Method for calculating number of passable lanes of beam bridge after ammunition strike based on image features

Abstract

The application discloses a kind of based on image feature's ammunition strike after beam bridge passable lane quantity calculation method: step 1, bridge deck and crater information extraction;Step 2, the center distance of any two craters is calculated and distance square matrix is formed;Step 3, if distance square matrix is not formed, jump to step 5;Otherwise, the number that two craters exist no passable path is obtained according to distance square matrix, and matrix is formed, two parameters in each row of matrix are respectively the row number and column number of required element;Step 4, read matrix according to left element consistency row, corresponding obtain each crater position matrix and diameter matrix;Step 5, obtain rendered image;Step 6, from the bottom of image, search bar is operated with rendered image;After searching, matrix is obtained;Step 7, the minimum value in matrix is calculated as passable lane quantity.The method of the application can quickly and accurately calculate the specific quantity of passable lane of large bridge after strike.

Description

Technical Field

[0001] This invention belongs to the field of damage assessment and mainly relates to an image processing technology, specifically a method for calculating the number of passable lanes on a beam bridge after munition impact based on image features. Background Technology

[0002] As a crucial node in transportation networks, beam bridges hold vital strategic significance in modern military operations. Modern large beam bridges are robustly designed with high structural redundancy, possessing strong resistance to damage. In military conflicts, a few conventional munitions are insufficient to completely destroy them; more commonly, partial damage, resulting in localized craters on the bridge surface, is caused. Therefore, quickly and accurately assessing the bridge's remaining capacity becomes a critical element in operational planning.

[0003] Currently, using images for bridge damage assessment has become a major research direction. For example, Zhang et al. used the Chishi Bridge in Chenzhou City, Hunan Province, China as a research object and designed a fully automatic detection system based on the classic target detection algorithm YOLOv7 to realize bridge damage detection. In her dissertation, Cao Haimei extracted the number of craters and the ratio of damaged area on the bridge target after demolition based on images and set different thresholds to calculate the damage level of the bridge target. Yu Tianchao proposed a bridge truncation feature based on the remote sensing image change detection method and judged the bridge damage level based on the conclusion of whether it can be opened to traffic. Liu Jiaqi used the changes in bridge deck satellite images to search for the narrowest traffic path of the demolished bridge by scanning binary images with a circular template to determine the bridge's traffic conditions.

[0004] For bridge targets, researchers have proposed a variety of image-based damage feature extraction methods, which are mostly focused on improving the identification accuracy of the main structure of the bridge or damaged parts (such as craters). However, they lack methods for accurately calculating the final decision problem of bridge functional changes and quantifying trafficability, and also suffer from low computational efficiency. Summary of the Invention

[0005] The purpose of this invention is to provide a method for calculating the number of passable lanes on a beam bridge after a munition strike based on image features, so as to solve the problem that existing methods lack a fast and accurate way to calculate the passability of bridges.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0007] A method for calculating the number of passable lanes on a beam bridge after munition impact based on image features includes the following steps: Step 1: Extract information on bridge surface and craters.

[0008] Based on images of the bridge before and after the impact (see...) Figure 1In (a) and (b)), crater extraction is performed using change detection (see [reference]). Figure 1 (c) in the middle), and then extract the bridge surface target through corner point localization (see (c) in the middle). Figure 1 (d)); for the extracted bridge deck target (see (d)); Figure 1 (The gray part in (d)) uses its two-dimensional image matrix as a coordinate system, with each pixel corresponding to a coordinate; in this coordinate system, a circular detection algorithm (see...) is used... Figure 1 (e) , to achieve crater identification, and obtain the crater identification result image (see Figure 1 (f) and generate the crater location matrix in this two-dimensional image matrix. and crater diameter matrix (See) Figure 1 (f)), where the order of the crater location matrix samples is according to Sort the coordinates from smallest to largest. for coordinate, for coordinate, The number of holes For the first The diameter of the crater, and the vehicle width is set to... W .

[0009] Step 2: Calculate the center distance between any two craters. D jk ( j,k=1,2,…,M ), and form a distance matrix The elements of the square matrix can be calculated using the following formula: .

[0010] Step 3, if a distance matrix is ​​not formed D If there is only one crater, proceed to step 5; if a distance matrix is ​​formed... D Then calculate the distance matrix. D All smaller than in the lower triangular part (excluding the diagonal) The rows and columns of the elements, i.e., the numbers of the two craters where there is no vehicular path, are used to form a matrix. ,matrix Each row in the algorithm has two parameters: the row number and column number of the elements that meet the requirements. The left column contains elements with smaller numbers, and the right column contains elements with larger numbers, thus sorting different rows in ascending order based on the size of the elements on the left.

[0011] in, n Distance matrix D The lower triangular part is smaller than The number of elements; , The first n The row and column numbers of each element that meets the requirements; Step 4: Read the matrix according to the consistent row of the left-hand elements. The matrix of each crater location is obtained accordingly. and diameter matrix Where K and N represent the read crater numbers, calculate the minimum bounding rectangle position matrix containing all craters. Rectangular length and width matrix This leads to the location of the new crater. And the length and width of the rectangle.

[0012] Delete the read data from the original crater location and diameter matrix, convert the remaining crater diameter matrix into the length and width information of the minimum bounding rectangle, and add the new crater location and rectangle length and width at the deleted locations to form a new crater location matrix. and length-width matrix .

[0013] Step 5, take the crater identification result image obtained in Step 1 (see...) Figure 1 The middle (f) area is rendered entirely in white, meaning the pixel value is 1. The new crater location matrix is ​​then read sequentially. and length-width matrix Q Based on the read position information, the corresponding rectangular areas are rendered as black (i.e., pixel values ​​of 0) to obtain the rendered image. See [link / reference]. Figure 5 .

[0014] Step 6: Read the new crater location matrix sequentially. and length-width matrix Q Set search bar J Width is The search bar is the width of the bridge surface and is rendered in white. Starting from the bottom of the rendered image obtained in step 5, an AND operation is performed between the search bar and the rendered image. The calculation formula is as follows:

[0015] Delete white areas whose width is less than the search bar width, and calculate the length of each white area after the deletion operation by dividing it by the vehicle width. Sum of the floor values After the retrieval is completed, a matrix is ​​obtained. .

[0016] Step 7, Calculate the matrix minimum value This is used as the number of passable lanes.

[0017] Compared with the prior art, the beneficial effects of the present invention are reflected in the following aspects: 1. This invention analyzes the features of bridge images after an impact, spatially correlates and logically judges the damaged area with the passable area, and can quickly calculate the specific number of passable lanes on a large bridge after an impact using end-to-end image analysis.

[0018] 2. This invention takes into account the impact of different vehicles and crater sizes, improving its applicability in real-world scenarios. By calculating the specific number of passable lanes, it improves the accuracy of the assessment, providing an effective method for assessing bridge damage effects and offering information support for subsequent operational planning. Attached Figure Description

[0019] Figure 1 Bridge deck and crater information extraction for an example; Figure 2 For example, the calculation of the crater matrix; Figure 3 Matrix for an example calculate; Figure 4 Cavity merging as an example; Figure 5 The bridge surface and craters are rendered for the example. Figure 6 Visualization of passable lanes in an embodiment; Figure 7 The result shows the number of passable lanes in the example. Detailed Implementation

[0020] The present invention will now be described in further detail with reference to the accompanying drawings and preferred embodiments.

[0021] The method in this embodiment includes the following steps: Step 1: Extract information on bridge surface and craters.

[0022] Using MATLAB R2014a, random functions were set up to simulate impact craters of random location and size. Experiments were conducted using images of the simulated damage, with the bridge measuring 30 m in length and 12 m in width, and the vehicle width set to 2.5 m (see [link to MATLAB R2014a]). Figure 1 (a) Calculations show that the original bridge had 4 lanes. Penetration attacks were simulated using randomly placed black circular areas of varying numbers, sizes, and positions (see [reference]). Figure 1 (b)). Crater extraction using change detection (see [reference]). Figure 1 (c) in the middle), and then extract the bridge surface target through corner point localization (see (c) in the middle). Figure 1 (d)). For the bridge surface target extracted from the image after change detection (see [reference]). Figure 1(The gray part in (d)) uses its two-dimensional image matrix as a coordinate system, with each pixel corresponding to a coordinate; in this coordinate system, a circular detection algorithm (see...) is used... Figure 1 (e) , to achieve crater identification, and obtain the crater identification result image (see Figure 1 In the middle (f), the crater location matrix and crater diameter matrix are generated in this two-dimensional image matrix.

[0023] Step 2: Calculate the center distance between any two craters. D jk ( j,k=1,2,…,M ), forming a distance matrix D The elements of the square matrix can be calculated using the following formula:

[0024] Obviously, the distance to the square array D The matrix is ​​symmetric, where the element in the i-th row and j-th column represents the distance between the i-th crater and the j-th crater. Therefore, the matrix can be calculated. D (See) Figure 2 ).

[0025] Step 3, due to the existence of a square matrix D Therefore, there are multiple craters on the bridge surface. Calculation D All values ​​smaller than the lower triangular portion (excluding the diagonal) of the square matrix The rows and columns of the elements, i.e., the numbers of the two craters where there is no vehicular path, are used to form a matrix. (See) Figure 3 The calculation result of the adjacent crater matrix is ​​as follows: each row is the row number and column number of the elements that meet the requirements, the left element is the value with the smaller number, and the right column element is the value with the larger number. Different rows are sorted in ascending order according to the size of the left element.

[0026] matrix This indicates that the intact width between the second and sixth craters is less than 2.5 meters, meaning there is no path for traffic. Similarly, there is no path for traffic between the third and fourth craters.

[0027] Step 4: Read the matrix according to the consistent row of the left-hand elements. This yields the position matrix of each crater (i.e., the second and sixth craters, the third and fourth craters). and diameter matrix Calculate the minimum bounding rectangle position matrix containing all craters. Rectangular length and width matrix This leads to the location of the new crater. and length and width information (see Figure 4 (The merged crater matrix).

[0028] Delete the read data from the original crater location and diameter matrix, convert the remaining crater diameter matrix into the length and width information of the minimum bounding rectangle, and add the new crater location and rectangle length and width data at the deleted locations to form a new crater location matrix. and length-width matrix .

[0029] Step 5, take the crater identification result image obtained in Step 1 (see...) Figure 1 The middle (f) area is rendered entirely in white, meaning the pixel value is 1. The new crater location matrix is ​​then read sequentially. and length-width matrix Q Based on the read position information, the corresponding rectangular area is rendered as black, i.e., the pixel value is 0, to obtain the rendered image (see...). Figure 5 ).

[0030] Step 6: Read the new crater location matrix sequentially. and length-width matrix Q Set search bar J Width is The search bar is the same length as the bridge width and is rendered in white. Starting from the bottom of the rendered image obtained in step 5, the search bar is aligned with the rendered image (see...). Figure 5 The calculation formula is as follows: (The text then goes on to describe the operation and calculation process.)

[0031]

[0032] Delete white areas whose width is less than the search bar width, and calculate the length of each white area after the deletion operation by dividing it by the vehicle width. Sum of the floor values After the retrieval is completed, a matrix is ​​obtained. .

[0033] Step 7, Calculate the matrix minimum value As the number of passable lanes (see Figure 7 ).

[0034] For the entire bridge, the number of passable lanes is the minimum number of passable lanes at each location (the final output). )0, meaning that the number of passable lanes remaining after the bridge was hit was 0.

[0035] Based on the actual execution process and results of the above specific examples, the method of the present invention can quickly and accurately calculate the number of passable lanes end-to-end based on the bridge's front and rear images. Compared with the prior art, the calculation method is simpler and the evaluation accuracy is better.

Claims

1. A method for calculating the number of passable lanes on a beam bridge after munition impact based on image features, comprising the following steps: Step 1, Extraction of bridge surface and crater information: Based on images of the bridge before and after the impact, craters are extracted using change detection, and then the bridge surface targets are extracted using corner point localization. For the extracted bridge surface targets, a two-dimensional image matrix is ​​used as a coordinate system, with each pixel corresponding to a coordinate. In this coordinate system, a circular detection algorithm is used to identify the craters, resulting in an image of the identified craters. A crater location matrix is ​​then generated within this two-dimensional image matrix. and crater diameter matrix The order of the crater location matrix samples is as follows: Sort the coordinates from smallest to largest. for coordinate, for coordinate, The number of holes For the first The diameter of the crater, and the vehicle width is set to... W ; Step 2: Calculate the center distance between any two craters. D jk ,in j,k=1,2,…,M and form a distance matrix Step 3, if a distance matrix is ​​not formed D If there is only one crater, proceed to step 5; if a distance matrix is ​​formed... D Then calculate the distance matrix. D All smaller than in the lower triangular part The rows and columns of the elements, i.e., the numbers of the two craters where there is no vehicular path, are used to form a matrix. ,matrix Each row in the algorithm has two parameters: the row number and column number of the elements that meet the requirements. The left column contains elements with smaller numbers, and the right column contains elements with larger numbers, thus sorting different rows in ascending order based on the size of the elements on the left. in, n Distance matrix D The lower triangular part is smaller than The number of elements; , The first n The row and column numbers of each element that meets the requirements; Step 4: Read the matrix according to the consistent row of the left-hand elements. The matrix of each crater location is obtained accordingly. and diameter matrix Where K and N represent the read crater numbers, calculate the minimum bounding rectangle position matrix containing all craters. Rectangular length and width matrix This leads to the location of the new crater. and the length and width of the rectangle; Delete the read data from the original crater location and diameter matrix, convert the remaining crater diameter matrix into the length and width information of the minimum bounding rectangle, and add the new crater location and rectangle length and width at the deleted locations to form a new crater location matrix. and length-width matrix ; Step 5: Render the entire crater identification result image obtained in Step 1 as white, i.e., with a pixel value of 1, and sequentially read the new crater location matrix. and length-width matrix Q Based on the read position information, the corresponding rectangular areas are rendered as black, i.e., the pixel value is 0, to obtain the rendered image; Step 6: Read the new crater location matrix sequentially. and length-width matrix Q Set search bar J Width is The length of the search bar is the width of the bridge surface, and the search bar is rendered in white. Starting from the bottom of the rendered image obtained in step 5, the search bar and the rendered image are ANDed. Delete white areas whose width is less than the search bar width, and calculate the length of each white area after the deletion operation by dividing it by the vehicle width. Sum of the floor values After the retrieval is completed, a matrix is ​​obtained. ; Step 7, Calculate the matrix minimum value This is used as the number of passable lanes.

2. The method for calculating the number of passable lanes on a beam bridge after munition impact based on image features as described in claim 1, characterized in that, In step 1, the elements of the distance matrix D are calculated using the following formula: 。 3. The method for calculating the number of passable lanes on a beam bridge after munition impact based on image features as described in claim 1, characterized in that, In step 6, the calculation formula is as follows: 。 4. An electronic device, characterized in that, The electronic device includes: Memory, used to store executable instructions; A processor, when executing executable instructions or computer programs stored in the memory, implements the method as described in any one of claims 1 to 3.

5. A computer-readable storage medium storing executable instructions or a computer program, characterized in that, When the executable instructions are executed by the processor, they implement the method as described in any one of claims 1 to 3.

Images