A field fast identification system for power transmission and transformation project soil ecological sensitivity zoning

Abstract

The present application relates to power transmission and transformation engineering ecological environment field investigation, and discloses a kind of fast identification system of soil ecological sensitivity zoning, including processing server, field control terminal, water supply terminal group and image terminal group;Processing server is according to engineering actual land boundary division disturbance unit, is configured with engineering mark for each disturbance unit, and determines the corresponding boundary water outlet position;Water supply terminal group includes the water supply terminal of being set respectively in each boundary water outlet position, each water supply terminal is equipped with flowmeter, electric control valve and water outlet, and field control terminal opens the electric control valve in predetermined order one by one;The image terminal group successively collects the corresponding downhill soil surface image before and after water supply of current boundary water outlet position;Disturbance unit is accurate with the actual land boundary after field review, and with engineering mark through boundary water outlet position, water supply record, surface image and water mark data, investigation result can be traced back to specific construction disturbance source.

Description

Technical Field

[0001] This invention belongs to the field of ecological environment investigation and evaluation technology for power transmission and transformation projects, specifically involving a system for rapidly identifying soil ecological sensitivity zones at construction sites. Background Technology

[0002] Power transmission and transformation project construction typically involves occupying areas such as tower foundation working faces, tensioning working faces, construction access roads, and substation working faces. Excavation, filling, leveling, and repeated passage by construction vehicles can all alter the original micro-topography, surface soil structure, and local runoff conditions. When conducting ecological surveys during construction, soil and water conservation monitoring, or designing subsequent restoration measures, it is essential not only to delineate the disturbed area but also to clarify where surface runoff originates, migrates along its path, and is ultimately received. Zoning based on these practical relationships is crucial to accurately reflect the sensitivity of different surface areas in the processes of runoff formation, transport, and reception.

[0003] Current field investigations are mostly based on project land occupation data, combined with slope, soil texture, surface cover, land use type, and existing erosion traces for comprehensive judgment. Possible confluence directions can also be inferred through geographic information overlay or elevation data. These methods are suitable for completing large-scale preliminary assessments and can provide investigators with route and key area references. When the site topography is simple and drainage facilities are fixed, the inferred results usually have a certain degree of usability.

[0004] However, the actual conditions at the construction site are often more fragmented than the design data. Construction boundaries may be adjusted during the operation, and temporary road ruts, temporary mounds of earth, shallow ditches, and temporary drainage outlets may alter the local flow direction; several outlets may appear in the same disturbed area, and surface water from different disturbed areas may share a channel downstream. It is difficult to continuously correlate specific disturbed units, boundary outlet locations, actual runoff paths, and receiving locations based solely on static topography, visible scour marks, or a few discrete survey points.

[0005] Controlled water injection on-site can visually demonstrate soil infiltration and surface water migration processes, but continuous testing also introduces new identification challenges: the wetted areas formed in the initial tests remain on the surface, and subsequent watermarks may overlap with them; the segmented imaging equipment has different fields of view, requiring synchronized timing and alignment of surface positions; when watermarks stop extending, whether they truly converge, and which disturbance source the downstream path after convergence belongs to cannot be determined solely from a single image. Therefore, water supply records, continuous images, surface elevation, and engineering sources need to be processed in the same data chain to generate traceable on-site sensitivity zoning results. Summary of the Invention

[0006] This invention aims to provide a rapid identification system suitable for power transmission and transformation engineering construction sites. The system establishes disturbance units marked with engineering symbols according to the actual land boundaries, sequentially implements controlled water supply at each boundary water outlet location, and extracts and reconstructs water traces from continuous images before and after water supply. After the identified water traces stop extending or merge, it further forms disturbance source sensitive areas, runoff channel sensitive areas, and runoff receiving sensitive areas. Thus, even with multiple water outlet locations, multiple disturbance sources, and shared downstream paths, the association between disturbance units and the actual runoff path, termination location, and receiving area can be maintained.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: A rapid on-site identification system for soil ecological sensitivity zones in power transmission and transformation projects includes a processing server, an on-site control terminal, a water supply terminal group, and an image terminal group.

[0008] The processing server divides the area into disturbance units according to the actual boundary of the project, configures an engineering identifier for each disturbance unit, and determines the corresponding boundary water outlet location.

[0009] The water supply terminal group consists of water supply terminals set at each boundary water outlet position. Each water supply terminal includes a flow meter, an electrically controlled valve, and a water outlet. The field control terminal controls the electrically controlled valves to open one by one in a predetermined order.

[0010] The image terminal group is used to continuously acquire images of the downhill soil surface before and after water supply at the current boundary water outlet location.

[0011] The processing server extracts newly added wet areas connected to the current boundary water outlet from the above image and reconstructs the current water trail accordingly. When the current water trail stops extending downhill or intersects with an already confirmed water trail, the processing server sends a water stop signal and determines the termination position of the water trail; if an intersection occurs, it also associates the already confirmed water trail downstream of the intersection position with the current disturbance unit.

[0012] After receiving a water outage signal, the on-site control terminal closes the electrically controlled valve that is currently open.

[0013] The processing server combines the disturbance unit, the associated water trace, and the water trace termination position to generate the disturbance source sensitive area, the runoff channel sensitive area, and the runoff receiving sensitive area.

[0014] In one implementation, the processing server includes a server processor, a server memory, and a server communication interface, while the field control terminal includes a terminal processor, a terminal memory, and a terminal communication interface.

[0015] The server communication interface is connected to the terminal communication interface and the image terminal group respectively; the terminal communication interface is connected to the flow meter and the electric control valve of each water supply terminal respectively.

[0016] When dividing the disturbance units, the processing server reads the actual land occupation boundaries of the tower foundation working face, tension working face, construction access road, and substation working face. The engineering identifier of each disturbance unit is determined by the engineering type, construction period, and spatial coordinates, and the surface runoff traces that cross the actual land occupation boundaries are recorded under the corresponding engineering identifier.

[0017] Furthermore, the processing server first identifies the junction between the surface runoff traces and the downhill side of the actual land boundary as candidate water outlet locations, and then verifies whether it is connected to the surface grooves, rut guide strips, or drainage outlets outside the boundary. Locations that meet the connection conditions are determined as boundary water outlet locations.

[0018] The on-site control terminal writes the cumulative water output fed back by the flow meter into the water supply record at the corresponding boundary water outlet position.

[0019] For image acquisition, the image terminal group consists of multiple portable image terminals and ground calibration devices. The portable image terminals are arranged in segments along the downhill direction of the boundary water outlet, and adjacent shooting ranges overlap in the area where the ground calibration devices are located.

[0020] The portable image terminal records the planar position and elevation of the ground calibration component and acquires continuous downhill soil surface images with timestamps generated by the field control terminal.

[0021] Water supply is conducted in a time-sharing manner. The on-site control terminal allocates the preset water supply volume of the current boundary water outlet position to multiple water supply time periods. After each time period is completed, the electric control valve is closed and the system waits for the processing server to return the water trace extension status. The next water supply time period will only continue if the newly added wet area is still expanding downhill and has not yet touched the confirmed water trace.

[0022] When the experiment switches from the current boundary water outlet position to a subsequent boundary water outlet position, the processing server uses the last frame image after the water supply stops at the current position as the new reference image. Areas within the existing wetted area of ​​this reference image in subsequent images are removed, and the remaining continuous wetted area is used as the new wetted area for the subsequent boundary water outlet position.

[0023] For newly added wet areas spanning multiple shooting ranges, the processing server uses ground calibration elements shared by adjacent fields of view to align the surface position, and then merges areas with continuous acquisition time and contiguous surface positions into the current water trail. The current water trail is stored as a directed path graph, where the boundary water outlet position is the source node, and the branch positions, confluence positions, and termination positions are the connection nodes.

[0024] When determining the convergence, the processing server simultaneously verifies the continuity of the wetted area and the surface elevation near the intersection. If the current watermark and the confirmed watermark form a continuous wetted area, and the elevation decreases sequentially downstream of the confirmed watermark, the watermarks are determined to converge. The downstream watermarks from the intersection to the termination position of the confirmed watermark are then associated with the engineering identifier of the current disturbance unit, triggering a water outage signal.

[0025] When generating sensitive areas, the actual boundary of the disturbance unit corresponding to the confirmed water trace is designated as the disturbance source sensitive area; all newly added wet areas from the boundary water outlet to the water trace termination point are designated as runoff channel sensitive areas; and the wet areas that continue to extend at the termination point after the water trace stops extending are designated as runoff receiving sensitive areas. Each sensitive area retains its engineering identifier, cumulative water discharge, and corresponding water trace association.

[0026] Compared with existing field investigation methods, the present invention has at least the following advantages: First, the disturbance unit is based on the actual land boundary verified on-site, and engineering markers are used to trace the water outlet location, water supply records, surface images, and water trace data along the boundary. The investigation results can be traced back to the specific source of construction disturbance. Second, the on-site control terminal controls the electrically controlled valves according to the water supply period, while the processing server decides whether to continue water supply based on whether the water trace continues to extend or touches a confirmed water trace, thus forming a closed loop between water supply operation and identification results. Third, by using ground calibration devices to align the positions of different shooting ranges, and combining temporal and surface continuity to reconstruct the water trace, the actual path characteristics such as branching, merging, and termination can be preserved.

[0027] Furthermore, when switching the boundary water outlet location, a reference mask is established using the last frame of the previous test, which can distinguish existing wet areas from subsequently added areas. The confluence judgment simultaneously employs continuous wetting relationships and changes in surface elevation, reducing misjudgments caused by image overlap, vegetation obstruction, or drainage structures. The resulting three types of sensitive areas simultaneously preserve the project source, cumulative water discharge, and water trace correlations, providing relatively complete field evidence for ecological surveys, soil and water conservation monitoring, and the deployment of restoration measures in the construction area.

[0028] The following description, in conjunction with the accompanying drawings and specific embodiments, further illustrates the system structure, on-site deployment, and processing procedures. Attached Figure Description

[0029] The accompanying drawings form part of this specification and are used to illustrate the system structure and processing procedures of the present invention. The contents shown in the drawings are merely illustrative examples of embodiments and do not constitute a limitation on the scope of protection.

[0030] Figure 1 This is a schematic diagram of the components and their communication relationships of the on-site rapid identification system in an embodiment of the present invention; Figure 2This is a flowchart illustrating the on-site identification and sensitive area generation process in an embodiment of the present invention; Figure 3 This is a flowchart illustrating the extraction of newly added wet areas from a surface image in an embodiment of the present invention; Figure 4 This is a flowchart of the time-sharing water supply and water trace extension status feedback in an embodiment of the present invention; Figure 5 This is a flowchart illustrating the water trace confluence verification and shared downstream path association in an embodiment of the present invention; Figure 6 This is a schematic diagram illustrating the generation and data association of three types of sensitive areas in an embodiment of the present invention. Detailed Implementation

[0031] A specific embodiment of the present invention is described below with reference to the accompanying drawings. This embodiment is used to explain the implementation process of the technical solution and does not limit the specific models, layout dimensions, or data processing platforms of the components.

[0032] For ease of understanding, the main terms used in this article are defined below. Project land occupation refers to the area actually occupied by tower foundation working faces, tension working faces, construction access roads, or substation working faces during the construction of a power transmission and transformation project, and which may alter the surface runoff pattern. A disturbance unit is a data object defined by a closed actual land occupation boundary and possessing an independent project identifier. A boundary water outlet location is the location where surface water crosses the actual land occupation boundary on its downslope side from within the disturbance unit and continues to flow outwards towards the boundary. A watermark is a newly added moist area formed after controlled water supply, continuous on the surface, and extending downslope. A confirmed watermark is a watermark that has undergone position alignment, direction verification, and termination position determination and is stored in the processing server.

[0033] Figure 1 The system composition and data connections are described. The system includes a processing server, a field control terminal, a water supply terminal group, and an image terminal group. The processing server can be located at a field workstation or deployed as a remote server; the field control terminal is carried by the operator. The water supply terminal group is installed at the boundary water outlet location to be verified, while the image terminal group is arranged along its downhill direction. The processing server issues test tasks and water stop signals, receives downhill soil surface images uploaded by the image terminal group, and returns the image processing results and water trail extension status to the field control terminal. The field control terminal controls the water supply terminals according to the returned status sequence, allowing only one boundary water outlet location to discharge water at a time, thus ensuring that each newly added wet area has a clear source.

[0034] The processing server is equipped with a server processor, server memory, and server communication interface. The server memory stores the actual land boundary, topographic survey data, engineering markers, boundary water outlet locations, water supply records, surface images, ground calibration data, and watermark data. The server processor performs tasks such as perturbation unit construction, image alignment, wet area extraction, watermark reconstruction, confluence relationship confirmation, and sensitive area generation. The server communication interface establishes communication connections with the field control terminal and image terminal group, respectively. On-site, wireless LAN, mobile communication network, or wired network can be selected depending on conditions. The uploaded data packet must include at least the engineering marker, boundary water outlet location marker, acquisition time, and image sequence number or water supply period sequence number to group image records and water supply records under the same experimental task.

[0035] The field control terminal includes a terminal processor, a terminal memory, a terminal communication interface, and an operation display unit. The terminal memory stores the test sequence, preset water supply volumes for each boundary outlet position, target water supply volumes for a single water supply period, and safety shutdown conditions in case of communication interruption. One end of the terminal communication interface connects to the processing server, and the other end connects to the flow meters and electrically controlled valves of each water supply terminal. The terminal processor sends an opening command to one of the electrically controlled valves according to the test sequence and continuously reads the instantaneous flow rate and cumulative water supply volume of the corresponding flow meter; if the target water supply volume for the current period is reached, a water outage signal is received, or a communication timeout or flow abnormality occurs, the electrically controlled valve is immediately shut off. The operation display unit shows the current project identification, boundary outlet positions, cumulative water supply volume, and valve status for easy verification by on-site personnel.

[0036] The water supply terminal can consist of a water storage container and a connected water supply pipeline, with a flow meter, an electrically controlled valve, and a water outlet installed sequentially on the pipeline; the flow meter can also be installed downstream of the electrically controlled valve depending on the installation space. The water outlet is fixed at the boundary water outlet position near the inner side of the disturbance unit, facing the existing surface confluence mark. Its installation height is limited to ensure that the water flow does not cause visible soil displacement when it contacts the ground surface. Each water supply terminal is assigned an independent communication address, and the field control terminal addresses the corresponding flow meter and electrically controlled valve based on the boundary water outlet position identifier.

[0037] The image terminal group consists of multiple portable image terminals and several ground calibration devices. The portable image terminals can be field acquisition devices integrating camera, positioning, and communication components, with their field of view and shooting direction fixed by brackets. Each terminal is deployed in segments along the expected confluence direction, with overlapping areas maintained between adjacent fields of view, and the ground calibration devices are placed within these overlapping areas. At least four distinguishable calibration points are set on the surface of each calibration device, and the actual distance between these points is pre-determined. After deployment, the planar position, elevation, and orientation of the calibration devices are entered into the portable image terminals and uploaded to the processing server, thereby establishing the conversion relationship between image pixel positions and actual ground positions. The average image position alignment error is calculated as the average distance between the converted ground position of each calibration point and its corresponding measured position. When the ground surface has significant undulations, additional calibration devices can be added within the same field of view, and position conversion can be performed separately for local surface areas enclosed by adjacent calibration points.

[0038] Figure 2 The process shown begins with the organization of engineering data and on-site verification. The processing server receives design data such as tower locations, line routes, construction access roads, and substation land use areas, while simultaneously reading on-site data such as the actual land occupation boundaries after construction, surface elevation points, and visible surface runoff traces. The actual land occupation boundaries can be measured point by point along the boundary using satellite positioning surveying equipment or a total station, and then connected and closed into a boundary polygon according to the sampling sequence. Two work surfaces connected by the same access road, if with different construction methods or surface drainage directions, are established as separate disturbance units; areas with continuous boundaries and the same construction method, construction period, and runoff outlet can be grouped into the same disturbance unit.

[0039] Each disturbance unit is equipped with an independent engineering identifier. This identifier corresponds to the engineering type, construction period, and spatial coordinates: the engineering type distinguishes between tower foundation work surfaces, tensioning work surfaces, construction access roads, and substation work surfaces; the construction period distinguishes between surface disturbances formed in different batches; and the spatial coordinates can be the coordinates of the disturbance unit's center point or a preset reference point. This identifier serves as the link between the disturbance unit, the boundary water outlet location, water supply records, surface images, and water traces. Even if two disturbance units enter the same surface channel downstream, their engineering identifiers and water supply records are still retained separately.

[0040] The processing server determines the downslope boundary of each disturbance unit based on the site elevation and records visible confluence traces crossing the actual land boundary under the corresponding engineering markers. Visible confluence traces include shallow gullies formed by rainwater erosion, fine-particle sediment zones, directional litter zones, continuous low-lying areas within ruts, and outlet traces of artificial drainage structures. On-site personnel collect several location points along these traces, and the processing server connects these points to form a confluence trace line, while also saving the intersection of the confluence trace line and the actual land boundary.

[0041] The location of the boundary water outlet is determined in two steps. First, the intersection of the confluence line and the downslope side of the actual occupied boundary is listed as a candidate location. Then, it is verified whether this location can be continuously connected by the terrain outside the boundary. If there is a continuous low-lying centerline extending from the candidate location to the outside of the boundary, or if the location connects with the centerline of the drainage outlet, and the elevation of the end of the centerline outside the boundary is lower than the elevation of the candidate location, and the difference between the two is greater than the elevation measurement error, then it is confirmed as the boundary water outlet location. Locations with only water accumulation traces inside the boundary and no continuous connected terrain outside the boundary are not included as boundary water outlet locations.

[0042] When the same disturbance unit has multiple boundary outlet locations, the test sequence should generally be arranged from downstream to upstream, with priority given to testing independent channels that are expected to not intersect. To make this prediction, the processing server first generates a digital elevation model (DEM) covering a local area on the downslope side using measured elevation points.

[0043] This embodiment uses inverse distance weighted (IDW) interpolation with a distance decay exponent of 2. The search neighborhood selects the 5 to 12 nearest measured elevation points around the target location to balance the overall slope smoothness and the preservation of micro-topographical changes such as ruts and soil mound edges.

[0044] Then, a single-flow algorithm based on maximum elevation drop, such as the D8 algorithm, is used to generate a predicted confluence path for each candidate outlet location.

[0045] If a path does not intersect with other paths within the survey area, it is marked as an independent channel; if there is an intersection, it is marked as a potential merging channel. The processing server establishes a separate record for each boundary water outlet location, including the project identifier, location coordinates, outlet installation direction, preset water supply volume, and cumulative water output.

[0046] Before formal water supply, portable imaging terminals are set up in sections along the downhill slope of the water outlet location. The field of view of the first terminal covers the water outlet and the starting surface downstream of it, and subsequent terminals sequentially cover the surface downstream of the previous field of view; the overlap width of adjacent fields of view should ensure that the same ground calibration object and a section of the surrounding surface are simultaneously captured in the image. After fixing the exposure range and shooting direction, each terminal first acquires a reference image before water supply. The field control terminal generates a timestamp according to a unified clock and writes it into the acquisition task, and the processing server then forwards the task to each portable imaging terminal. The terminal writes the timestamp and image sequence number into the continuous images. The maximum time difference between the arrival of the same acquisition task at each imaging terminal is recorded as the time correction error.

[0047] The processing server checks whether the field of view covers the expected downhill passage, water outlet, and ground calibration points based on the reference image. If a fixed shadow exists in the image, the area is marked first, and the overall brightness variation during water supply is corrected using adjacent areas where the surface color has not changed. If the calibration points or image terminal are displaced, the original reference data is not continued to be used; instead, the position of the calibration points is remeasured and a new reference image is acquired.

[0048] After water supply begins, the on-site control terminal does not release the entire test water volume at once, but instead records the preset water supply volume at the current boundary outlet position as... And calculate using the following formula: In the formula, This refers to the upstream catchment width measured on-site. To determine the expected length of surface runoff testing; This represents the volumetric saturated water content of this soil type. The measured initial volumetric moisture content of the topsoil; The penetration depth parameter is taken as empirical, ranging from 0.1 m to 0.3 m in this embodiment; The maximum permissible test water volume on site is typically 10 to 20 liters. All parameters should be converted to compatible units before being substituted.

[0049] For extremely dry silty soils, or surfaces exhibiting significant hydrophobicity due to mechanical oil spills or other reasons, water droplet contact morphology should be observed simultaneously during the first water supply period. If the water flow in the image shows bead-like aggregation and rolling, and no lateral capillary extension occurs at the local permeability boundary, a hydrophobic compensation coefficient should be introduced when generating the runoff channel sensitive area. For example, the lateral buffer range should be expanded to 1.2 to 1.5 times the original calculated value to compensate for the underestimation of the potential lateral influence width.

[0050] In this embodiment, the judgment time for water supply pipeline anomalies is set to 15 to 30 seconds; the image acquisition interval is set to 1 to 5 seconds; and the observation time after the electrically controlled valve is closed is set to 5 to 10 minutes. The elevation measurement error is directly based on the nominal vertical accuracy of the total station or satellite positioning equipment, typically ±2 cm to ±5 cm. When calculating the normalized change, if the background median deviation is close to zero, a quantization step size is used as the minimum divisor: 1 / 255 is used when the 8-bit image channel is normalized to 0 to 1, and 1 is used when using integer pixel values ​​from 0 to 255, to avoid division by zero or abnormal magnification.

[0051] The target output for a single water supply period is less than the preset output. The upper limit of the anti-scour flow rate is determined through a short-term test flow before water supply, i.e., selecting the maximum instantaneous flow rate that has not yet caused visible soil displacement. During the test flow, the processing server gradually increases the flow rate in increments of 0.5 liters per minute and continuously acquires local images of the water flow contacting the ground surface. The dynamic region is first extracted using the adjacent frame difference method, and then opening and closing operations are performed sequentially using 3×3 or 5×5 pixel cross-shaped structural elements to filter out water splashes and light and shadow noise and fill in the internal pores of the soil block area.

[0052] When a dark connected region with an area greater than 50 to 100 pixels moves, and the displacement in an adjacent image exceeds 5 pixels for the first time, visible soil displacement is determined, the electronically controlled valve is immediately closed, and the previous flow step is recorded as the upper limit of the anti-scour flow rate. For example, the preset water supply at a certain boundary outlet position on the tower base working face can be 12 liters, divided into 4 time periods, with 3 liters in each time period; the actual value is still determined based on the scale of the disturbance unit and the soil infiltration status.

[0053] At the start of each water supply period, the field control terminal opens the currently controlled electric valve, reads the flow meter data, and sends the initial cumulative value, instantaneous flow rate, and current cumulative value for this period to the processing server. The image terminal group continuously collects data during the valve opening period. When the difference between the current cumulative value and the initial cumulative value reaches the target water output, or when the field control terminal receives a water outage signal, it immediately closes the electric valve and saves the cumulative water output at the time of closure.

[0054] If the cumulative water output does not increase within the preset time, the water supply pipeline is deemed abnormal and the valve is closed. If the instantaneous flow rate exceeds the upper limit of the anti-scour flow rate, the valve is also closed first, and the water supply period is restarted after the outlet or water supply pressure is adjusted on-site. The actual water output for each period is recorded by the flow meter and assigned to the current boundary outlet position.

[0055] Even after the electronically controlled valve is closed, the image terminal group continues to acquire data within the preset observation time. Figure 3 The process of extracting newly added wet areas is illustrated. The processing server first pairs the baseline image before water supply with the continuous images after water supply according to the timestamp, then uses a ground calibration component to complete the position correction, and subsequently calculates the decrease in brightness, the change in chromaticity, and the change in local texture at the same surface location. The decrease in brightness is taken as the positive difference in brightness before and after water supply; the change in chromaticity is taken as the square root of the sum of the squares of the differences between the two chromaticity components; and the change in local texture is taken as the magnitude of the difference in the mean gray-level gradient within the neighborhood of the same pixel.

[0056] In the specific calculation, the surface image is first converted to the CIELAB color space. The brightness decrease is calculated using the L channel value, and negative brightness differences are treated as zero; the chromaticity change is the Euclidean distance between the feature differences of the a and b channels; the grayscale gradient is obtained by summing the horizontal and vertical gradient magnitudes obtained by the Sobel operator in the 3×3 neighborhood centered on the target pixel.

[0057] Before water supply, a background area is selected within a fixed surface area that does not connect with the expected confluence channel. The three types of changes are each subtracted from their respective background median values ​​and divided by their corresponding background median deviations to obtain three normalized changes. The background median deviation is the median of the absolute values ​​of the differences between each change within the background area and its background median value; when this median deviation is less than one quantization step size, the quantization step size is used as the divisor to avoid numerical instability.

[0058] The processing server first synthesizes three normalized variables with each of the three initial weights being one-third. Considering the local illumination variations and differences in subsoil reflectivity in the field, fixed weights may amplify a certain high-noise dimension. Therefore, the sample variance of the three normalized variables in the background area is calculated in real time, and the weights are updated according to the inverse variance rule: the weight of a variable is equal to the proportion of its inverse sample variance to the sum of the inverses of the three variances.

[0059] This reduces the impact of noisy features on the overall change. The median value of the overall background change and the sum of three times the median deviation are used as the change threshold; pixels exceeding this threshold and remaining there for three consecutive frames constitute candidate wet regions. When the image acquisition frequency changes, the number of consecutive frames is adjusted accordingly to cover the same observation duration.

[0060] Once candidate wet areas are formed, the processing server begins a connectivity search from the surface location corresponding to the current water outlet. Areas directly connected to the water outlet, or those that can be traced back to the water outlet through adjacent candidate areas, are retained; scattered dark areas and existing puddles that do not have a continuous surface relationship with the water outlet are removed.

[0061] For suspected areas affected by wind-swept vegetation or large-area swaying shadows, perform Fast Fourier Transform (FFT) on their brightness curves across multiple consecutive frames; if the dominant frequency falls within the typical wind-swept frequency band of 0.5 Hz to 5 Hz and there is no continuous monotonous darkening trend, remove it as environmental dynamic interference.

[0062] The newly added wet area relative to the baseline image is the area that is finally retained. Its boundary, area, downhill front position, acquisition time and corresponding cumulative water output are recorded.

[0063] Figure 4The water supply feedback shown uses the downslope front of the newly moistened area as the judgment object. Here, the downslope front is the continuous boundary that is farthest from the water outlet position along the local elevation decrease direction. To reduce the impact of edge jitter caused by non-uniform capillary water absorption or small soil clods, the processing server calculates the planar Euclidean distance between each pixel on this continuous boundary after it is mapped to the ground surface and the current boundary water outlet position, and sorts them in descending order of distance.

[0064] The top 5% of points are selected as the frontier feature point set, and their geometric centroids at the Earth's surface are taken as the equivalent frontier position at that observation time. The frontier movement distance between adjacent observation periods is the straight-line distance between the two geometric centroids.

[0065] Using geometric centroid tracking instead of a single farthest point reduces the impact of edge noise on the travel distance. After the current water supply period ends, if the equivalent front continues to move downhill relative to the previous period, and the newly added wet area has not yet touched the confirmed water trace, the processing server returns a "continue to extend" status to the field control terminal, and only then will the field control terminal allow the start of the next water supply period; if this status is not received, the electrically controlled valve remains closed.

[0066] When the distance traveled by the leading edge of the downhill slope in two consecutive observation periods does not exceed the spatial determination threshold, the processing server determines that the current watermark has stopped extending downhill. The spatial determination threshold is the greater of twice the average image position alignment error and the actual ground distance between the two pixels, to avoid mistaking registration error for watermark movement.

[0067] When a newly added wet area is continuous with an already confirmed water stain location, the processing server records the intersection. Regardless of whether the water stain stops extending or intersects, the processing server sends a water stop signal with the current boundary water outlet location identifier to the field control terminal; the field control terminal verifies the identifier and closes the corresponding electrically controlled valve. When the preset water supply volume is reached, the valve remains closed and the current water supply task ends.

[0068] When water supply to a boundary outlet is completed and the system is ready to switch, the processing server sets the last frame image after water supply to that location stops as the current reference image, and saves it together with the project identifier, the completed boundary outlet identifier, and the last frame acquisition time.

[0069] After water supply begins at the subsequent boundary water outlet location, the processing server aligns the new image with the current reference image, identifies the already wetted surface area in the reference image, and subtracts that area from the subsequent candidate wetted areas.

[0070] The above subtraction is accomplished using a surface location mask. The processing server first merges the union of the extracted wet areas from each water supply period at the current boundary water outlet location, as well as the areas in the last frame image that still meet the wet change conditions, and then writes the merged surface coordinates into the reference mask.

[0071] After subsequent images are transformed to the same surface coordinate system, areas falling within the reference mask and not extending beyond it are excluded from the newly added wetted areas at subsequent water outlet locations; however, portions extending continuously from subsequent outlets beyond the reference mask are retained. This method distinguishes between wetted areas formed by successive experiments while preserving the intersection points where subsequent water traces connect with previous ones.

[0072] Crossing the field of view requires prior alignment of the ground positions. For two adjacent portable imaging terminals, the processing server identifies the ground calibration points they both captured. Based on the image position, actual distance, planar position, and elevation of the calibration points, it calculates the transformation relationship between the two images and the same ground coordinate system. After applying this relationship, the difference in ground position between the common calibration points is calculated.

[0073] The actual ground distance corresponding to a single pixel after conversion is defined as the ground pixel size. The preset registration error is the larger of the average image position alignment error and the ground pixel size. The alignment is accepted if the difference between the common calibration points does not exceed the preset registration error; otherwise, a new calibration point is selected or the corresponding field of view image is reacquired.

[0074] If the difference in position of the common calibration point still exceeds the preset registration error after using a single-plane transformation, the processing server establishes transformation relationships for each local surface region. Specifically, Delaunay triangulation is performed using the two-dimensional pixel coordinates of the calibration points within the same field of view as input nodes to obtain non-overlapping triangular meshes, with each triangular facet corresponding to a local surface region. The six-parameter affine transformation matrix of the region is solved by taking the two-dimensional pixel coordinates of the three vertices of the facet and the corresponding measured three-dimensional surface coordinates.

[0075] Newly added wet regions within a region are mapped to the 3D surface coordinate system by calling the affine transformation matrix of their respective patches. Adjacent patches in the Delaunay triangulation share boundary vertices, so even after independent transformation, each region maintains surface position continuity at the common boundary, reducing local perspective distortion caused by micro-topographic undulations. When a newly added wet region spans two patches, the portions on either side of the common boundary are transformed first, and then merged in the surface coordinate system.

[0076] After alignment, temporal continuity and surface continuity must be verified separately. Newly added wet areas belonging to the same boundary water outlet location, and where the difference in adjacent acquisition times is no greater than the sum of the set acquisition interval and the time correction error, are considered temporally continuous. Surface continuity is determined by the converted surface coordinates: if two areas intersect, or their shortest surface distance is no greater than the sum of the average image alignment error and the size of one surface pixel, and there are no unwetted protruding surfaces or drainage obstructions at the interval, they are considered to be connected at the surface location.

[0077] For newly added humid areas that simultaneously meet the conditions of temporal continuity and surface contact, the processing server merges them sequentially from earliest to latest based on the collection time. The merged records retain the first occurrence time, last occurrence time, water supply period, and cumulative water outflow of each area, and are bound to the current boundary water outflow location.

[0078] If a subsequent wet region connects to two separate parts of a previous wet region, then the three merge; if a subsequent wet region extends in two directions from the previous region, then two downstream branches are retained. All merged areas form a continuous surface region, constituting the current watermark.

[0079] The current water trail is stored in the form of a directed path graph. The water outlet position at the boundary is used as the source node, the branch position of the water trail, the confluence position of water trails from different sources, and the position where the extension stops are used as the connecting nodes, and the continuous wet region between adjacent connecting nodes is used as the directed edge.

[0080] The direction of directed edges is determined first from earliest to latest according to the first occurrence time of the newly added humid area, and verified by surface elevation. When two adjacent nodes have the same first occurrence time or the difference is less than the time resolution, the node with higher elevation points to the node with lower elevation. Each directed edge stores the surface boundary, length, elevation change, formation period, cumulative water discharge, and source engineering identifier of the humid area.

[0081] Branching, merging, and termination points are all determined by the actual water trace relationships. When a continuous wet region forms two or more separate downstream extensions from the same upstream location, the common connection point is recorded as the branching point; the location where wet regions from different sources first form a continuous connection is recorded as the merging point; the location furthest from the source node along the directed edge direction, and where no downstream directed edges appear in subsequent observation periods, is recorded as the termination point. When the current water trace has multiple downstream branches, the termination point of each branch is recorded separately, and no branch is discarded when generating sensitive areas.

[0082] Figure 5 This demonstrates the verification of the confluence of the current watermark with existing confirmed watermarks. The processing server first transforms the two watermarks to the same surface coordinate system and checks whether the wet areas are truly contiguous.

[0083] If the two watermarks share a common surface area, and this area connects the upstream continuous area of ​​the current watermark and the downstream continuous area of ​​the confirmed watermark, then a continuous wetting relationship is established. The location where the current watermark first enters the common area along the directed edge is recorded as the intersection location to be verified. If the watermarks only overlap in the image projection but do not connect after being transformed to surface coordinates, they will not proceed to the next step of convergence judgment.

[0084] Then, the surface elevation near the intersection location to be verified is read. Along the directed edge of the confirmed water trace, elevation sampling points are selected upstream and downstream of the intersection location, respectively, requiring that there be continuous wet areas between the upstream sampling point and the intersection location, and between the intersection location and the downstream sampling point.

[0085] When the elevation of the upstream sampling point is higher than that of the intersection point, and the elevation of the intersection point is higher than that of the downstream sampling point, the surface elevation is considered to decrease along the downstream direction of the confirmed watermark. If measurement fluctuations exist, the elevation difference between the two segments should also be greater than the elevation measurement error. The processing server only confirms the watermark convergence when both conditions—continuous humidity and a sequential decrease in elevation—are simultaneously met.

[0086] If the current watermark intersects with a confirmed watermark in the image but fails the elevation verification, the location is marked as a location to be verified, and the downstream part of the confirmed watermark is not associated with the current disturbance unit for the time being.

[0087] On-site personnel can supplement the elevation points at the intersection and add surface images, and the processing server will then re-evaluate based on the new data. This measure is used to distinguish between the actual confluence relationships of flows entering the same surface channel and the apparent intersections caused by shooting angles, vegetation obstruction, or drainage structures on different levels.

[0088] After the convergence is confirmed, the processing server traverses downstream from the intersection point along the directed edges of the confirmed water trails until the termination point. The directed edges and termination points reached during the traversal are marked with the engineering identifier of the current disturbance unit, while the original engineering identifiers are retained.

[0089] Therefore, the upstream paths from each disturbance unit to the intersection point are stored separately, while the shared downstream path from the intersection point to the termination point serves multiple sources simultaneously. After the association is completed, the processing server sends a water outage signal to the field control terminal.

[0090] When a current watermark does not intersect with a previous watermark, it is updated to a confirmed watermark after verifying surface continuity, downstream direction, and termination location. When a watermark merges with a previous watermark, it is updated to a confirmed watermark after verifying the merging relationship and associating it with a shared downstream path. In both cases, the processing server saves the corresponding directed path map and project identifier.

[0091] A water outage signal triggered by intersection determination includes at least the current project identifier, the current boundary water outlet location identifier, the intersection location, the reason for the water outage, and the signal generation time. The field control terminal checks the boundary water outlet location identifier in the signal against the address of the currently opened electrically controlled valve, closes the corresponding electrically controlled valve, and returns the closing time and the cumulative water output from the flow meter to the processing server.

[0092] like Figure 6As shown, after completing water trace reconstruction, confluence judgment, and disturbance unit association, the processing server generates disturbance source sensitive areas, runoff channel sensitive areas, and runoff receiving sensitive areas in the actual surface coordinate system.

[0093] The disturbance source sensitive area is directly taken from the actual land boundary of the disturbance unit corresponding to the confirmed water trace. The processing server queries the corresponding disturbance unit based on the engineering identifier saved in the water trace and writes its actual land boundary into the disturbance source sensitive area. When a disturbance unit has multiple boundary water outlet locations, the water traces formed at each location all point to the same disturbance source sensitive area, while each retains its own boundary water outlet location record.

[0094] The sensitive area of ​​the runoff channel consists of all newly added wetted areas between the boundary outlet position and each termination position. The processing server traverses from the source node to each termination position along the directed edges of the current water trail and the inherited confirmed water trails, merging the traversed wetted areas. When there are branches, all downstream branches are included; when there is a shared downstream path, the path retains the engineering identifiers of all upstream disturbance units. The boundary of the sensitive area is the actual extracted outer boundary of the wetted area, rather than a straight line between the source node and the termination position.

[0095] The runoff receiving sensitive zone is taken from the wet area that continues to expand from the termination point. The processing server uses the observation time when the current water trail stops extending downhill as the boundary, and extracts the wet portion from the image after that time that is continuous with the termination point but does not form a new downstream directed edge.

[0096] If this portion extends laterally around the termination point, extends into local low-lying areas, or continuously increases in area, the channel area already formed when the extension stops is deducted, and the remaining portion is designated as the runoff receiving sensitive area. Multiple branches with different termination points are extracted separately; adjacent extended areas can be merged, but their respective corresponding watermark branches are still retained.

[0097] If the cumulative outflow has reached the preset water supply volume, but the downhill front is still moving, the on-site control terminal keeps the electronically controlled valve closed, and the image terminal group continues to observe until the front stops moving or the moisture change is below the change threshold. If the termination position still cannot be determined after the observation, the already formed runoff channel sensitive area is saved first, and the runoff receiving sensitive area is generated after the termination position is obtained by supplementary experiments.

[0098] All three sensitive areas retain engineering identifiers, cumulative outflow volumes, and water trace relationships. The cumulative outflow volume includes both the total outflow at each boundary outflow location and the cumulative values ​​for each stage when the directed edges are formed. The water trace relationships sequentially record the connections from the disturbance unit to the boundary outflow location, from the boundary outflow location to the source node, from the source node to the branch or confluence location, from the confluence location to the termination location, and from the termination location to the runoff receiving sensitive area.

[0099] Therefore, from any sensitive area, its upstream disturbance unit, water supply records and complete water traces can be retrieved, and the three types of sensitive areas corresponding to the disturbance unit can also be retrieved from any engineering identifier.

[0100] The following is a field application example. The survey object is a power transmission tower foundation working face and its adjacent construction access road. After on-site verification, the two actual land occupation boundaries were established as the first disturbance unit and the second disturbance unit, respectively: the water outlet of the boundary on the downslope side of the first disturbance unit is connected to a shallow ditch outside the boundary, and the water outlet of the boundary on the downslope side of the second disturbance unit is connected to a rut guide strip leading to the shallow ditch.

[0101] The processing servers were configured with engineering identifiers, and the two locations were marked as the first and second water outlets according to the order in which the downstream watermarks formed. Three portable image terminals were deployed along the shallow ditch and the rut guide strip, with adjacent fields of view including the same ground calibration element.

[0102] Four water supply periods were first implemented at the first water outlet location. After the first three periods, the downhill front continued to move downstream, and the processing server returned to the continued extension state. After the fourth period, the movement distance of the downhill front in the two consecutive observation periods did not exceed the spatial judgment threshold, but the moist area near the termination location continued to expand towards the low-lying side of the ditch.

[0103] The on-site control terminal keeps the electric control valve at the first water outlet position closed; the processing server reconstructs the continuous wet area from the first water outlet position to the termination position as the first water trace, updates it as the confirmed water trace, and saves the area that continues to expand near the termination position as the first runoff receiving sensitive area.

[0104] The system then switches to the second water outlet location of the second disturbance unit. The processing server uses the last frame image after the water supply to the first water outlet location stops as the current reference image, and first subtracts the surface area that has already been moistened by the first water mark. The newly moistened area formed by the second water outlet location is then connected to the first water mark after two complete water supply periods.

[0105] The processing server uses a common ground calibration component to complete the position alignment and issues a water stop signal when the two water tracks form a continuous wet area. It then verifies that the elevations of the upstream, intersecting, and downstream positions decrease sequentially, confirms that the second water track merges into the first water track, and simultaneously associates the downstream portion of the first water track from the intersecting position to the termination position with the engineering identifiers of the two disturbance units.

[0106] After the tests at the two boundary outlet locations are completed, the actual land boundaries of the first and second disturbance units respectively form the disturbance source sensitive areas; all the newly added wet areas extending from the first and second outlet locations to the common termination location form the runoff channel sensitive area; and the wet area that continues to expand at the common termination location forms the runoff receiving sensitive area.

[0107] The results data save the cumulative discharge volume, independent upstream water trace, shared downstream water trace, and termination location of the two discharge locations, thereby distinguishing the two sources of disturbance and their common downstream receiving area.

[0108] The above values ​​and deployment methods can be adjusted according to site conditions. For example, the water supply volume, number of water supply periods, image acquisition interval, spatial judgment threshold, and observation duration can be determined in conjunction with the scale of the construction site, surface slope, soil infiltration status, and image surface resolution. The serial numbers "first," "second," etc., in this text are only used to distinguish similar objects and do not indicate quantity, priority, or fixed chronological order.

[0109] Without altering the above-described technical concept, those skilled in the art can make equivalent substitutions or conventional adjustments to the specific implementation based on the on-site equipment and terrain conditions; all corresponding changes falling within the scope of the claims are within the scope of protection of this invention.

Claims

1. A rapid on-site identification system for soil ecological sensitivity zoning in power transmission and transformation projects, characterized in that, This includes a processing server, a field control terminal, a water supply terminal group, and an image terminal group; The processing server divides the disturbance units according to the actual land area boundary of the project, configures the project identifier for each disturbance unit, and determines the corresponding boundary water outlet position. The water supply terminal group includes water supply terminals respectively set at each of the boundary water outlet positions. Each water supply terminal is equipped with a flow meter, an electric control valve and a water outlet. The field control terminal opens the electric control valve one by one in a predetermined order. The image terminal group continuously acquires images of the downhill soil surface before and after water supply at the current boundary water outlet location. The processing server extracts newly added wet areas connected to the current boundary water outlet location from the downhill soil surface image and reconstructs the newly added wet areas as the current water trail; when the current water trail stops extending downhill or intersects with a confirmed water trail, the processing server sends a water stop signal and determines the termination position of the water trail; when the current water trail intersects with the confirmed water trail, the processing server also associates the confirmed water trail downstream of the intersection position with the current disturbance unit; After receiving the water outage signal, the field control terminal closes the currently open electrically controlled valve; The processing server generates a disturbance source sensitive area, a runoff channel sensitive area, and a runoff receiving sensitive area based on the disturbance unit, the associated water trace, and the termination position of the water trace.

2. The on-site rapid identification system for soil ecological sensitivity zoning in power transmission and transformation projects according to claim 1, characterized in that, The processing server is equipped with a server processor, a server memory, and a server communication interface; the field control terminal is equipped with a terminal processor, a terminal memory, and a terminal communication interface. The server communication interface is connected to the terminal communication interface and the image terminal group, respectively, and the terminal communication interface is connected to the flow meter and the electric control valve, respectively.

3. The on-site rapid identification system for soil ecological sensitivity zoning in power transmission and transformation projects according to claim 1, characterized in that, The processing server forms disturbance units according to the actual land boundaries of the tower foundation working face, tension working face, construction access road and substation working face; each project identifier is determined by the project type, construction period and spatial coordinates of the corresponding disturbance unit; the processing server also associates the surface runoff traces that cross the actual land boundaries with the corresponding project identifiers.

4. The on-site rapid identification system for soil ecological sensitivity zoning in power transmission and transformation projects according to claim 3, characterized in that, The processing server first determines the junction of the surface runoff trace and the downslope side of the actual land occupation boundary as the candidate water outlet location, and then selects the location from the candidate water outlet locations that is connected to the surface groove, rut guide strip or drainage outlet outside the boundary as the boundary water outlet location. The field control terminal records the cumulative water output fed back by the flow meter at the corresponding boundary water outlet position.

5. The on-site rapid identification system for soil ecological sensitivity zoning in power transmission and transformation projects according to claim 1, characterized in that, The image terminal group includes multiple portable image terminals and ground calibration components; each of the portable image terminals is arranged in segments along the downhill direction of the boundary water outlet position, and the shooting range of adjacent portable image terminals overlaps with the area where the ground calibration component is located. The portable image terminal records the planar position and elevation of the ground calibration component, and acquires continuous downhill soil surface images with timestamps generated by the field control terminal.

6. The on-site rapid identification system for soil ecological sensitivity zoning in power transmission and transformation projects according to claim 1, characterized in that, The field control terminal allocates the preset water supply volume of the current boundary water outlet position to multiple water supply periods; after each water supply period is completed, the field control terminal closes the current electric control valve and receives the water trace extension status returned by the processing server; only when the newly added wet area is still expanding downhill and has not yet touched the confirmed water trace, the field control terminal reopens the current electric control valve to execute the next water supply period.

7. The on-site rapid identification system for soil ecological sensitivity zoning in power transmission and transformation projects according to claim 6, characterized in that, When the field control terminal switches from the current boundary water outlet position to the subsequent boundary water outlet position, the processing server sets the last frame image after the current boundary water outlet position stops supplying water as the current reference image. The processing server removes the already moistened surface areas from the downhill soil surface images obtained during subsequent water supply periods, and identifies the remaining and continuous moist areas as the newly added moist areas corresponding to the subsequent boundary water outlet positions.

8. The on-site rapid identification system for soil ecological sensitivity zoning in power transmission and transformation projects according to claim 5, characterized in that, The processing server uses the ground calibration component, which is shared by the shooting range of adjacent portable image terminals, to align the surface position of each newly added wet area. The processing server merges the newly added wet areas with continuous acquisition time and contiguous surface positions into the current water trail, and saves the current water trail as a directed path graph. The directed path graph uses the boundary water outlet position as the source node and the branch position, confluence position and termination position as connection nodes.

9. The on-site rapid identification system for soil ecological sensitivity zoning in power transmission and transformation projects according to claim 5, characterized in that, When the current water mark and the confirmed water mark form a continuous wet area, and the surface elevation on both sides of the intersection point decreases sequentially in the downstream direction of the confirmed water mark, the processing server confirms that the two water marks have merged; the processing server associates the downstream water mark from the intersection point to the termination point of the confirmed water mark with the engineering identifier corresponding to the current disturbance unit, and sends the water outage signal to the field control terminal.

10. The on-site rapid identification system for soil ecological sensitivity zoning in power transmission and transformation projects according to claim 4, characterized in that, The processing server determines the actual boundary of the disturbance unit corresponding to the confirmed water trace as the disturbance source sensitive area, determines all newly added wet areas extending from the boundary water outlet position to the water trace termination position as the runoff channel sensitive area, and determines the wet area that continues to expand at the termination position after the current water trace stops extending downhill as the runoff receiving sensitive area; each sensitive area stores the corresponding engineering identifier, cumulative water discharge, and water trace association relationship from the boundary water outlet position to the water trace termination position.

Images