Image processing-based power construction cable laying path optimization method and system

Abstract

The application discloses a power construction cable laying path optimization method and system based on image processing, and relates to the technical field of image processing, and comprises the following steps: acquiring multi-angle images of a construction trench, extracting a sediment spatial distribution boundary, and obtaining a sediment feature set; performing connectivity analysis on the sediment feature set to obtain a sediment continuous feature set; dividing the spatial range of the sediment continuous feature set, analyzing a sediment height gradient, and forming a height change feature set; performing constraint analysis on the sediment continuous feature set and the sediment height change feature set, generating a sediment constraint matrix, and extracting a constraint feature; dividing a preset cable path based on the constraint matrix to obtain a first path set; calculating the distribution proportion of the constraint feature in the first path set and adjusting the path according to the distribution proportion to obtain a second path set, so that the potential scouring risk and structural damage problem caused by the interaction between the continuous accumulation of sediments and the local height mutation of the sediments can be solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and more particularly, to a power construction cable laying path optimization method and system based on image processing. BACKGROUND

[0002] In the process of power engineering construction, for the laying path of the cable from the starting point to the end point, the construction environment, topography, obstacle distribution, construction cost, safety and post-maintenance and other factors are comprehensively considered, and the cable laying route is reasonably planned and selected through scientific methods and technical means, so as to realize the maximization of resource utilization and the minimization of construction risk. The core goal is to avoid unnecessary bending, detouring and crossing, reduce the cable length and construction difficulty, reduce the amount of civil excavation and engineering cost, and at the same time ensure that the path meets the electrical performance and safety specification requirements. In recent years, with the development of image processing and intelligent analysis technology, the construction site environment information can be automatically extracted through image acquisition, recognition and analysis, and combined with optimization algorithm to generate more efficient and reasonable cable laying path, thereby significantly improving the construction efficiency and intelligent level.

[0003] However, in the process of power construction cable laying, there are different forms and distributions of sediments at the bottom of the construction trench or pipe. These sediments may present continuous accumulation fragments or form isolated accumulation points. Continuous accumulation fragments will form a stable but extensive stress area under the action of water flow or seepage, while local isolated accumulation points with high mutation may cause local acceleration or change of water flow, thereby producing local scouring similar to "water knife effect". The interaction between the continuous fragments and local protrusions of the height variation not only aggravates the local movement and redistribution of sediments, but also may cause structural damage to the trench or cable sheath, thereby affecting the stability and construction safety of the cable. Traditional construction path planning usually cannot consider the coupling effect between the spatial continuity and height mutation of sediments at the same time, thereby existing potential construction risks and structural damage hazards.

[0004] In view of the above problems, the present application provides a solution. SUMMARY

[0005] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present application provide a power construction cable laying path optimization method and system based on image processing, which analyzes the continuity and height variation characteristics of sediments in the construction trench to solve the problem of potential scouring risk and structural damage of the construction path caused by the interaction between the continuous accumulation fragments and local height mutation of sediments.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0007] The power construction cable laying path optimization method based on image processing comprises the following steps: acquiring multi-angle images of a construction trench, extracting the spatial distribution boundary of sediments based on the multi-angle images to obtain a sediment feature set; performing connectivity analysis on the sediment feature set to obtain a sediment continuous feature set;

[0008] dividing the spatial range of the sediment continuous feature set and analyzing the height gradient of the sediments in the divided spatial range to obtain a sediment height variation feature set; performing constraint analysis on the sediment continuous feature set and the sediment height variation feature set to obtain a sediment constraint matrix and extract constraint features in the sediment constraint matrix; dividing a preset cable path based on the sediment constraint matrix to obtain a first cable path set; calculating the distribution proportion of the constraint features in the first cable path set and adjusting the first cable path set based on the distribution proportion to obtain a second cable path set.

[0009] In a preferred embodiment, the multi-angle images of the construction trench are acquired, and the spatial distribution boundary of the sediments is extracted based on the multi-angle images to obtain a sediment feature set, specifically: multi-angle images of the construction trench are collected by a multi-view camera device deployed above the construction trench, and distortion correction and illumination equalization processing are performed on the multi-angle images to obtain a multi-angle image set; depth information recovery is performed on the multi-angle image set, and a three-dimensional point cloud model of the construction trench is constructed; surface data of the sediments is extracted based on the three-dimensional point cloud model of the construction trench; a morphological segmentation algorithm is used to perform region growing on the surface data of the sediments to obtain the spatial distribution boundary of the sediments; geometric features of the sediments are extracted based on the spatial distribution boundary of the sediments to obtain the sediment feature set.

[0010] In a preferred embodiment, the connectivity analysis on the sediment feature set to obtain a sediment continuous feature set is specifically: the sediment feature set is discretized into a uniform grid, and a sediment distribution binary matrix is constructed; a region growing algorithm is used to traverse the sediment distribution binary matrix to identify connected sediment regions to obtain a plurality of connected regions; geometric features of each connected region are extracted to obtain the sediment continuous feature set.

[0011] In a preferred embodiment, the spatial range of the sediment continuous feature set is divided, specifically: the sediment continuous feature set is spatially grid-divided to obtain a plurality of grid cells; the average height and height variance of the sediments in each grid cell are calculated, and region merging is performed based on the height variance to obtain a plurality of sub-regions; the spatial range boundary of each sub-region is calculated to obtain a divided spatial range set.

[0012] In a preferred implementation, the height gradient of the deposit in each divided spatial range is analyzed to obtain a deposit height variation feature set, specifically: in each divided spatial range in the set of divided spatial ranges, deposit height data in each divided spatial range is obtained; the deposit height data in each divided spatial range is subjected to surface fitting to obtain a height gradient vector, and the amplitude of the height gradient vector is calculated; and the deposit height variation feature set is constructed based on the amplitude of the height gradient vector.

[0013] In a preferred implementation, the deposit continuity feature set and the deposit height variation feature set are subjected to constraint analysis to obtain a deposit constraint matrix and extract constraint features in the deposit constraint matrix, specifically: the deposit continuity feature set and the deposit height variation feature set are subjected to feature alignment to obtain first features of each spatial position; feature difference evaluation is performed between each spatial position based on the first features of each spatial position; a constraint matrix is constructed based on the feature difference evaluation results; the deposit constraint matrix is subjected to singular value decomposition to extract a singular vector corresponding to a maximum singular value; the spectral radius and the condition number of the deposit constraint matrix are calculated based on the maximum singular value; the deposit constraint matrix is subjected to structural analysis based on the spectral radius and the condition number of the deposit constraint matrix to obtain a constraint cluster; and features of the constraint cluster are extracted based on the singular vector to obtain the constraint features in the deposit constraint matrix.

[0014] In a preferred implementation, the preset cable path is divided based on the deposit constraint matrix to obtain a first cable path set, specifically: a coordinate sequence of the preset cable path is obtained, and the preset cable path is discretized into a plurality of path points based on the coordinate sequence; the plurality of path points are clustered into a plurality of path segments according to the values of the deposit constraint matrix; the plurality of path segments are subjected to smooth reconstruction by a B-spline curve to obtain the first cable path set.

[0015] In a preferred implementation, the distribution proportion of the constraint features in the first cable path set is calculated, and the first cable path set is adjusted based on the distribution proportion to obtain a second cable path set, specifically: for each path segment in the first cable path set, a distribution density function of the constraint features along the length of each path is calculated; a proportion threshold of the constraint features is determined based on the distribution density function to obtain the distribution proportion; and the positions and orientations of the path segments in the first cable path set are dynamically adjusted according to the distribution proportion to obtain the second cable path set.

[0016] The power construction cable laying path optimization method and system based on image processing have the following technical effects and advantages:

[0017] The application forms a sediment feature set by acquiring multi-angle images of the construction trench and extracting the spatial distribution boundary of the sediment; performs connectivity analysis on the feature set, identifies the continuous sediment region, and divides the spatial range thereof, while analyzing the height gradient of the sediment in each spatial range to form a sediment height variation feature set; then performs constraint analysis on the sediment continuity feature set and the height variation feature set to obtain a sediment constraint matrix and extract constraint features, and divides and adjusts the preset cable path based on the constraint matrix, calculates the distribution proportion of the constraint features, and dynamically optimizes the path, to finally obtain the optimized cable laying path. The method can accurately identify the continuous segment of the sediment accumulation and the local height mutation region, and significantly reduce the erosion risk and structural damage that the cable may suffer in the laying process by analyzing and adjusting the distribution proportion and spatial correlation of the constraint features, to ensure the smoothness and construction safety of the path. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 FIG. 1 is a flowchart of the image processing-based power construction cable laying path optimization method of the application.

[0019] Figure 2 FIG. 2 is a structural diagram of the image processing-based power construction cable laying path optimization system of the application. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0021] Embodiment 1, Figure 1 The image processing-based power construction cable laying path optimization method of the application is given, which comprises the following steps:

[0022] S1, acquiring multi-angle images of the construction trench, and extracting the spatial distribution boundary of the sediment based on the multi-angle images to obtain a sediment feature set;

[0023] In this example, multi-angle images of the construction trench are acquired, and the spatial distribution boundary of the sediment is extracted based on the multi-angle images to obtain a sediment feature set, specifically as follows:

[0024] Multi-angle images of the construction trench are collected by multi-view camera equipment deployed above the construction trench, and distortion correction and illumination equalization processing are performed on the multi-angle images to obtain a multi-angle image set;

[0025] Depth information recovery is performed on the multi-angle image set, and a three-dimensional point cloud model of the construction trench is constructed;

[0026] Surface data of the sediment is extracted based on the three-dimensional point cloud model of the construction trench;

[0027] Regional growth is performed on the surface data of the sediment using a morphological segmentation algorithm to obtain a spatial distribution boundary of the sediment;

[0028] Geometric features of the sediment are extracted based on the spatial distribution boundary of the sediment to obtain a sediment feature set.

[0029] It should be noted that the multi-angle images of the construction trench refer to a set of images obtained by imaging the same trench area from different angles through multi-view camera devices (such as fixed cameras, drones, or multi-lens arrays) arranged above or around the trench. These images not only include longitudinal and lateral views of the trench, but also may include oblique and overhead views to ensure that the internal sediment and the trench wall can be completely captured. The purpose of obtaining multi-angle images is to overcome the information missing problem under a single view, so that the subsequent trench three-dimensional morphology and sediment distribution can be more accurately recovered. These images are usually subjected to distortion correction and illumination equalization processing to reduce the interference caused by lens distortion and uneven on-site lighting conditions, thereby ensuring the usability and accuracy of the data.

[0030] Secondly, depth information recovery is achieved based on the principle of multi-view stereo vision. First, feature point extraction and matching are performed on the multi-angle image set to identify corresponding pixel points of the same object surface in different images. Then, multi-view geometry (such as the triangulation method) is used to calculate the depth coordinates of these pixel points in space, thereby obtaining the three-dimensional spatial structure of the scene. This process can use algorithms such as structured light method, SfM (structure from motion), or multi-view stereo reconstruction (MVS). After completing the depth information recovery, a dense three-dimensional point set of the trench surface and sediment is obtained, and these points have (x, y, z) coordinates in space, i.e., a three-dimensional point cloud model of the construction trench is constructed. The point cloud model can completely describe the geometric shape of the trench and the accumulation position and height variation of the sediment in the trench, and is the core data basis for subsequent analysis.

[0031] Further, in the three-dimensional point cloud model, each point corresponds to a surface position in the trench scene, including the bottom of the trench, the side wall, and the surface of the deposit. In order to extract the surface data of the deposit, it is necessary to first distinguish the deposit from the trench boundary or the bottom soil in the point cloud by semantic segmentation or clustering algorithm. Color features, surface normal vector changes, and point height differences are usually combined to identify which points belong to the deposit. The extracted deposit surface data is a set of point clouds that constitute the external boundary and surface profile of the deposit, which describes the shape, distribution range, and surface ups and downs of the deposit in the trench. Simply put, the surface data of the deposit is a three-dimensional point cloud of the visible part of the deposit exposed to the air, which reflects the direct impact of the deposit on the cable laying space.

[0032] Then, the morphological segmentation algorithm is a kind of segmentation method based on the geometric structure of the image or point cloud, and the common operations include dilation, erosion, opening operation, closing operation, etc. For the surface point cloud data of the deposit, it is first mapped to a regular grid or a projection plane to form a dense two-dimensional height map or a voxel representation. Then, through morphological filtering, isolated points and noise points are removed, and the main structure is retained. Then, a region growing algorithm is used: starting from a seed point, according to the similarity of the neighboring points and the current region in height difference, normal vector direction, or spatial distance, the region is gradually expanded until the similarity condition is not met, thereby dividing the complete deposit region. The final obtained deposit spatial distribution boundary is the contour range of the deposit separated from the surrounding environment in three-dimensional space, which is used to accurately define the influence area of the deposit on the construction channel.

[0033] Finally, after obtaining the spatial distribution boundary of the deposit, the geometric features of the deposit can be further extracted to form a feature set for the path optimization model. The geometric features are as follows:

[0034] Shape features: such as the length, width, perimeter, area, and volume of the deposit boundary; height features: including the average height, maximum height, minimum height, and height variance of the deposit; surface features: such as surface roughness, curvature, and inclination angle; spatial distribution features: such as the position coordinates of the deposit in the trench, the occupied space range, and the distance between adjacent deposits; directional features: the angle between the long axis direction of the deposit and the trend of the trench. These geometric features comprehensively reflect the shape and spatial occupation of the deposit in the trench, forming the deposit feature set. In cable path optimization, these features can be used to determine which areas have more deposits and are more rugged, thereby providing quantitative basis for the feasibility and laying difficulty of the cable path.

[0035] S2, performing connectivity analysis on the deposit feature set to obtain a continuous deposit feature set;

[0036] In the present example, a connectivity analysis is performed on the sediment feature set to obtain a set of sediment continuous features, specifically:

[0037] The sediment feature set is discretized into a uniform grid, and a sediment distribution binary matrix is constructed.

[0038] A region growing algorithm is used to traverse the sediment distribution binary matrix to identify connected sediment regions, resulting in a number of connected regions.

[0039] The geometric features of each connected region are extracted to obtain the set of sediment continuous features.

[0040] It should be noted that first, a uniform grid is established in the three-dimensional space of the trench (it can be a two-dimensional projection grid or a three-dimensional voxel grid), and then each point in the sediment feature set is projected into the corresponding grid cell; if there is a sediment point in a grid cell, it is assigned a value of "1", indicating that the cell is occupied by sediment; if there is no sediment point in the cell, it is assigned a value of "0", indicating that it is empty. In this way, a matrix composed of "0" and "1" is obtained, which is the sediment distribution binary matrix. It is actually a discretized representation of the spatial distribution of sediment, where the "1" aggregation region corresponds to the sediment entity, and the "0" corresponds to the void or non-sediment region.

[0041] Further, after obtaining the sediment distribution binary matrix, a region growing algorithm can be used for traversal and segmentation. The specific method is: starting from an arbitrary "1" value cell in the matrix as a seed point, checking the cells above, below, left and right (in two-dimensional case) or surrounding 26 neighbors (in three-dimensional voxel case), if the neighbor cell is also "1", it will be included in the same region, and continue to expand to its neighbors, until there is no new neighbor that meets the conditions. Through this step-by-step expansion, all "1" regions in the matrix can be divided out, and each set of mutually connected "1"s is a connected region. The so-called connected sediment region refers to the adjacent and uninterrupted sediment entity in space, which physically represents a whole sediment accumulation. In this way, the scattered or continuous sediment in the trench can be distinguished, resulting in a number of clear sediment connected regions.

[0042] Finally, after identifying several connected regions, it is necessary to further quantify the morphology and spatial characteristics of each region as a constraint condition in cable path optimization. The specific method is: statistics of point cloud or grid cells in each connected region is performed to extract the geometric features of the region. For example: 1. Volume feature, calculate the number of all cells or point cloud voxels in the region to estimate the volume of the sediment; 2. Size feature, including maximum length, maximum width, thickness or height range; 3. Morphology feature, such as the perimeter of the boundary, area, compactness, aspect ratio; 4. Position feature, i.e. the absolute coordinate range of the connected region in the trench (for example, the height from the bottom of the trench, the horizontal distance from the boundary of the trench); 5. Surface feature, such as the roughness of the region surface, inclination angle, etc. These geometric features together constitute the continuous feature set of the sediment. The continuous feature set can more completely describe the distribution state and physical characteristics of the sediment in the trench, providing important spatial constraint information for cable path selection, such as avoiding the cable passing through the area with the most dense sediment accumulation or the most dramatic height change.

[0043] S3, dividing the spatial range of the continuous feature set of the sediment, and analyzing the height gradient of the sediment in the divided spatial range to obtain a height variation feature set of the sediment;

[0044] In this example, the spatial range of the continuous feature set of the sediment is divided, specifically:

[0045] The continuous feature set of the sediment is spatially grid divided to obtain a plurality of grid cells;

[0046] The average height and height variance of the sediment in each grid cell are calculated, and the regions are merged based on the height variance to obtain a plurality of sub-regions;

[0047] The spatial range boundary of each sub-region is calculated to obtain a set of divided spatial ranges.

[0048] It should be noted that in the construction trench, the continuous feature set of the sediment refers to the continuous sediment region and its geometric features that have been identified. In order to perform more detailed height analysis, the continuous sediment region needs to be discretized into regular spatial cells. The specific method is: a uniform three-dimensional grid or two-dimensional projection grid is established in the three-dimensional spatial range of the trench, and the continuous sediment region is covered on the grid, and each grid cell records the sediment point information in the corresponding region. In this way, the originally continuous sediment block is divided into several small grid cells, each grid cell becomes an independent analysis unit, providing a basis for subsequent height calculation and region merging.

[0049] Secondly, in each grid cell, there are a plurality of sediment point cloud data, and the z coordinate of each point represents the height of the sediment. The average height is obtained by summing the z coordinates of all points in the grid cell and dividing by the number of points, reflecting the overall sediment height of the grid cell. The height variance is obtained by calculating the square sum of the difference between the height of each point and the average height and dividing by the number of points, which is used to measure the degree of fluctuation or concave-convex change of the sediment surface in the unit. The average height and the height variance not only quantitatively describe the spatial distribution of the sediment in the grid cell, but also provide key parameters for subsequent regional smoothing and cable path planning.

[0050] Further, after completing the height statistics of each grid cell, the units with relatively uniform height changes can be merged into larger regions according to the similarity of the height variances of adjacent grid cells. The specific method is: traverse all grid cells, and group the units with height variances within a certain threshold range and spatially adjacent into the same sub-region, and keep independent if the height difference is too large. In this way, the originally discrete or too fine grid can be integrated into several sub-regions. At this time, the sub-region represents a spatial block with relatively consistent and continuous sediment height change in the trench, and each sub-region can be regarded as an independent sediment obstacle unit in the cable path analysis, which is used to evaluate the path feasibility and obstacle avoidance strategy.

[0051] Finally, after the sub-regions are formed, the actual occupied space of each sub-region in the trench needs to be determined. The specific method is: the x, y, z coordinates of all grid cells or point cloud points in the sub-region are counted to determine the maximum and minimum coordinate values, thereby obtaining the boundary range along the three-dimensional coordinate axes; further, a convex hull or boundary envelope algorithm can be used to generate the spatial profile of the sub-region. By performing such boundary calculation on all sub-regions, a complete set of divided spatial ranges is obtained. This set of spatial ranges accurately describes the spatial position, size and occupied range of each sediment sub-region in the trench, and provides accurate spatial constraint information for subsequent sediment height gradient analysis and cable path optimization.

[0052] In this example, the height gradient of the sediment in the divided spatial range is analyzed to obtain a set of sediment height change characteristics, which are specifically:

[0053] In each divided spatial range in the set of divided spatial ranges, the sediment height data in each divided spatial range is obtained;

[0054] The sediment height data in each divided spatial range is subjected to surface fitting to obtain a height gradient vector, and the amplitude of the height gradient vector is calculated;

[0055] Based on the amplitude of the height gradient vector, a set of sediment height change characteristics is constructed.

[0056] It should be noted that the spatial range of each sub-region has a clear boundary determined by grid division and region merging. When obtaining the sediment height data within this boundary, the z coordinates (vertical direction coordinates) of all sediment point cloud points or grid cells falling within the sub-region need to be extracted, and these height values constitute the sediment height data set. The sediment height data specifically refers to the sediment accumulation height at each location within the sub-region, reflecting the height variation of the sediment surface in the vertical direction.

[0057] In each sub-region, the discrete sediment height data collected is used for surface fitting, and methods such as polynomial fitting, B-spline surface, or least squares plane fitting can be used to fit the discrete points into a continuous surface z = f(x, y). After fitting the surface, the height gradient vector at each point can be calculated, representing the rate of change of height with spatial position. The calculation method of the height gradient vector is to find the partial derivatives of the surface in the x and y directions, i.e. G(x, y) = (∂z / ∂x, ∂z / ∂y), which reflects the direction and size of the sediment surface slope at that point. Square the partial derivatives in the x and y directions respectively, and add the two squared values, then take the square root of the sum, to get the magnitude of the height gradient vector at that point. The larger the vector magnitude, the more abrupt the change in the sediment surface at that location, and the more significant the potential obstacle to cable laying.

[0058] Finally, the steps to construct the sediment height variation feature set based on the height gradient vector magnitude are as follows: sample or map the surface height gradient magnitude in the sub-region to discrete grid cells at a fixed interval; calculate statistical indicators such as mean, maximum, and variance of the height gradient magnitude in each grid cell to quantify the local surface variation intensity; divide the height gradient magnitude into different levels (such as low, medium, and high) according to a pre-set threshold to distinguish between gentle and steep regions; bind the height variation level of each grid cell or sampling point to its spatial position to form a structured feature representation; integrate the height variation features of all sub-regions to generate a complete sediment height variation feature set.

[0059] S4, constraint analysis is performed on the sediment continuous feature set and the sediment height variation feature set to obtain a sediment constraint matrix and extract constraint features in the sediment constraint matrix;

[0060] In this example, constraint analysis is performed on the sediment continuous feature set and the sediment height variation feature set to obtain a sediment constraint matrix and extract constraint features in the sediment constraint matrix, specifically as follows:

[0061] The sediment continuous feature set and the sediment height variation feature set are aligned to obtain a first feature at each spatial position;

[0062] Based on the first feature at each spatial position, a feature difference evaluation is performed between each spatial position.

[0063] constructing a constraint matrix based on the feature difference evaluation result;

[0064] performing singular value decomposition on the sediment constraint matrix to extract a singular vector corresponding to a maximum singular value;

[0065] calculating a spectral radius and a condition number of the sediment constraint matrix based on the maximum singular value;

[0066] performing structural analysis on the sediment constraint matrix based on the spectral radius and the condition number of the sediment constraint matrix to obtain a constraint cluster;

[0067] extracting features of the constraint cluster based on the singular vector to obtain constraint features in the sediment constraint matrix.

[0068] It should be noted that in the construction trench, the sediment continuity feature set describes the spatial distribution and geometric characteristics of the sediment, and the sediment height variation feature set describes the steepness and undulation of the sediment surface. In order to comprehensively utilize these two types of information, it is necessary to align them in the same spatial coordinate system. The specific method is as follows: first, map the continuity feature set and the height variation feature set to the same grid or sampling point, then in each grid element or sampling point position, combine the sediment geometric features (such as volume, length, width, surface roughness) with the corresponding height variation features (such as height gradient amplitude, steepness level) to form a vector, this combined vector is the first feature of the spatial position. The first feature of each spatial position reflects the geometric shape and surface undulation of the sediment at that point, providing basic information for subsequent constraint analysis.

[0069] In addition, the purpose of feature difference evaluation is to quantify the difference degree of sediment features at different spatial positions. The steps are as follows: traverse all grid elements or sampling points to form a set of spatial position pairs; for each pair of positions, compare the numerical difference of each dimension in their first feature vectors, such as height gradient difference, volume difference, roughness difference, etc., the total difference value can be calculated using Euclidean distance, weighted Euclidean distance or Manhattan distance; normalize or standardize all difference values to eliminate dimensional influence; record the difference value of each pair of spatial positions in the matrix to provide basic data for constraint matrix construction.

[0070] Further, the constraint matrix is a matrix that reflects the potential constraint relationship of the sediments on the cable path at the spatial position level. The specific method is: taking the spatial position as the row and column of the matrix, and storing the feature difference value or the constraint strength calculated based on the difference of the corresponding position pair in each element. For example, the larger the difference value indicates that the feature difference of the two positions is large, which may form a strong constraint on the path; the smaller the difference value, the weaker the constraint. The constraint matrix specifically includes: matrix dimension (corresponding to the number of spatial positions), matrix element value (indicating constraint strength or feature difference), which reflects the potential influence relationship of different sediment regions in the construction trench on the cable laying.

[0071] Then, singular value decomposition (SVD) is a matrix decomposition method that decomposes the constraint matrix into the product of three matrices: left singular vector matrix, diagonal singular value matrix and right singular vector matrix. The specific method is: decompose the constraint matrix M to get M = UΣVᵀ, where the diagonal elements of Σ are singular values, sorted by size. Select the singular vector corresponding to the largest singular value (usually take the left singular vector or the right singular vector), which represents the most significant mode or direction in the constraint matrix, that is, the most important constraint relationship.

[0072] In addition, the spectral radius is the absolute value of the largest eigenvalue of the constraint matrix, and the condition number is the ratio of the largest singular value to the smallest singular value of the matrix. The specific method is: calculate the eigenvalue or singular value of the constraint matrix; the largest singular value is used for spectral radius calculation; the condition number = the largest singular value / the smallest singular value, which is used to measure the stability or solvability of the matrix. The spectral radius and the condition number can be used to analyze the potential strong constraint region and sensitive region in the constraint matrix.

[0073] Finally, the structure analysis determines the sediment combination that has the greatest impact on the cable path by identifying the constraint aggregation area with high constraint strength in the matrix. The specific method is: combined with the spectral radius and condition number information, find the sub-matrix or block with high value density in the matrix, which indicates that the constraint relationship between these spatial positions is strong and interrelated. Each constraint aggregation block is a group of adjacent or similar sediment regions in space, which form a common constraint on the cable path and usually need to be avoided or adjusted in path planning.

[0074] The singular vector corresponding to the largest singular value reflects the most significant constraint mode in the constraint matrix. The extraction method is: analyze the singular vector component size, mark the sediment regions corresponding to the spatial positions with larger component values as the main constraint region; summarize the geometric features and height variation characteristics of these regions to form the feature description of the constraint aggregation block. Finally, these features constitute the constraint features in the sediment constraint matrix, which are used to guide the optimization of the cable path to ensure that the path avoids the most significant sediment constraint region, reducing the construction difficulty and risk.

[0075] S5, dividing the preset cable path based on the sediment constraint matrix to obtain a first cable path set;

[0076] In this example, the preset cable path is divided based on the sediment constraint matrix to obtain a first cable path set, specifically:

[0077] The coordinate sequence of the preset cable path is obtained, and the preset cable path is discretized into a plurality of path points based on the coordinate sequence;

[0078] The plurality of path points are clustered into a plurality of path segments according to the values of the sediment constraint matrix;

[0079] The plurality of path segments are smoothed and reconstructed by B-spline curve to obtain the first cable path set.

[0080] It should be noted that in power construction, the preset cable path is usually determined by a design scheme or construction plan, indicating the ideal direction of the cable from the starting point to the ending point. The way to obtain the coordinate sequence of the preset path can be to extract the path points from the construction design drawings, or to obtain the three-dimensional coordinates of the path through CAD models or GIS information systems. The coordinate sequence specifically refers to the spatial coordinate set of each discrete position on the path, each coordinate contains x, y, z three-dimensional information, reflecting the absolute position of the point in the construction trench. In order to perform numerical analysis and optimization, it is necessary to discretize the continuous preset path, that is, to generate a plurality of path points on the path according to a fixed interval or key nodes, each path point corresponds to a three-dimensional coordinate, providing basic data for subsequent analysis combined with the sediment constraint matrix.

[0081] Further, after obtaining the discrete path points, it is necessary to evaluate the construction constraints of each point on the path in combination with the sediment constraint matrix. The specific method is: traversing the path points, extracting the constraint strength corresponding to each path point in the sediment constraint matrix, and then clustering according to the size of the constraint value and the spatial proximity. Generally, path points with similar constraint values and spatial proximity are classified into the same path segment, while points with significantly different constraint strengths are used as the boundary between segments. In this way, the originally continuous cable path is divided into a plurality of path segments, and the constraint characteristics inside each path segment are relatively uniform, providing a unitized processing basis for subsequent smoothing and optimization.

[0082] Finally, after the path segment division, the discrete path points need to be smoothed to ensure the continuity and smoothness of the cable laying path. The specific method is to apply B-spline curve fitting to the path points of each path segment, and by selecting appropriate control points and curve orders, the discrete path points are smoothly connected to form a continuous curve. B-spline curve has local controllability and smoothness, which can avoid sharp turning points or unreasonable twists in the path, while preserving the key positions passed by the path. After smoothing and reconstructing each path segment, all path segments are combined to obtain the complete first cable path set. This path set has considered the sediment distribution constraint while maintaining the feasibility of construction, providing a basis for further optimization and adjustment.

[0083] S6, calculating the distribution proportion of the constraint feature in the first cable path set and adjusting the first cable path set based on the distribution proportion to obtain a second cable path set.

[0084] In this example, the distribution proportion of the constraint feature in the first cable path set is calculated and the first cable path set is adjusted based on the distribution proportion to obtain a second cable path set, specifically:

[0085] For each path segment in the first cable path set, the distribution density function of the constraint feature along the length of each path is calculated;

[0086] Based on the distribution density function, the proportion threshold of the constraint feature is determined to obtain the distribution proportion;

[0087] According to the distribution proportion, the position and direction of the path segment in the first cable path set are dynamically adjusted to obtain a second cable path set.

[0088] It should be noted that in the first cable path set, each path segment contains a number of discrete path points, and each path point corresponds to a sediment constraint feature in space. When calculating the distribution density function of the constraint feature, statistical analysis of the constraint feature along the length of the path segment is required. The specific method is: divide the path segment into continuous small intervals or equidistant sampling points, and count the number or intensity of the constraint features in each interval, then divide by the length of the interval to obtain the density value of each interval. Arrange the density values of all intervals along the path segment to form the distribution density function. The distribution density function reflects the spatial distribution of the constraint feature on the path segment, i.e., where the sediment has a greater impact on cable laying and has a high density, and where the impact is smaller and the density is lower.

[0089] Further, the purpose of calculating the distribution ratio based on the distribution density function is to quantify the proportion of high constraint areas in the path segment. The specific steps are as follows: first, statistically analyze the distribution density function of the path segment to determine the maximum and average values of the density value, or distinguish high and low constraint segments by setting a percentile threshold (such as an interval higher than 80% of the maximum density is considered a high constraint area); then, the ratio of the total length of the high constraint interval in the path segment to the length of the entire path segment is calculated, i.e. the distribution ratio of the constraint characteristics of the path segment is obtained. This ratio represents the proportion of space affected by sediment constraints in the entire path segment, providing a quantitative basis for dynamic adjustment of the path. For example, if a path segment is 10 meters long, 4 meters of which belong to a high constraint density segment, the distribution ratio is 40%, which can be used to determine whether the path segment needs to be adjusted or avoided in the high constraint area.

[0090] Finally, according to the distribution ratio of the path segment, the first cable path can be optimized and adjusted to avoid areas with dense constraint characteristics, thereby obtaining a second cable path set. The specific method is as follows: for each path segment, according to the location and proportion of its high constraint area distribution, the path points are fine-tuned or relocated in three-dimensional space, so that the path segment is routed around or slightly offset near the high constraint area, while maintaining the continuity and constructability of the overall path. B-spline curves or other smoothing algorithms are used to regenerate the adjusted path segment to ensure smoothness. An example is given: suppose a path segment is originally laid along the center line of the trench, but the dense sediment area is concentrated in the first half of the center line length. Through dynamic adjustment, the path can be offset 0.5 meters to the left side of the trench to bypass the high constraint area, while maintaining the second half of the path along the original route. After adjustment, the curve is regenerated, forming an optimized second cable path segment. After all path segments are adjusted, the complete second cable path set is formed, achieving a balance between obstacle avoidance and construction feasibility for cable laying paths.

[0091] Embodiment 2, Figure 2 The power construction cable laying path optimization system based on image processing of the present application is given, which includes a data acquisition module, a connectivity analysis module, a height analysis module, a constraint analysis module, a path division module, and a path optimization module:

[0092] The data acquisition module is used to acquire multi-angle images of the construction trench, and extract the spatial distribution boundary of the sediment based on the multi-angle images to obtain a sediment feature set;

[0093] The connectivity analysis module is used to analyze the connectivity of the sediment feature set to obtain a continuous sediment feature set;

[0094] The height analysis module is used to divide the spatial range of the continuous sediment feature set and analyze the height gradient of the sediment in the divided spatial range to obtain a sediment height variation feature set;

[0095] The constraint analysis module is configured to perform constraint analysis on the set of continuous characteristics of the deposit and the set of characteristics of the height variation of the deposit, to obtain a constraint matrix of the deposit and extract constraint characteristics in the constraint matrix of the deposit;

[0096] The path division module is configured to divide the preset cable path based on the constraint matrix of the deposit, to obtain a first set of cable paths.

[0097] The path optimization module is configured to calculate a distribution proportion of the constraint characteristics in the first set of cable paths and adjust the first set of cable paths based on the distribution proportion, to obtain a second set of cable paths.

[0098] The above embodiments can be realized wholly or partially by software, hardware, firmware or any other combination. When realized by software, the above embodiments can be realized wholly or partially in the form of a computer program product.

[0099] Those skilled in the art can appreciate that the modules and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized by hardware or software depends on the specific application and design constraints of the technical solutions. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0100] In addition, the functional modules in each of the embodiments of the present application can be integrated in one processing module, or each module can exist physically independently, or two or more modules can be integrated in one module.

[0101] The above is merely specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any modification or replacement within the technical scope disclosed in the present application can be easily thought of by those skilled in the art, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0102] Finally, the above is merely preferred embodiments of the present application, and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for optimizing the laying path of power cables in construction based on image processing, characterized in that, Includes the following steps: Acquire multi-angle images of the construction trench and extract the spatial distribution boundary of sediments based on the multi-angle images to obtain a sediment feature set; Connectivity analysis was performed on the sediment feature set to obtain a continuous sediment feature set. The spatial range of a continuous sediment feature set is defined, and the height gradient of sediments is analyzed within the defined spatial range to obtain a sediment height variation feature set; Constraint analysis was performed on the continuous feature set and the sediment height variation feature set of sediments to obtain the sediment constraint matrix and extract the constraint features from the sediment constraint matrix, specifically: By aligning the continuous feature set of sediments and the feature set of sediment height variation, the first feature of each spatial location is obtained; Based on the first feature of each spatial location, feature differences between each spatial location are evaluated; The sediment constraint matrix is ​​constructed based on the feature difference assessment results. Specifically, the spatial location is used as the row and column of the sediment constraint matrix, and each element stores the feature difference value of the corresponding location pair or the constraint strength calculated based on the difference; singular value decomposition is performed on the sediment constraint matrix to extract the singular vector corresponding to the maximum singular value; the spectral radius and condition number of the sediment constraint matrix are calculated based on the maximum singular value. Structural analysis of the sediment constraint matrix is ​​performed based on the spectral radius and condition number of the sediment constraint matrix to obtain constrained clusters. Features of the constrained clusters are extracted based on singular vectors to obtain the constraint features in the sediment constraint matrix. The specific method for extracting the features of the clusters is as follows: the magnitude of the singular vector components is analyzed, the sediment regions corresponding to the spatial locations with larger component values ​​are marked as the main constraint regions, and the geometric features and height variation features of these regions are summarized to form a feature description of the constrained clusters. The preset cable paths are divided based on the sediment constraint matrix to obtain the first set of cable paths, specifically: Obtain the coordinate sequence of the preset cable path, and discretize the preset cable path into several path points based on the coordinate sequence; cluster the several path points into several path segments according to the value of the sediment constraint matrix; use B-spline curves to smoothly reconstruct the several path segments to obtain the first cable path set; The distribution ratio of the constraint features in the first cable path set is calculated, and the first cable path set is adjusted based on the distribution ratio to obtain the second cable path set, specifically: For each path segment in the first cable path set, the distribution density function of the constraint features along each path length is calculated. Specifically, the path segment is divided into continuous small intervals or equidistant sampling points. The number or intensity of the constraint features existing in each interval is counted, and then divided by the length of the interval to obtain the density value of each interval. The density values ​​of all intervals are arranged along the path segment to form the distribution density function. The proportion threshold of the constraint features is determined based on the distribution density function to obtain the distribution ratio. The position and direction of the path segments in the first cable path set are dynamically adjusted according to the distribution ratio to obtain the second cable path set.

2. The power cable laying path optimization method based on image processing according to claim 1, characterized in that, The process of acquiring multi-angle images of the construction trench and extracting the spatial distribution boundaries of sediments based on these images to obtain a sediment feature set specifically involves: Multi-angle images of the construction trench are collected by multi-view camera equipment deployed above the construction trench, and the multi-angle images are processed for distortion correction and illumination equalization to obtain a multi-angle image set. Depth information is recovered from multi-angle image sets, and a three-dimensional point cloud model of the construction trench is constructed. Surface data of sediments were extracted based on a 3D point cloud model of the construction trench. Morphological segmentation algorithms were used to perform region growing on the surface data of sediments to obtain the spatial distribution boundaries of the sediments. Geometric features of sediments are extracted based on the spatial distribution boundaries of sediments to obtain a sediment feature set.

3. The power cable laying path optimization method based on image processing according to claim 2, characterized in that, The connectivity analysis of the sediment feature set to obtain a continuous sediment feature set is specifically as follows: The sediment feature set is discretized into a uniform grid, and a binary matrix of sediment distribution is constructed. The region growing algorithm is used to traverse the binary matrix of sediment distribution, identify connected sediment regions, and obtain several connected regions. Geometric features of each connected region are extracted to obtain a continuous feature set of sediments.

4. The power cable laying path optimization method based on image processing according to claim 3, characterized in that, The spatial range for dividing the continuous feature set of sediments is specifically as follows: The continuous feature set of sediments is spatially gridded to obtain several grid cells; The average height and height variance of sediments within each grid cell are calculated, and regions are merged based on the height variance to obtain several sub-regions; Calculate the spatial extent boundary of each sub-region to obtain the set of divided spatial extents.

5. The power cable laying path optimization method based on image processing according to claim 4, characterized in that, The analysis of sediment height gradients within the defined spatial range yields a set of sediment height variation characteristics, specifically: Within each spatial division of the spatial range set, sediment height data is obtained for each spatial division. For the sediment height data within each defined spatial range, a surface is fitted to obtain the height gradient vector, and the magnitude of the height gradient vector is calculated. A feature set of sediment height variation is constructed based on the magnitude of the height gradient vector.

6. A power cable laying path optimization system based on image processing, applied to the power cable laying path optimization method based on image processing according to any one of claims 1-5, characterized in that, It includes a data acquisition module, a connectivity analysis module, a height analysis module, a constraint analysis module, a path partitioning module, and a path optimization module. The data acquisition module is used to acquire multi-angle images of the construction trench and extract the spatial distribution boundary of the sediments based on the multi-angle images to obtain the sediment feature set; The connectivity analysis module is used to perform connectivity analysis on sediment feature sets to obtain continuous sediment feature sets. The height analysis module is used to divide the spatial range of a continuous set of sediment features and analyze the height gradient of sediments within the divided spatial range to obtain a set of sediment height variation features. The constraint analysis module is used to perform constraint analysis on the continuous feature set and the feature set of sediment height variation, to obtain the sediment constraint matrix and extract the constraint features in the sediment constraint matrix; The path partitioning module is used to partition the preset cable paths based on the sediment constraint matrix to obtain the first cable path set; The path optimization module is used to calculate the distribution ratio of the constraint features in the first cable path set and adjust the first cable path set based on the distribution ratio to obtain the second cable path set.

Images