Material de-weighting statistical method in distribution network project acceptance

Abstract

The invention provides a material de-weight statistical method in distribution network project acceptance, and aims to provide a de-weight statistical method in combination with a material volume, and design an algorithm for de-weight through the spatial position and the material volume of a material, and the algorithm can carry out de-weight calculation according to the spatial position and the material volume of the material from a plurality of pictures shot from different angles. The positions of the same material are accurately matched, repeated calculation is avoided, and efficient and accurate material statistics and acceptance are achieved. The second purpose of the invention is to provide a multi-view image fusion algorithm. According to the algorithm, tower pictures shot from different angles can be fused, and a more complete and consistent view is constructed through feature matching and image splicing technologies. The problem of shielding possibly existing under a single view angle is solved, the specific position of each material can be recognized more accurately, and therefore repeated counting in statistics is avoided.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method for deduplicating and counting materials during the acceptance of power distribution network projects. Background Technology

[0002] With the development of smart grids, the maintenance and management of power systems increasingly rely on automation and information technology. During the construction of distribution networks, the installation quality of poles and related components directly affects the safety and reliability of power transmission. Traditional distribution network project acceptance methods typically rely on manual on-site inspections, which are not only inefficient but also prone to errors due to human factors. In recent years, with the development of image recognition and spatial positioning technologies, automated acceptance of distribution network equipment using computer vision has become possible. This not only improves work efficiency but also reduces human error and ensures project quality.

[0003] However, existing image recognition-based power distribution network acceptance technologies still have some limitations. On the one hand, in complex environments, such as changes in lighting, shading, and different weather conditions, image recognition algorithms may make misjudgments or omissions. On the other hand, due to the large number and dense layout of components on power distribution network towers, it is difficult to accurately distinguish components of the same type but located in different places when using a single image for identification, leading to duplication or omissions in material statistics. In addition, existing systems often lack effective mechanisms to integrate image information taken from multiple angles, making it impossible to accurately determine the spatial location of materials and affecting the accuracy of the final statistical results.

[0004] Therefore, there is an urgent need for a method that can efficiently and accurately count the deduplication of materials on power distribution towers. Summary of the Invention

[0005] This invention provides a method for deduplication and statistical analysis of materials during the acceptance of power distribution network projects. The method designs an algorithm for deduplication based on the spatial location and volume of materials. This algorithm can accurately match the location of the same material from multiple images taken from different angles, avoiding duplicate calculations and achieving efficient and accurate material statistics and acceptance.

[0006] To address the aforementioned technical problems, this invention provides a method for deduplicating and counting materials during the acceptance of power distribution network projects, comprising the following steps:

[0007] A1. For each type of material to be accepted, a three-dimensional model of it should be established in advance or its standard volume Vs should be estimated. This information can be obtained through actual measurement or by referring to the material specifications.

[0008] A2. Material spatial location calculation: Using the multi-view image fusion positioning algorithm described above, the position coordinates (X, Y, Z) of each material in three-dimensional space are obtained;

[0009] A3. For all located materials, calculate the Euclidean distance between their spatial positions; let the positions of two materials be (X1, Y1, Z1) and (X2, Y2, Z2), then the distance d between them can be calculated using the following formula:

[0010]

[0011] A4. Compare the calculated distance d with the standard material volume Vs; when the distance d is less than a certain threshold dthreshold and the material volume difference is within the allowable range, determine that these materials are the same material and perform deduplication; the specific judgment conditions are as follows:

[0012] d <d threshold And |V observed -Vs| <V tolerance

[0013] Where Vobserved is the material volume inferred from the image, and Vtolerance is the allowable volume error range;

[0014] A5. After deduplication, count the final quantity of materials and generate an acceptance report to ensure that each material is accurately identified and not counted repeatedly from different perspectives.

[0015] Preferably, the material deduplication and counting method is based on an application method of a multi-view image fusion algorithm, which includes the following steps:

[0016] B1. Take multiple images of the power distribution towers from different angles to ensure that all materials to be inspected can be photographed from at least two different angles; generally, the shooting positions are directly above the top of the tower, and five images in four directions: front, back, left, and right of the tower; the image directly above the top of the tower can also provide a basis for calculating the tower's position;

[0017] B2. Use the key material identification model of the power distribution network to identify materials in each image and mark the pixel coordinates of the materials in the image;

[0018] B3. Use the Scale Invariant Feature Transform (SIFT) algorithm to match feature points between images from different viewpoints to find the corresponding positions of the same material in multiple images; for images i and j, calculate the corresponding points (xi,yi) and (xj,yj) of the material in the image using the SIFT algorithm;

[0019] B4. Perform camera calibration to obtain the camera's intrinsic parameter matrix K and extrinsic parameter matrix (rotation matrix R and translation vector t), converting pixel coordinates to coordinates in the camera coordinate system; then, use the essential matrix E or fundamental matrix F in multi-view geometry to calculate the view transformation matrix, as shown in the following formula:

[0020] Calculate the essential matrix E:

[0021] E=K T ×F×K

[0022] Where K is the camera intrinsic parameter matrix;

[0023] The triangulation method is used to calculate the three-dimensional position (X, Y, Z) of the material in space; using corresponding points of the material in two viewpoint images, the following system of linear equations is solved:

[0024] (x i ×P1)=(X,Y,Z,1)×P1

[0025] (x j ×P2)=(X,Y,Z,1)×P2

[0026] P1 and P2 are the projection matrices calculated from images i and j, respectively; the three-dimensional coordinates (X, Y, Z) of the material are obtained by solving the system of equations using the least squares method.

[0027] Compared with related technologies, the present invention has the following beneficial effects:

[0028] (1) The multi-view image fusion algorithm provided by this invention. By combining multi-angle shooting and image recognition technology, it can not only effectively solve the problems of missed detection and misjudgment that are prone to occur in the traditional manual inspection process, but also use three-dimensional spatial positioning to ensure accurate identification and deduplication of materials. This method shows higher accuracy and reliability in complex tower environments, can automatically locate and deduplicate materials, and significantly reduce the time and cost of manual acceptance.

[0029] (2) The deduplication method combining material volume provided by this invention further improves the accuracy and comprehensiveness of the acceptance process by introducing material volume information. The algorithm combining spatial location and material volume can more intelligently filter duplicate counts, adapting to acceptance needs in various complex environments. This not only improves the efficiency of material statistics but also significantly enhances the accuracy and consistency of acceptance results, making the acceptance process of power distribution network projects more efficient and intelligent, and helping to accelerate project delivery and completion. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the power distribution tower image acquisition method of the present invention. The drone takes pictures of the tower head from five directions: directly in front, directly behind, directly to the left, directly to the right, and directly above. Detailed Implementation

[0031] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0032] The first objective of this invention is to propose a deduplication method that combines material volume. An algorithm is designed to deduplicatize materials based on their spatial location and volume. This algorithm can accurately match the location of the same material from images taken from multiple different angles, avoiding duplicate calculations and achieving efficient and accurate material statistics and acceptance.

[0033] The technical solution to achieve the second objective of this invention is:

[0034] 1. For each type of material to be accepted, a three-dimensional model of it should be created in advance or its standard volume Vs should be estimated. This information can be obtained through actual measurement or by referring to the material specifications.

[0035] 2. Material spatial location calculation: Using the multi-view image fusion positioning algorithm described above, the position coordinates (X, Y, Z) of each material in three-dimensional space are obtained.

[0036] 3. For all located materials, calculate the Euclidean distance between their spatial positions. Let the positions of two materials be (X1, Y1, Z1) and (X2, Y2, Z2), then the distance d between them can be calculated using the following formula:

[0037]

[0038] 4. Compare the calculated distance d with the standard material volume Vs. When the distance d is less than a certain threshold d... threshold If the volume difference of the materials is within the allowable range, these materials are considered to be the same material, and deduplication is performed. The specific judgment criteria are as follows:

[0039] d <d thredhold And |V observed -Vs| <V tolerance

[0040] Among them, V observed The volume of material inferred from the image, V tolerance This is the allowable volume error range.

[0041] 5. After deduplication, count the final quantity of materials and generate an acceptance report to ensure that each material is accurately identified and not counted repeatedly from different perspectives.

[0042] The second objective of this invention is to provide a multi-view image fusion algorithm. This algorithm can fuse images of towers taken from different angles, constructing a more complete and consistent view through feature matching and image stitching techniques. It overcomes the occlusion problem that may exist in a single viewpoint and can more accurately identify the specific location of each material, thereby avoiding duplicate counting in statistics.

[0043] The technical solution to achieve the first objective of this invention is:

[0044] 1. Take multiple images of the power distribution towers from different angles to ensure that all materials to be inspected are captured from at least two different angles. Generally, five images are taken: one directly above the top of the tower, and four in the front, back, left, and right directions. The image directly above the top of the tower also provides a basis for calculating the tower's position.

[0045] 2. Use the key material identification model of the power distribution network to identify materials in each image and mark the pixel coordinates of the materials in the image.

[0046] 3. The Scale Invariant Feature Transform (SIFT) algorithm is used to match feature points between images from different viewpoints to find the corresponding positions of the same material in multiple images. For images i and j, the corresponding points (xi, yi) and (xj, yj) of the material in the image are calculated using the SIFT algorithm.

[0047] 4. Perform camera calibration to obtain the camera's intrinsic parameter matrix K and extrinsic parameter matrix (rotation matrix R and translation vector t), converting pixel coordinates to coordinates in the camera coordinate system. Then, use the essential matrix E or fundamental matrix F from multi-view geometry to calculate the view transformation matrix, as shown in the following formula:

[0048] Calculate the essential matrix E:

[0049] E=K T ×F×K

[0050] Where K is the camera intrinsic parameter matrix.

[0051] The triangulation method is used to calculate the three-dimensional position (X, Y, Z) of the material in space. Using corresponding points of the material in two viewpoint images, the following system of linear equations is solved:

[0052] (x i ×P1)=(X,Y,Z,1)×P1

[0053] (x j ×P2)=(X,Y,Z,1)×P2

[0054] P1 and P2 are the projection matrices calculated from images i and j, respectively. The three-dimensional coordinates (X, Y, Z) of the material are obtained by solving the system of equations using the least squares method.

[0055] The above are merely embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for deduplication and statistical analysis of materials during the acceptance of power distribution network projects, characterized in that, Includes the following steps: A1. For each type of material to be accepted, a three-dimensional model of it should be established in advance or its standard volume Vs should be estimated. This information can be obtained through actual measurement or by referring to the material specifications. A2. Material spatial location calculation: Using the multi-view image fusion positioning algorithm described above, the position coordinates (X, Y, Z) of each material in three-dimensional space are obtained; A3. For all located materials, calculate the Euclidean distance between their spatial positions; let the positions of two materials be (X1, Y1, Z1) and (X2, Y2, Z2), then the distance d between them can be calculated using the following formula: A4. Compare the calculated distance d with the standard material volume Vs; when the distance d is less than a certain threshold dthreshold and the material volume difference is within the allowable range, determine that these materials are the same material and perform deduplication; the specific judgment conditions are as follows: d <d threshold And |V observed -V s | <V tolerance Where Vobserved is the material volume inferred from the image, and Vtolerance is the allowable volume error range; A5. After deduplication, count the final quantity of materials and generate an acceptance report to ensure that each material is accurately identified and not counted repeatedly from different perspectives.

2. The method for deduplication and statistical analysis of materials during the acceptance of power distribution network projects according to claim 1, characterized in that: The material deduplication and statistical method is based on an application of a multi-view image fusion algorithm, which includes the following steps: B1. Take multiple images of the power distribution towers from different angles to ensure that all materials to be inspected can be photographed from at least two different angles; generally, the shooting positions are directly above the top of the tower, and five images in four directions: front, back, left, and right of the tower; the image directly above the top of the tower can also provide a basis for calculating the tower's position; B2. Use the key material identification model of the power distribution network to identify materials in each image and mark the pixel coordinates of the materials in the image; B3. Use the Scale Invariant Feature Transform (SIFT) algorithm to match feature points between images from different viewpoints to find the corresponding positions of the same material in multiple images; for images i and j, calculate the corresponding points (xi,yi) and (xj,yj) of the material in the image using the SIFT algorithm; B4. Perform camera calibration to obtain the camera's intrinsic parameter matrix K and extrinsic parameter matrix (rotation matrix R and translation vector t), converting pixel coordinates to coordinates in the camera coordinate system; then, use the essential matrix E or fundamental matrix F in multi-view geometry to calculate the view transformation matrix, as shown in the following formula: Calculate the essential matrix E: E=K T ×F×K Where K is the camera intrinsic parameter matrix; The triangulation method is used to calculate the three-dimensional position (X, Y, Z) of the material in space; using corresponding points of the material in two viewpoint images, the following system of linear equations is solved: (x i ×P1)=(X,Y,Z,1)×P1 (x j ×P2)=(X,Y,Z,1)×P2 P1 and P2 are the projection matrices calculated from images i and j, respectively; the three-dimensional coordinates (X, Y, Z) of the material are obtained by solving the system of equations using the least squares method.

Images