 60 and 90 degrees. When any of the above conditions are met, the area pointing state is determined to be in a distance-away state. If neither the pointing state condition nor the distance-away state condition is met, the area pointing state of the previous time sub-window is maintained until either the pointing state or distance-away state condition is met.

[0119] Step S2652: Determine the operational-side change state of the dynamic leader based on the state transition relationship between the pointing state and the far-away state in the continuous event stream segment, as well as the continuity relationship of the pointing state.

[0120] In some embodiments, step S2652, determining the operational-side change state of the dynamic leader based on the state transition relationship between the pointing state and the far-away state in the continuous event stream segment, and the persistence relationship of the pointing state, includes steps D1 to D4: Step D1: Determine that the area pointing state changes from the far-away state to the pointing state, and determine the change state of the working side of the dynamic leader as the swing-in state.

[0121] Step D2: Determine that the region pointing state changes from pointing state to moving away state, and determine the change state of the working side of the dynamic leader as the swing-back state.

[0122] Step D3: Determine that the area pointing state changes back to pointing state after undergoing a transition from pointing state to far away state, and determine the change state of the working side of the dynamic leader as the re-entry state.

[0123] Step D4: Determine that the region pointing state remains continuously in the pointing state in the continuous event stream segment, and determine the operation-side change state of the dynamic leader as the continuous swing-in state.

[0124] "Continuously maintaining a pointing state" can refer to the area being in a pointing state for N consecutive time sub-windows, where N can be 3 to 5; or it can refer to the area maintaining a pointing state for a duration of 0.15 to 0.5 seconds. If both a "re-swinging in" state and a "continuous swinging in" state occur simultaneously in the same continuous event stream segment, the "continuous swinging in" state can be used as the main operation-side change state for the current segment, while the "re-swinging in" state is retained as a state change record.

[0125] Step S2653: Generate operation-side leader change information based on the operation-side change status of the dynamic leader terminal.

[0126] The operational leader change information may include the dynamic leader position, operational leader change status, leader direction, leader pointing line, area pointing status change record, and continuous event stream segment time range. The operational leader change information is output to step S3 to determine the operational leader's position relative to the adjacent hazardous area of ​​the conductor, its entry position, and the pointing receiving area.

[0127] It should be noted that step S265 generates work-side pilot change information that characterizes the dynamic swing-in, swing-back, re-swing-in, or continuous swing-in of the pilot end by analyzing the extension results, spacing changes, and state transition relationships of the pilot guideline. This information is not a single distance value, but rather reflects the changing trend of a local part of the suspended object relative to the adjacent danger zone of the guideline.

[0128] It should be noted that step S26 establishes the pointing relationship of the dynamic pilot end before and after the change of the operation action by using the operation action change node, the position of the first pilot end, the position of the second pilot end, and the pilot pointing line, and generates the operation-side pilot change information to provide input for the subsequent determination of the operation-side receiving point.

[0129] Preferably, step S2 identifies the follow-up relationship between the UAV body and the suspended object through event flow, determines the dynamic leader from the near-line boundary of the suspended object, and generates the operation-side leader change information based on the pointing change of the dynamic leader, so that subsequent processing can focus on the local part of the suspended object that is likely to approach the guide wire first.

[0130] Step S3: Determine the work-side receiving point based on the change information of the pilot on the work side and the receiving relationship between the candidate receiving point and the receiving point in the adjacent danger zone of the conductor.

[0131] It should be noted that conventional monitoring methods, upon receiving information that a drone or suspended object is approaching the conductor, typically only output an overall alarm indicating the target's proximity to the conductor, making it difficult to pinpoint the specific location within the adjacent hazardous area corresponding to this inclination trend. Step S3 receives the candidate acceptance point generated in Step S1 and the operational-side leader change information generated in Step S2. Based on the operational-side of the dynamic leader end, the position of the leader's pointing line entering the adjacent hazardous area of ​​the conductor, and the positional relationship of the candidate acceptance points, the operational-side acceptance point is determined. The operational-side acceptance point is used to map the inclination trend of the dynamic leader end to a specific acceptance location on the conductor side and serves as input for Step S4 to generate the moving hazard encroachment window.

[0132] In some embodiments, step S3 includes steps S31 to S34: Step S31: Based on the change information of the working side leader, determine the working side of the dynamic leader relative to the dangerous area adjacent to the conductor.

[0133] Using the current projection or the traverse line as a boundary, calculate which side of the current projection the dynamic leader is located on. If the dynamic leader is on one side of the traverse line and the leader's pointing line points towards the adjacent danger zone on the same side of the traverse, then that side is designated as the working side. If the dynamic leader is on one side of the traverse line, but the leader's pointing line enters the adjacent danger zone on the other side of the traverse, then the side corresponding to the leader's pointing line entering the adjacent danger zone is designated as the working side. The working side is used to limit the selection range of subsequent candidate acceptance points.

[0134] Step S32: Determine the position where the pilot guide line enters the dangerous area adjacent to the conductor as the entry position, and determine the receiving area in the dangerous area adjacent to the conductor based on the entry position.

[0135] If the leader pointing line intersects with the adjacent danger zone of the conductor, the point where the first entry into the adjacent danger zone along the leader direction is the entry position. If the leader pointing line does not intersect with the adjacent danger zone of the conductor, but the area is pointing, the point in the adjacent danger zone of the conductor closest to the leader pointing line is the entry position, and this entry position is marked as the approaching entry position. When determining the directional receiving area, it is determined which local danger zone the entry position falls into. If the entry position is located inside or on the boundary of a local danger zone, that local danger zone is determined as the directional receiving area. If the entry position is located at the boundary of two local danger zones, the distance from the candidate receiving point in each of the two local danger zones to the entry position is calculated, and the local danger zone with the smaller distance is selected as the directional receiving area; if the distances are the same, the local danger zone with the smaller angle to the leader direction is selected as the directional receiving area.

[0136] Step S33: Select candidate acceptance points located on the working side and corresponding to the acceptance area from the candidate acceptance points to obtain the acceptance candidate points.

[0137] Whether a candidate acceptance point is located on the work side can be determined by its position relative to the traverse line. Whether a candidate acceptance point corresponds to a directional acceptance area can be determined by whether the local hazard zone to which the candidate acceptance point belongs is also a directional acceptance area. If a candidate acceptance point is located on the boundary of a directional acceptance area, it is also considered to correspond to that area. If no candidate acceptance point simultaneously meets both the conditions of being on the work side and directional acceptance area, then candidate acceptance points on the same work side are searched in the local hazard zone adjacent to the directional acceptance area, and the point closest to the entry position is used as a supplementary acceptance candidate point, and simultaneously marked as adjacent acceptance in the moving hazard event reporting information.

[0138] Step S34: Based on the connection relationship between the candidate receiving point and the entry point, determine the receiving point on the operation side.

[0139] In some embodiments, step S34 includes steps S341 to S343: Step S341: When the number of candidate sites for acceptance is determined to be a single one, the single candidate site for acceptance is determined as the acceptance site on the operation side.

[0140] If there is only one candidate receiving point obtained in step S33, then a unique receiving relationship is formed between the candidate receiving point and the entry position, and the candidate receiving point is directly determined as the receiving point on the operation side.

[0141] Step S342: When the number of candidate receiving points is determined to be multiple, the multiple candidate receiving points are sorted according to the receiving distance between each candidate receiving point and the entry position along the leading direction to obtain the receiving point sequence.

[0142] The acceptance distance is the absolute value of the projection length of the vector from the entry position to the candidate acceptance point onto the leader direction. A point vector is formed with the entry position as the starting point and the candidate acceptance point as the ending point. This point vector is then projected onto the unit vector in the leader direction to obtain the acceptance distance. If the acceptance distances are the same, the Euclidean distances from the candidate acceptance point to the entry position are compared, with the smaller Euclidean distance ranked higher. If the Euclidean distances are still the same, the candidate acceptance point with the larger overlap length between the corresponding local danger zone and the leader pointing line is ranked higher.

[0143] Step S343: The candidate receiving point with the smallest receiving distance in the receiving point sequence is determined as the receiving point on the operation side.

[0144] The candidate receiving point at the beginning of the receiving point sequence has the minimum receiving distance, and this candidate receiving point is determined as the receiving point on the operation side. The receiving point on the operation side is output to step S4 to determine the receiving area on the conductor side and the window for encroachment of moving hazards.

[0145] It should be noted that step S34 determines the work-side receiving point that best corresponds to the current pointing trend of the dynamic pilot end by the receiving distance between the receiving candidate point and the entry position, so that the local swinging trend of the suspended object can fall on the specific receiving position in the danger zone adjacent to the conductor.

[0146] Preferably, step S3 determines the receiving point on the work side by establishing the correspondence between the work side, the entry position, the target receiving area, and the candidate receiving point. This ensures that the candidate receiving point is no longer just a static point distributed along the guide line, but forms a receiving relationship with the changes on the work side of the dynamic pilot end, providing a positioning basis for the subsequent generation of the mobile hazard encroachment window.

[0147] Step S4: Generate a mobile hazard encroachment window based on the work-side acceptance point, and generate evidence collection data and mobile hazard event reporting information based on the mobile hazard encroachment window.

[0148] It should be noted that conventional evidence collection typically generates image records around the drone itself or a fixed monitoring area. However, when a drone is operating near a conductor, the key evidence to be collected often lies in the relationship between the dynamic leader of the suspended object, the area the leader points towards, and the conductor-side receiving area. Step S4 receives the receiving point on the operating side determined in step S3 and generates a moving hazard intrusion window by combining the area traversed by the dynamic leader and the area the leader points towards. The moving hazard intrusion window is used to limit the range of the RGB evidence image and the evidence event stream segment, ensuring that the evidence data and moving hazard event reporting information are formed around the process of the suspended object partially entering the dangerous area adjacent to the conductor.

[0149] In some embodiments, step S4 includes steps S41 and S42: Step S41: Generate a window for mobile hazard encroachment based on the work-side acceptance point.

[0150] In some embodiments, step S41 includes steps S411 to S414: Step S411: Determine the conductor side receiving area based on the direction of the receiving area corresponding to the receiving point on the working side.

[0151] Centered on the work-side receiving point, extend forward and backward by half the length of a local section of the conductor along the conductor's direction. Extend inward along the receiving side to cover the corresponding local section of the conductor, and outward to cover the boundary of the receiving side, thus obtaining the conductor-side receiving area. If the work-side receiving point is close to the edge of the receiving area, the conductor-side receiving area does not exceed the boundary of the receiving area. The conductor-side receiving area is used to represent the local extent of the adjacent hazardous area of ​​the conductor currently being pointed to by the dynamic leader.

[0152] Step S412: Determine the region traversed by the dynamic leader based on the continuous position of the dynamic leader in the continuous event stream segment.

[0153] The positions of the dynamic leader in each time sub-window within a continuous event stream segment are connected sequentially to form a leader path. Then, the dynamic leader path area is expanded outwards from the leader path by an end-expansion width to obtain the dynamic leader path area. The end-expansion width can be 10% to 30% of the shorter side length of the bounding box of the hanging object event change area, or 5 to 20 pixels. If the dynamic leader is missing in a certain time sub-window, linear interpolation is performed using the positions of the previous and next time sub-windows. If more than two consecutive time sub-windows are missing, the generation of the dynamic leader path area is stopped, and the continuous event stream segment is marked as an insufficient evidence segment.

[0154] Step S413: Determine the area traversed by the leader pointing line based on the line segment between the second leader end position and the entry position.

[0155] The leader line segment is defined as the line segment from the second leader position to the entry position. The leader line segment is then expanded outwards along the leader extension width to obtain the leader-traversed area. The leader extension width can be 20% to 50% of the bearing width, or 10 to 30 pixels. If the entry position is an approaching entry position, the leader line segment is the line segment from the second leader position to the approaching entry position. The leader-traversed area represents the trend path of the dynamic leader moving or approaching the adjacent danger zone of the conductor.

[0156] Step S414: Based on the conductor side receiving area, the area traversed by the dynamic leader end, and the area traversed by the leader direction, generate a moving hazard encroachment window.

[0157] First, calculate the union of the conductor-side receiving area, the area traversed by the dynamic leader, and the area traversed by the leader's pointing direction. Then, calculate the minimum bounding rectangle of this union. Next, expand the window margins around the minimum bounding rectangle to obtain the moving hazard encroachment window. The window margins can be 10 to 40 pixels, or 10% to 20% of the receiving width. If the expanded window exceeds the monitoring field of view boundary, the window boundary is clipped to within the monitoring field of view boundary. If the area of ​​the union of the three is less than the minimum window area, a rectangular window corresponding to the minimum window area is generated with the midpoint of the line connecting the work-side receiving point and the dynamic leader position as the center. The minimum window area can be no less than 80×80 pixels. The moving hazard encroachment window is output to step S42.

[0158] It should be noted that step S41 generates a moving hazard intrusion window by means of the conductor side receiving area, the area through which the dynamic leader end passes, and the area through which the leader points, so that the window simultaneously covers the conductor side receiving position, the actual position through which the dynamic leader end of the suspended object passes, and its pointing path toward the adjacent danger area of ​​the conductor.

[0159] Step S42: Generate evidence data and mobile hazard event reporting information based on the mobile hazard intrusion window.

[0160] In some embodiments, step S42 includes steps S421 to S424: Step S421: Based on the monitoring field of view correspondence between the event camera and the RGB camera, the moving hazard intrusion window is mapped to the evidence collection area in the RGB image.

[0161] The correspondence between the monitoring fields of view of the event camera and the RGB camera can be established through common calibration points. Common calibration points can be the ends of transmission lines, the edges of towers, insulator hanging points, or the corners of calibration plates. During calibration, at least four sets of corresponding points are recorded in both the event camera coordinate system and the RGB image coordinate system, and the mapping matrix from the event camera coordinates to the RGB image coordinates is calculated. The coordinates of the four corner points of the moving hazard encroachment window are input into the mapping matrix to obtain the four corresponding corner points in the RGB image; the area enclosed by these four corner points is used as the evidence collection area. If the mapped area is a non-rectangular quadrilateral, the circumscribed rectangle of the quadrilateral is taken as the RGB image evidence collection area; if the evidence collection area exceeds the RGB image boundary, it is cropped to within the RGB image boundary.

[0162] Step S422: Generate an RGB evidence image with a moving hazard encroachment window marker, a work-side acceptance point marker, and a pilot directional line marker based on the evidence collection area, and extract the evidence collection event stream segment corresponding to the moving hazard encroachment window.

[0163] RGB evidence images can be generated based on the current frame of the RGB camera or the RGB image frame corresponding to the time of a continuous event stream segment. The moving hazard intrusion window marker is displayed as a bounding box surrounding the evidence collection area; the work-side receiving point marker is displayed as a dot or crosshair; and the leader pointing line marker is displayed as an arrow pointing from the second leader position to the entry position. All of the above markers are converted to RGB image coordinates using a mapping matrix from event camera coordinates to RGB image coordinates. Evidence collection event stream segments are extracted from continuous event stream segments. The extraction time range is centered on the time of the moving hazard intrusion window generation, trunculating forward by T1 seconds and backward by T2 seconds, with T1 and T2 each ranging from 0.5 seconds to 1 second. The extraction spatial range is the moving hazard intrusion window and its outer area, with the width of the outer area equal to the window margin. Event data falling within this time range and whose event point coordinates are within this spatial range constitute the evidence collection event stream segment.

[0164] Step S423: Generate evidence data based on the RGB evidence image and the evidence event stream fragment.

[0165] The evidence data includes RGB evidence images, event stream clips, coordinates of the moving hazard encroachment window, coordinates of the work-side receiving point, coordinates of the pilot pointing line, dynamic changes in the work-side status of the pilot end, and event time. RGB evidence images provide a visual correspondence between the guide wire, the suspended drone, the suspended object, and the risk window; event stream clips preserve evidence of dynamic pilot end swing and pointing changes. Together, they constitute the evidence data for moving hazard events.

[0166] Step S424: Based on the position of the moving hazard encroachment window in the monitoring field of view, the position of the work side acceptance point, the correspondence between the pilot line and the target acceptance area, the work side pilot change information, and the evidence data, generate the moving hazard event reporting information.

[0167] The reporting information for moving hazard incidents should include at least the monitoring point number, the time of the incident, the identification of the danger zone adjacent to the conductor, the work-side receiving point, the change status of the dynamic pilot end on the work side, the location of the moving hazard encroachment window, the RGB evidence image index, and the evidence collection event stream segment index. If the dynamic pilot end is in the observation state or has no valid entry position, the moving hazard incident reporting information is marked as an observation report; if the dynamic pilot end is in the swing-in state, the re-swing-in state, or the continuous swing-in state, and the pilot pointing line enters the danger zone adjacent to the conductor, then a moving hazard incident reporting information is generated.

[0168] It should be noted that step S42 generates RGB evidence images and evidence collection event stream segments by moving the hazard encroachment window, and writes the window position, the work side acceptance point, the correspondence between the pilot line and the target acceptance area, and the work side pilot change information into the reporting information, so that the reporting results can correspond to the change process of the local pilot part of the suspended object toward the dangerous area adjacent to the conductor.

[0169] Preferably, step S4 generates a mobile hazard encroachment window through the work-side acceptance point, and generates evidence collection data and mobile hazard event reporting information around the window, so that the evidence collection range corresponds with the dynamic pilot end, the conductor-side acceptance position and the pilot pointing path, so that the reporting information is more suitable for expressing the pre-hanging signs of the suspended operation drone.

[0170] Based on the above steps, this application also includes the following embodiments: See Figure 3 This is a schematic diagram of the structure of the mobile hazard event perception system provided in the embodiments of this application, including: The conductor module, based on the conductor projection position information in the monitoring field of view and the safety interval requirements of the transmission conductor, determines the dangerous area adjacent to the conductor and generates candidate receiving points along the dangerous area adjacent to the conductor. The pilot module is used to generate operational-side pilot change information for moving hazard objects based on the event stream output by the event camera. The operational-side pilot change information is used to characterize the operational-side change state of the dynamic pilot end of the moving hazard object relative to the adjacent danger zone of the conductor. The receiving module is used to determine the receiving point on the work side based on the receiving relationship between the pilot change information on the work side and the candidate receiving point in the dangerous area adjacent to the conductor. The window module generates a mobile hazard encroachment window based on the work-side receiving point, and generates evidence collection data and mobile hazard event reporting information based on the mobile hazard encroachment window.

[0171] Figure 3 The system of the illustrated embodiment can be used to perform corresponding operations. Figure 1 The steps in the method embodiments shown are implemented in a similar manner and have similar technical effects, and will not be repeated here.

[0172] See Figure 4 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. The electronic device 40 includes: a processor 41, a memory 42, and a computer program; wherein, The memory 42 is used to store the computer program, and the memory may also be flash memory. The computer program is, for example, an application program or functional module that implements the above method.

[0173] The processor 41 is configured to execute the computer program stored in the memory to implement the various steps performed by the device in the above method. For details, please refer to the relevant descriptions in the preceding method embodiments.

[0174] Alternatively, the memory 42 can be either standalone or integrated with the processor 41.

[0175] When the memory 42 is a device independent of the processor 41, the device may further include: Bus 43 is used to connect the memory 42 and the processor 41.

[0176] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for sensing mobile potential hazard events, characterized in that, include: Based on the projection position information of the transmission line in the monitoring field of view and the safety interval requirements of the transmission line, the dangerous area adjacent to the transmission line is determined, and candidate receiving points are generated along the dangerous area adjacent to the transmission line. Based on the event stream output by the event camera, operational-side leader change information of the moving hazard object is generated. The operational-side leader change information is used to characterize the operational-side change state of the dynamic leader end of the moving hazard object relative to the operational-side change state of the adjacent danger zone of the conductor. Based on the connection between the pilot change information on the work side and the candidate receiving points in the dangerous area adjacent to the conductor, the receiving points on the work side are determined. Based on the work-side receiving point, a mobile hazard encroachment window is generated, and evidence collection data and mobile hazard event reporting information are generated according to the mobile hazard encroachment window.

2. The method according to claim 1, characterized in that, The determination of the adjacent danger zone based on the conductor projection position information of the transmission conductor in the monitoring field of view and the safety interval requirements of the transmission conductor includes: Based on the endpoint position, direction change position, and preset segmentation interval of the current projection of the conductor in the conductor projection position information, multiple conductor projection points on the current projection of the conductor are determined; Connecting adjacent traverse projection points yields the traverse route, and the line connecting two adjacent traverse projection points is defined as a local segment of the traverse. Based on the safety spacing requirements of transmission lines, local danger zones are set up on the side of local sections of each line. The adjacent danger zone of the conductor is generated based on each of the local danger zones.

3. The method according to claim 2, characterized in that, Based on the safety spacing requirements of transmission lines, local danger zones are set up beside local sections of each conductor, including: Any local segment of the conductor is identified as the target local segment, and the positions of the first and second ends of the target local segment are determined. The two opposite sides of the local segment of the target conductor in the monitoring field of view are respectively defined as the first local generation side and the second local generation side; The acceptance width is determined based on the aforementioned safety interval requirements; The target local generation side is determined from the first local generation side and the second local generation side, and the direction from the target conductor local segment to the target local generation side is taken as the target lateral direction; In the target lateral direction, the positions that are respectively distanced from the first segment end position and the second segment end position from the bearing width are determined as the first outer end point and the second outer end point; Connect the first outer endpoint and the second outer endpoint to obtain the local outer boundary; Enclosing the local segment of the target conductor, the first lateral boundary from the end position of the first segment to the first outer endpoint, the second lateral boundary from the end position of the second segment to the second outer endpoint, and the local outer boundary, a local danger zone corresponding to the local generation side of the target is obtained.

4. The method according to claim 2, characterized in that, The generation of candidate acceptance points along the conductor near the danger zone includes: The local segment of the conductor that shares a side with any of the aforementioned local danger zones is identified as the target local segment of the conductor, and the boundary in any of the aforementioned local danger zones that is opposite to the target local segment of the conductor is identified as the receiving side boundary. Starting from the midpoint of the local segment of the target conductor, a receiving lateral line is generated along the direction from the local segment of the target conductor toward the receiving lateral boundary; The intersection of the receiving side line and the receiving side boundary is determined as the boundary receiving position, and the boundary receiving position is used as the candidate receiving point.

5. The method according to claim 1, characterized in that, The step of generating operational-side precursor change information for moving hazard objects based on the event stream output by the event camera includes: Extract the change region of the main body event of the suspended UAV in the event stream; Based on the position of the changed region of the ontology event in the event flow, determine the candidate follow-up event regions in the event flow that are adjacent to the changed region of the ontology event; Identify candidate follower event regions in which there are continuous event portions between the candidate follower event regions and the main event change regions, and use the continuous event portions as hanging connection belts; Based on the relative changes of the suspension connection belt, the main body event change region, and the candidate follower event region in a continuous event stream segment, the suspension event change region in the candidate follower event region is determined, and the near-line boundary of the suspension event change region toward the adjacent danger area of ​​the conductor is determined. The dynamic leader is determined based on the near-line boundary, and the operational-side leader change information is generated based on the dynamic leader.

6. The method according to claim 5, characterized in that, Determining the dynamic leader based on the nearline boundary includes: The relative swing direction of the nearline boundary with respect to the body event change region is determined based on the boundary position sequence of the nearline boundary in the continuous event flow segment and the body position sequence of the body event change region in the continuous event flow segment. Based on the relative swing direction and the side of the adjacent danger zone of the conductor relative to the near-line boundary, a local boundary segment is determined from the near-line boundary; The end portion of the local boundary segment located at the front end of the relative swing direction is identified as a candidate leader end; Among the candidate leader terminals, the candidate leader terminal whose extension line along the corresponding relative swing direction intersects the adjacent danger zone of the conductor and whose intersection point is the closest is selected as the dynamic leader terminal.

7. The method according to claim 5, characterized in that, The generation of the operation-side leader change information based on the dynamic leader includes: Based on the displacement change of the main event change region in the continuous event stream segment and the offset change of the suspended object event change region relative to the main event change region, the operation action change node is determined; Determine the position of the dynamic pilot end before the operation action change node and the position of the second pilot end after the operation action change node; Connect the first leader position and the second leader position to obtain a leader change segment, and take the direction from the first leader position to the second leader position as the leader direction; Starting from the position of the second leader end, a leader pointing line is generated along the leader direction; Based on the directional relationship between the pilot line and the adjacent danger zone of the conductor, the operational-side pilot change information is generated.

8. The method according to claim 7, characterized in that, The step of generating the work-side leader change information based on the pointing relationship of the leader line relative to the adjacent danger zone of the conductor includes: Based on the extension of the leading pointing line along the leading direction and the change in the distance between the leading pointing line and the adjacent danger zone of the conductor, the regional pointing state of the leading pointing line relative to the adjacent danger zone of the conductor is determined; the regional pointing state includes a pointing state and a moving away state. Based on the state transition relationship between the pointing state and the far-away state in the continuous event stream segment, and the persistence relationship of the pointing state, the operational side change state of the dynamic leader is determined. The operation-side leader change information is generated based on the operation-side change status of the dynamic leader terminal.

9. The method according to claim 8, characterized in that, The step of determining the work-side receiving point location based on the receiving relationship between the work-side pilot change information and the candidate receiving point locations in the adjacent hazardous area of ​​the conductor includes: Based on the change information of the working side leader, the working side of the dynamic leader relative to the dangerous area adjacent to the conductor is determined; The position where the pilot directional line enters the dangerous area adjacent to the conductor is determined as the entry position, and the directional receiving area is determined in the dangerous area adjacent to the conductor based on the entry position; Select the candidate receiving points located on the working side and corresponding to the pointing receiving area from the candidate receiving points to obtain the receiving candidate points; Based on the connection relationship between the candidate receiving points and the entry position, the receiving point on the work side is determined.

10. The method according to claim 9, characterized in that, The process of determining the receiving point on the work side based on the receiving relationship between the candidate receiving points and the entry position includes: When the number of candidate sites for acceptance is determined to be a single point, the single candidate site for acceptance is determined as the acceptance site on the work side. When the number of the receiving candidate points is determined to be multiple, the multiple receiving candidate points are sorted according to the receiving distance between each receiving candidate point and the entry position along the leading direction to obtain a receiving point sequence; The candidate receiving point with the smallest receiving distance in the receiving point sequence is determined as the receiving point on the working side.

11. The method according to claim 1, characterized in that, The process of generating a mobile hazard encroachment window based on the work-side acceptance point includes: The conductor-side receiving area is determined based on the direction of the receiving area corresponding to the receiving point on the working side. Based on the continuous positions of the dynamic leader in the continuous event stream segment, the region traversed by the dynamic leader is determined; The pilot pointing area is determined based on the line segment between the second pilot end position and the entry position of the pilot pointing line; The moving hazard intrusion window is generated based on the conductor side receiving area, the area traversed by the dynamic leader end, and the area traversed by the leader direction.

12. The method according to claim 11, characterized in that, The step of generating evidence data and mobile hazard event reporting information based on the mobile hazard intrusion window includes: Based on the monitoring field of view correspondence between the event camera and the RGB camera, the moving hazard intrusion window is mapped to the evidence collection area in the RGB image; An RGB evidence image is generated based on the evidence collection area, which includes a moving hazard encroachment window marker, a work-side acceptance point marker, and a pilot directional line marker. The evidence collection event stream segment corresponding to the moving hazard encroachment window is then extracted. The evidence data is generated based on the RGB evidence image and the evidence event stream fragment; Based on the position of the mobile hazard encroachment window in the monitoring field of view, the position of the work-side acceptance point, the correspondence between the pilot directional line and the directional acceptance area, the work-side pilot change information, and the evidence collection data, the mobile hazard event reporting information is generated.

13. A mobile hazard event sensing system, employing the mobile hazard event sensing method as described in any one of claims 1 to 12, characterized in that, include: The conductor module, based on the conductor projection position information in the monitoring field of view and the safety interval requirements of the transmission conductor, determines the dangerous area adjacent to the conductor and generates candidate receiving points along the dangerous area adjacent to the conductor. The pilot module is used to generate operational pilot change information of the moving hazard object based on the event stream output by the event camera. The operational pilot change information is used to characterize the operational pilot change state of the moving hazard object relative to the operational side of the adjacent danger zone of the conductor. The receiving module is used to determine the receiving point on the working side based on the receiving relationship between the pilot change information on the working side and the candidate receiving point in the dangerous area adjacent to the conductor. The window module generates a mobile hazard encroachment window based on the work-side receiving point, and generates evidence collection data and mobile hazard event reporting information based on the mobile hazard encroachment window.

Images