Concrete roughness detection method based on improved bitterfish optimization algorithm

Abstract

The application discloses a concrete roughening roughness detection method based on an improved bitterfish optimization algorithm, S1. Obtain the original image data covering the detection area; S2. Perform multi-scale grid segmentation on the original image data; S3. Perform image preprocessing on the plurality of grid segmentation image data respectively; S4. Form a concrete surface roughness feature vector; S5. Obtain the optimal roughness parameter in each grid area; S6. Perform global fusion processing on the optimal roughness parameter in each grid area, integrate the multi-scale detection results by using a unified fitting model, and generate the overall concrete roughening roughness detection result; S7. Generate a detection report based on the overall concrete roughening roughness detection result. The application can finely adjust the roughness parameter, avoid the local optimal trap, and more stably converge to the optimal solution when processing the complex non-homogeneous concrete roughening surface.

Description

Technical Field

[0001] This invention relates to the field of concrete roughness detection technology, and in particular to a concrete roughness detection method based on an improved bitterfish optimization algorithm. Background Technology

[0002] With the development of building construction and infrastructure construction, concrete roughening technology, as an important means to improve the bonding strength between new and old concrete, has been widely used in the construction of bridges, tunnels, subways and large building structures. The main purpose of concrete roughening is to remove the smooth layer on the concrete surface by mechanical or manual means to form a rough texture of a certain depth and uniformity, so as to enhance the interfacial shear strength.

[0003] Currently, the roughness inspection of roughened concrete mainly relies on manual visual inspection or traditional contact measurement methods. Manual visual inspection is usually carried out by experienced inspectors who judge the roughness level based on their vision. However, this method is highly subjective and greatly affected by ambient lighting and the experience of the inspectors, making it difficult to guarantee the consistency and repeatability of the test results. Traditional contact measurement methods, such as the pin gauge method and profilometer measurement method, can provide quantitative indicators, but these methods have many limitations in practical applications, including complex operation, low detection efficiency, and measurement accuracy affected by the contact area and pressure of the probe. In addition, contact measurement methods cannot fully cover the microscopic features of the roughened concrete surface, resulting in local deviations in the measurement data and making it difficult to reflect the overall roughness.

[0004] In recent years, the development of image processing and computational intelligence technologies has provided new ideas for non-contact concrete surface inspection. Some studies have proposed using high-precision image acquisition equipment combined with texture analysis methods for roughness detection. However, existing image analysis methods still have the following shortcomings:

[0005] Insufficient accuracy in image data feature extraction: Traditional image processing methods are easily affected by changes in lighting and surface noise when dealing with complex textures, leading to a decrease in feature extraction accuracy.

[0006] Optimization algorithms are prone to getting trapped in local optima: Existing roughness optimization algorithms are prone to getting trapped in local optima when faced with the heterogeneous and multi-scale texture features of concrete surfaces, leading to a decrease in detection accuracy.

[0007] Lack of integration of multi-scale texture features: The scale span of the texture of roughened concrete surface is large, and existing methods often cannot take into account both local and global features at the same time, resulting in unstable final roughness assessment results.

[0008] In summary, existing technologies for detecting the roughness of rubbed concrete suffer from problems such as high subjectivity, low measurement efficiency, insufficient algorithm optimization capabilities, and inadequate multi-scale feature fusion, making it difficult to meet the needs of high-precision, standardized, and intelligent detection. Summary of the Invention

[0009] One objective of this invention is to propose a concrete roughness detection method based on an improved bitterfish optimization algorithm. This invention can finely adjust the roughness parameters, avoid local optimum traps, and converge to the optimal solution more stably when dealing with complex heterogeneous concrete roughened surfaces.

[0010] A method for detecting the roughness of concrete chiseling based on an improved bitterfish optimization algorithm according to an embodiment of the present invention includes the following steps:

[0011] S1. Use an image acquisition device to acquire images of the roughened concrete surface and obtain the original image data covering the detection area;

[0012] S2. Perform multi-scale grid segmentation on the original image data, dividing the original image data into several grid segmentation image data of different scales according to the preset scale parameters;

[0013] S3. Perform image preprocessing on multiple grid-segmented image data to generate optimized preprocessed grid-segmented image data;

[0014] S4. Extract the texture features of the roughened concrete surface from the optimized preprocessed mesh segmentation image data to form a concrete surface roughness feature vector;

[0015] S5. Input the concrete surface roughness feature vector into the improved bitter fish optimization algorithm, and optimize the roughness features in each grid region according to the dynamic weight adjustment mechanism and adaptive iteration strategy to obtain the optimal roughness parameters in each grid region.

[0016] S6. Perform global fusion processing on the optimal roughness parameters in each grid region, and use a unified fitting model to integrate the multi-scale detection results to generate the overall concrete roughness detection results.

[0017] S7. Generate a test report based on the overall concrete roughness test results, including roughness numerical evaluation, distribution diagram and abnormal area prompts.

[0018] Optionally, S2 includes the following steps:

[0019] S21. Set the grid segmentation scale parameters for the roughened concrete surface. Based on the size range and surface roughness characteristics of the detection area, define different scale levels L = {l1, l2, ..., l i}, where li This represents the grid segmentation scale of the i-th layer, where i is the set number of scale levels;

[0020] S22. Divide the original image data I into grid regions of different scale levels according to the grid segmentation scale parameter. Indicates the grid segmentation scale l i The segmented image data, grid region The size is determined by the following relationship:

[0021]

[0022] Where W and H are the width and height of the original image data I, respectively. and They represent the grid segmentation scale l i The number of horizontal and vertical grid divisions below;

[0023] S23. For each grid segmentation scale l i The grid area below Set grid area overlap rate And obtain the final segmentation result of the grid region.

[0024]

[0025] in, The preset overlap ratio;

[0026] S24. For all grid segmentation scales l i The grid area below Index and number the images to form a multi-scale grid segmentation image dataset.

[0027] Optionally, S3 includes the following steps:

[0028] S31. Noise filtering is performed on the multi-scale grid segmented image data G. The image data of each grid region is processed to filter out random noise caused by changes in illumination, equipment noise and environmental interference, and the processed grid segmented image data is obtained.

[0029] S32. Enhance the contrast of the noise-filtered mesh segmentation image data and optimize the features of different roughness regions on the concrete surface to obtain the adjusted mesh segmentation image data.

[0030] S33. Perform edge detection on the contrast-enhanced grid segmentation image data, extract the key boundaries of the texture features of the roughened concrete surface, and obtain the extracted edge grid segmentation image data.

[0031] S34. The edge-refined image data of the edge grid segmentation is processed to remove redundant edge information, and the effective edge features are enhanced by combining double threshold filtering to finally obtain optimized preprocessed grid segmentation image data.

[0032] S35. Optimize the preprocessing of grid-segmented image data at all grid segmentation scales. Data storage is performed to form an optimized preprocessed image data set:

[0033]

[0034] Optionally, S4 includes the following steps:

[0035] S41. Based on the optimized preprocessed grid segmentation image data set G final For each grid segmentation scale, the image data is segmented into grids. Calculate the gray-level co-occurrence matrix

[0036]

[0037] Where x and y are gray levels, P and Q are the number of rows and columns of the grid segmented image data, respectively, Δp and Δq are relative displacements, and δ is an indicator function used to statistically analyze the joint probability distribution of gray values ​​of adjacent pixels;

[0038] S42. Based on the gray-level co-occurrence matrix Calculate the contrast at each grid segmentation scale entropy Correlation and energy

[0039]

[0040]

[0041]

[0042]

[0043] Where, μ x μ y Let σ be the mean of gray levels x and y, respectively. x , σ y These are the standard deviations;

[0044] S43. Calculate the eigenvectors of local binary patterns.

[0045]

[0046] Among them, (p k ,q k Let ) be the neighboring pixels of pixel (p, q), K be the number of neighboring pixels, and s(x) be the binary mapping function;

[0047] S44. Vectorize the calculation results of contrast, entropy, correlation, energy, and local binary mode at all grid segmentation scales to construct a set of concrete surface roughness feature vectors:

[0048]

[0049] Optionally, S5 includes the following steps:

[0050] S51. Input the set of concrete surface roughness feature vectors V into the improved Bitterfish optimization algorithm, and calculate the roughness contrast coefficient of each grid based on the information contrast for different grid regions of the roughened concrete surface:

[0051]

[0052] Where ∈0 is a constant to prevent the denominator from being zero;

[0053] Roughness comparison coefficient Each individual S in the initial population P0 is generated as a weighting factor. j Initial roughness estimate of (0):

[0054]

[0055] Where β and γ are preset calibration coefficients, rand(0,1) is a random number that follows a uniform distribution, and j = 1,2,…,N represents the number of the candidate solution in the population;

[0056] S52. Set the roughness difference factor Describes the roughness variation between grid regions to reflect the differences in local roughness on the concrete surface:

[0057]

[0058] Where, N adj (l i ) for the lth i The set of adjacent grids in a scaled grid region. For the l i Comprehensive roughness index at the scale:

[0059]

[0060] Wherein, ω1, ω2, ω3, ω4, and ω5 are preset weight coefficients;

[0061] S53. Introduce a dynamic weight adjustment mechanism to adjust the roughness difference factor. Integrating the weight updates of bitter fish and prey, we set the bitter fish influence factor W at the t-th iteration. S (t) and prey influence factor W P (t):

[0062]

[0063] W P (t)=1-W S (t);

[0064] in, and These represent the maximum and minimum values ​​of the influence factor of bitter fish, respectively; T is the maximum number of iterations; and λ is the amplification factor of the roughness difference effect.

[0065] S54. In each iteration, for each candidate solution in the population Adopt improved update rules:

[0066]

[0067] Among them, S best (t) represents the individual bitter fish with the smallest detection error based on concrete surface roughness in the current population, P best (t) represents the current best prey individual, r1 and r2 are uniformly distributed random numbers, and α is the roughness gradient adjustment coefficient;

[0068] S55. Define a fitness function to evaluate candidate solutions. The deviation between the measured value and the actual concrete surface roughness, fitness function for:

[0069]

[0070] in, The first one obtained by experiment or field testing i Actual values ​​of concrete surface roughness under scale;

[0071] S56. When the fitness function When the preset convergence threshold ∈ is met or the maximum number of iterations T is reached, the l-th iteration is determined. i Optimal roughness parameters at scale The optimal candidate solution for the current fitness is given, and the set of optimal roughness parameters for all grid regions is represented as:

[0072]

[0073] Each parameter reflects both the local differences and overall characteristics of the texture of the roughened concrete surface.

[0074] Optionally, S6 includes the following steps:

[0075] S61. Calculate the multi-scale roughness weighting coefficient based on the set of optimal roughness parameters for each grid region. Multi-scale roughness weighting coefficients are used to reflect the contribution of different scale grid regions to the overall roughness:

[0076]

[0077] in, For the l i Comprehensive roughness index at the scale, is the average roughness across all grid scales, and κ is the local roughness variation influence factor, used to amplify or reduce the influence of the grid region on the overall roughness.

[0078] The multi-scale roughness weighting coefficient compares the differences between the comprehensive roughness index of each grid region and the global mean, giving regions with roughness changes greater than a threshold higher weight during the fusion process.

[0079] S62. Based on the multi-scale roughness weighting coefficient W li Optimal roughness parameters for each grid region Weighted fusion is performed by introducing a local deviation correction term during the fusion process, so that the overall roughness assessment value reflects the true roughness distribution of the concrete surface, generating the fused roughness parameter P. fusion :

[0080]

[0081] in, λ1 is the mean of the optimal roughness parameters for all grid regions, and λ1 is the roughness deviation adjustment coefficient, which is used to correct the impact of local anomalies on the overall evaluation during the fusion process.

[0082] S63. Calculate the fusion roughness error E fusion This is used to evaluate the rationale for multi-scale fusion.

[0083]

[0084] The first term measures the deviation between the locally optimal roughness parameter and the fusion parameter, while the second term assesses the degree of deviation in the comprehensive roughness index, ensuring that the fusion result characterizes the roughness features of the roughened concrete surface. If the fusion roughness error meets the preset fusion error threshold ∈ f If the result is positive, the fusion result is valid; otherwise, the weighting coefficient W is adjusted. liAnd recalculate the fusion roughness parameter P fusion ;

[0085] S64. Using a nonlinear fitting model F fit For the fusion roughness parameter P fusion The fitting and optimization process is performed to make the fused roughness parameters conform to the actual surface roughness distribution of concrete through an adaptive optimization process.

[0086]

[0087] Among them, P final The final output is the overall roughness test of the concrete.

[0088] The beneficial effects of this invention are:

[0089] (1) The present invention adopts a multi-scale grid segmentation strategy to divide the concrete roughened surface into multi-level regions, and extracts the texture features of the concrete surface through grid segmentation image data of different scales. It can capture the micro and macro features of the roughness of the concrete surface at different scales, and weighted integrate the optimal roughness parameters of each grid region through an adaptive fusion strategy, so that the overall roughness detection results can reflect the detailed features and maintain global consistency.

[0090] (2) This invention improves the bitter fish optimization algorithm. Based on the traditional bitter fish optimization algorithm, a roughness comparison coefficient is introduced. The comprehensive roughness difference of different grid regions is used as the optimization weight to improve the algorithm's adaptability to complex texture features. In addition, a dynamic weight adjustment mechanism is adopted to dynamically adjust the influence factors of bitter fish and prey according to the roughness gradient during the optimization process. This ensures that the algorithm has a strong global search capability in the early stage and can finely adjust the roughness parameters in the later stage to avoid local optimal traps. When dealing with complex heterogeneous concrete roughened surfaces, it can more stably converge to the optimal solution.

[0091] (3) In the process of multi-scale roughness fusion, the present invention adopts a nonlinear fitting model and uses an adaptive weight adjustment mechanism to perform secondary optimization on the roughness parameters after fusion, so as to reduce the impact of local outliers on the overall detection results. The nonlinear fitting strategy can automatically adjust the parameters according to the changing trend of concrete surface texture, so that the final detection results are more in line with the real roughness distribution. Attached Figure Description

[0092] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0093] Figure 1This is a flowchart of a concrete roughness detection method based on an improved bitter fish optimization algorithm proposed in this invention. Detailed Implementation

[0094] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0095] refer to Figure 1 A method for detecting the roughness of concrete chiseling based on an improved bitter fish optimization algorithm includes the following steps:

[0096] S1. Use an image acquisition device to acquire images of the roughened concrete surface and obtain the original image data covering the detection area;

[0097] S2. Perform multi-scale grid segmentation on the original image data, dividing the original image data into several grid segmentation image data of different scales according to the preset scale parameters;

[0098] S3. Perform image preprocessing on multiple grid-segmented image data to generate optimized preprocessed grid-segmented image data;

[0099] S4. Extract the texture features of the roughened concrete surface from the optimized preprocessed mesh segmentation image data to form a concrete surface roughness feature vector;

[0100] S5. Input the concrete surface roughness feature vector into the improved bitter fish optimization algorithm, and optimize the roughness features in each grid region according to the dynamic weight adjustment mechanism and adaptive iteration strategy to obtain the optimal roughness parameters in each grid region.

[0101] S6. Perform global fusion processing on the optimal roughness parameters in each grid region, and use a unified fitting model to integrate the multi-scale detection results to generate the overall concrete roughness detection results.

[0102] S7. Generate a test report based on the overall concrete roughness test results, including roughness numerical evaluation, distribution diagram and abnormal area prompts.

[0103] In this embodiment, S2 includes the following steps:

[0104] S21. Set the grid segmentation scale parameters for the roughened concrete surface. Based on the size range and surface roughness characteristics of the detection area, define different scale levels L = {l1, l2, ..., l i}, where l i This represents the grid segmentation scale of the i-th layer, where i is the set number of scale levels;

[0105] S22. Divide the original image data I into grid regions of different scale levels according to the grid segmentation scale parameter. Indicates the grid segmentation scale l i The segmented image data, grid region The size is determined by the following relationship:

[0106]

[0107] Where W and H are the width and height of the original image data I, respectively. and They represent the grid segmentation scale l i The number of horizontal and vertical grid divisions below;

[0108] S23. For each grid segmentation scale l i The grid area below Set grid area overlap rate And obtain the final segmentation result of the grid region.

[0109]

[0110] in, The preset overlap ratio;

[0111] S24. For all grid segmentation scales l i The grid area below Index and number the images to form a multi-scale grid segmentation image dataset.

[0112] In this embodiment, S3 includes the following steps:

[0113] S31. Noise filtering is performed on the multi-scale grid segmented image data G. The image data of each grid region is processed to filter out random noise caused by changes in illumination, equipment noise and environmental interference, and the processed grid segmented image data is obtained.

[0114] S32. Enhance the contrast of the noise-filtered mesh segmentation image data and optimize the features of different roughness regions on the concrete surface to obtain the adjusted mesh segmentation image data.

[0115] S33. Perform edge detection on the contrast-enhanced grid segmentation image data, extract the key boundaries of the texture features of the roughened concrete surface, and obtain the extracted edge grid segmentation image data.

[0116] S34. The edge-refined image data of the edge grid segmentation is processed to remove redundant edge information, and the effective edge features are enhanced by combining double threshold filtering to finally obtain optimized preprocessed grid segmentation image data.

[0117] S35. Optimize the preprocessing of grid-segmented image data at all grid segmentation scales. Data storage is performed to form an optimized preprocessed image data set:

[0118]

[0119] In this embodiment, S4 includes the following steps:

[0120] S41. Based on the optimized preprocessed grid segmentation image data set G final For each grid segmentation scale, the image data is segmented into grids. Calculate the gray-level co-occurrence matrix

[0121]

[0122] Where x and y are gray levels, P and Q are the number of rows and columns of the grid segmented image data, respectively, Δp and Δq are relative displacements, and δ is an indicator function used to statistically analyze the joint probability distribution of gray values ​​of adjacent pixels;

[0123] S42. Based on the gray-level co-occurrence matrix Calculate the contrast at each grid segmentation scale entropy Correlation and energy

[0124]

[0125]

[0126]

[0127]

[0128] Where, μ x μ y Let σ be the mean of gray levels x and y, respectively. x , σ y These are the standard deviations;

[0129] S43. Calculate the eigenvectors of local binary patterns.

[0130]

[0131] Among them, (p k ,q kLet ) be the neighboring pixels of pixel (p, q), K be the number of neighboring pixels, and s(x) be the binary mapping function;

[0132] S44. Vectorize the calculation results of contrast, entropy, correlation, energy, and local binary mode at all grid segmentation scales to construct a set of concrete surface roughness feature vectors:

[0133]

[0134] In this embodiment, S5 includes the following steps:

[0135] S51. Input the set of concrete surface roughness feature vectors V into the improved Bitterfish optimization algorithm, and calculate the roughness contrast coefficient of each grid based on the information contrast for different grid regions of the roughened concrete surface:

[0136]

[0137] Where ∈0 is a constant to prevent the denominator from being zero;

[0138] Roughness comparison coefficient Each individual S in the initial population P0 is generated as a weighting factor. j Initial roughness estimate of (0):

[0139]

[0140] Where β and γ are preset calibration coefficients, rand(0,1) is a random number that follows a uniform distribution, and j = 1,2,…,N represents the number of the candidate solution in the population;

[0141] S52. Set the roughness difference factor Describes the roughness variation between grid regions to reflect the differences in local roughness on the concrete surface:

[0142]

[0143] Where, N adj (l i ) for the lth i The set of adjacent grids in a scaled grid region. For the l i Comprehensive roughness index at the scale:

[0144]

[0145] Wherein, ω1, ω2, ω3, ω4, and ω5 are preset weight coefficients;

[0146] S53. Introduce a dynamic weight adjustment mechanism to adjust the roughness difference factor ΔR.li Integrating the weight updates of bitter fish and prey, we set the bitter fish influence factor W at the t-th iteration. S (t) and prey influence factor W P (t):

[0147]

[0148] W P (t)=1-W S (t);

[0149] in, and These represent the maximum and minimum values ​​of the influence factor of bitter fish, respectively; T is the maximum number of iterations; and λ is the amplification factor of the roughness difference effect.

[0150] S54. In each iteration, for each candidate solution in the population Adopt improved update rules:

[0151]

[0152] Among them, S best (t) represents the individual bitter fish with the smallest detection error based on concrete surface roughness in the current population, P best (t) represents the current best prey individual, r1 and r2 are uniformly distributed random numbers, and α is the roughness gradient adjustment coefficient;

[0153] S55. Define a fitness function to evaluate candidate solutions. The deviation between the measured value and the actual concrete surface roughness, fitness function for:

[0154]

[0155] in, The first one obtained by experiment or field testing i Actual values ​​of concrete surface roughness under scale;

[0156] S56. When the fitness function When the preset convergence threshold ∈ is met or the maximum number of iterations T is reached, the l-th iteration is determined. i Optimal roughness parameters at scale The optimal candidate solution for the current fitness is given, and the set of optimal roughness parameters for all grid regions is represented as:

[0157]

[0158] Each parameter reflects both the local differences and overall characteristics of the texture of the roughened concrete surface.

[0159] In this embodiment, S6 includes the following steps:

[0160] S61. Calculate the multi-scale roughness weighting coefficient W based on the set of optimal roughness parameters for each grid region. li Multi-scale roughness weighting coefficients are used to reflect the contribution of different scale grid regions to the overall roughness:

[0161]

[0162] in, For the l i Comprehensive roughness index at the scale, is the average roughness across all grid scales, and κ is the local roughness variation influence factor, used to amplify or reduce the influence of the grid region on the overall roughness.

[0163] The multi-scale roughness weighting coefficient compares the differences between the comprehensive roughness index of each grid region and the global mean, giving regions with roughness changes greater than a threshold higher weight during the fusion process.

[0164] S62. Based on multi-scale roughness weighting coefficients Optimal roughness parameters for each grid region Weighted fusion is performed by introducing a local deviation correction term during the fusion process, so that the overall roughness assessment value reflects the true roughness distribution of the concrete surface, generating the fused roughness parameter P. fusion :

[0165]

[0166] in, λ1 is the mean of the optimal roughness parameters for all grid regions, and λ1 is the roughness deviation adjustment coefficient, which is used to correct the impact of local anomalies on the overall evaluation during the fusion process.

[0167] S63. Calculate the fusion roughness error E fusion This is used to evaluate the rationale for multi-scale fusion.

[0168]

[0169] The first term measures the deviation between the locally optimal roughness parameter and the fusion parameter, while the second term assesses the degree of deviation in the comprehensive roughness index, ensuring that the fusion result characterizes the roughness features of the roughened concrete surface. If the fusion roughness error meets the preset fusion error threshold ∈ f If the result is correct, the fusion is valid; otherwise, the weighting coefficients are adjusted. And recalculate the fusion roughness parameter P fusion ;

[0170] S64. Using a nonlinear fitting model F fit For the fusion roughness parameter P fusion The fitting and optimization process is performed to make the fused roughness parameters conform to the actual surface roughness distribution of concrete through an adaptive optimization process.

[0171]

[0172] Among them, P final The final output is the overall roughness test of the concrete.

[0173] Example 1:

[0174] On October 15, 2024, at the construction site of a large bridge, the construction team conducted a roughness test on the surface of the concrete pier after it had been roughened. The goal was to assess the roughening quality and ensure that the bonding strength between the old and new concrete met the design requirements. The construction team started the testing operation at 8:00 a.m. The on-site temperature was 22°C and the humidity was 55%. A standardized roughening machine was used to treat the surface of the pier, and the roughening depth was set within the range of 5mm to 8mm.

[0175] To ensure the scientific rigor of the testing, the construction team selected 50 roughened concrete samples from different construction areas. Before the formal testing, the team used traditional manual visual inspection methods for preliminary evaluation and found that there were significant differences in the scores given by different inspectors. In the example, for concrete sample numbered A15, inspector 1 gave it a roughness level of 3, while inspector 2 gave it a level of 4, a difference of 25%, which made it difficult for the construction team to make an accurate quality assessment decision.

[0176] The construction team used a Baslerac A2500-60uc industrial camera (5000×4000 pixels resolution) to acquire high-precision images of the roughened surfaces of 50 samples. Five images were taken of each sample from different angles to reduce the impact of lighting and angle. During the image acquisition process, sample number B07 had low image brightness due to insufficient lighting conditions. The system automatically adjusted the exposure parameters and finally obtained a clear texture image.

[0177] The acquired image data is input into the detection system. First, noise filtering and edge enhancement are performed to remove image artifacts caused by uneven lighting. After processing, sample C12 showed slight over-enhancement at the edges of the roughened area. The system detected this anomaly and automatically adjusted the filtering parameters to restore the edge detection effect to normal. Subsequently, the system uses a multi-scale grid segmentation algorithm to divide each image into a 10*10 grid region to capture texture features at different scales.

[0178] After completing the mesh segmentation, the system begins to extract the gray-level co-occurrence matrix and key texture features of the local binary mode for each mesh region, and generates a roughness feature vector. During this process, the feature vector of the sample with the number D22 deviates significantly from that of the standard roughened sample due to the shallow chiseling depth. The system automatically marks this sample as "low roughness risk" and prompts the construction party to review it.

[0179] Subsequently, the system employs an improved bitterfish optimization algorithm to optimize the feature vectors. The optimization process involves 30 iterations, ultimately obtaining the optimal roughness parameters for each grid region. In the example, the initial roughness parameter calculated for sample E35 was 5.12 μm, but after optimization, it was ultimately adjusted to 5.28 μm, improving the accuracy of the data.

[0180] After obtaining the optimal roughness parameters for all grid regions, the system uses a multi-scale weighted fusion model to calculate the overall roughness evaluation value. In the example, for sample number F19, the initial roughness parameters of each grid region ranged from 4.7 μm to 5.3 μm. After fusion, the final roughness evaluation value was determined to be 5.05 μm.

[0181] The system generates a test report, which includes the following:

[0182] 1. Testing time: 10:45 AM, October 15, 2024;

[0183] 2. Inspection location: A bridge construction site, pier number: B03;

[0184] 3. Detection method: Concrete roughness detection based on improved bitterfish optimization algorithm;

[0185] 4. Test Results:

[0186] Total sample size: 50;

[0187] Average roughness value: 5.08 μm;

[0188] Maximum roughness value: 5.51 μm (sample number: G41);

[0189] Minimum roughness value: 4.72 μm (sample number: D22, marked as low roughness risk);

[0190] Roughness pass rate: 96% (48 samples passed, 2 samples need to be re-inspected);

[0191] Experimental results show that the method of the present invention not only has higher detection accuracy, but also maintains stable detection performance under different lighting, temperature and humidity conditions. At the same time, the method of the present invention can automatically mark abnormal samples, reduce the cost of manual intervention, and improve the level of intelligence in construction quality management.

[0192] Ultimately, the construction team adjusted the construction process based on the test report, performing a second roughening treatment on the pier area where sample D22 was located, and completed the re-inspection at 12:30. The test results met the qualification standard. Thus, the concrete roughness test was successfully completed. The construction team confirmed that the method of this invention has superior testing performance and plans to fully promote its application in subsequent bridge construction.

[0193] This invention employs a multi-scale grid segmentation strategy to divide the roughened concrete surface into multi-level regions. It extracts the texture features of the concrete surface through grid segmentation image data at different scales, enabling the capture of micro and macro features of concrete surface roughness at different scales. Furthermore, it uses an adaptive fusion strategy to weighted integrate the optimal roughness parameters of each grid region, ensuring that the overall roughness detection results reflect detailed features while maintaining global consistency.

[0194] This invention improves the bitter fish optimization algorithm by introducing a roughness comparison coefficient on the basis of the traditional bitter fish optimization algorithm. The comprehensive roughness difference of different grid regions is used as the optimization weight to improve the algorithm's adaptability to complex texture features. In addition, a dynamic weight adjustment mechanism is adopted to dynamically adjust the influence factors of bitter fish and prey according to the roughness gradient during the optimization process. This ensures that the algorithm has a strong global search capability in the early stage and can finely adjust the roughness parameters in the later stage to avoid local optimum traps. When dealing with complex heterogeneous concrete roughened surfaces, it can more stably converge to the optimal solution.

[0195] This invention employs a nonlinear fitting model in the multi-scale roughness fusion process and utilizes an adaptive weight adjustment mechanism to perform secondary optimization on the fused roughness parameters in order to reduce the impact of local outliers on the overall detection results. The nonlinear fitting strategy can automatically adjust the parameters according to the changing trend of concrete surface texture, making the final detection results more consistent with the real roughness distribution.

[0196] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for detecting the roughness of concrete chiseling based on an improved bitterfish optimization algorithm, characterized in that, Includes the following steps: S1. Use image acquisition equipment to acquire images of the roughened concrete surface and obtain the original image data covering the detection area; S2. Perform multi-scale grid segmentation on the original image data, dividing the original image data into several grid segmentation image data of different scales according to the preset scale parameters; S3. Perform image preprocessing on multiple grid-segmented image data to generate optimized preprocessed grid-segmented image data; S4. Extract the texture features of the roughened concrete surface from the optimized preprocessed mesh segmentation image data to form a concrete surface roughness feature vector; S5. Input the concrete surface roughness feature vector into the improved bitter fish optimization algorithm, and optimize the roughness features in each grid region according to the dynamic weight adjustment mechanism and adaptive iteration strategy to obtain the optimal roughness parameter value in each grid region. S6. Perform global fusion processing on the optimal roughness parameter values ​​in each grid region, and use a unified fitting model to integrate the multi-scale detection results to generate the overall concrete roughness detection results. S7. Generate a test report based on the overall concrete roughness test results, including roughness numerical evaluation, distribution diagram and abnormal area prompts.

2. The method for detecting concrete roughness based on the improved bitterfish optimization algorithm according to claim 1, characterized in that, S2 includes the following steps: S21. Set the grid segmentation scale parameters for the roughened concrete surface, and define different scale levels based on the size range and surface roughness characteristics of the inspection area. ,in This represents the grid segmentation scale of the i-th layer, where i is the set number of scale levels; S22. Divide the original image data I into grid regions of different scale levels according to the grid segmentation scale parameter. Indicates the grid segmentation scale The segmented image data, grid region The size is determined by the following relationship: ； Where W and H are the width and height of the original image data I, respectively. and These represent the grid segmentation scales. The number of horizontal and vertical grid divisions below; S23. For each grid segmentation scale The grid area below Set the grid area overlap rate And obtain the final segmentation result of the grid region. : ； in, The preset overlap ratio; S24. For all mesh segmentation scales The grid area below Index and number the images to form a multi-scale grid segmentation image dataset. .

3. The method for detecting concrete roughness based on the improved bitterfish optimization algorithm according to claim 1, characterized in that, S3 includes the following steps: S31. Noise filtering is performed on the multi-scale grid segmented image data G. The image data of each grid region is processed to filter out random noise caused by changes in illumination, equipment noise and environmental interference, and the processed grid segmented image data is obtained. S32. Enhance the contrast of the noise-filtered mesh segmented image data and optimize the features of different roughness regions on the concrete surface to obtain the adjusted mesh segmented image data; S33. Perform edge detection on the contrast-enhanced grid segmentation image data, extract the key boundaries of the texture features of the roughened concrete surface, and obtain the extracted edge grid segmentation image data; S34. The edge-refined image data of the edge grid segmentation is processed to remove redundant edge information, and the effective edge features are enhanced by combining double threshold filtering to finally obtain optimized preprocessed grid segmentation image data. ; S35. Optimize preprocessing of grid-segmented image data for all grid segmentation scales. Data storage is performed to form an optimized preprocessed image data set: 。 4. The method for detecting concrete roughness based on the improved bitterfish optimization algorithm according to claim 1, characterized in that, S4 includes the following steps: S41. Based on the optimized preprocessed grid segmentation of the image data set For each grid segmentation scale, the image data is segmented into grids. Calculate the gray-level co-occurrence matrix : ； Where x and y are gray levels, and P and Q are the number of rows and columns of the grid-segmented image data, respectively. and This is relative displacement. This is an indicator function used to statistically analyze the joint probability distribution of grayscale values ​​of adjacent pixels; S42. Based on the gray-level co-occurrence matrix Calculate the contrast at each grid segmentation scale ,entropy Correlation and energy : ； in, , These are the mean values ​​of gray levels x and y, respectively. , These are the standard deviations; S43. Calculate the eigenvectors of the local binary pattern. : ； in, Let K be the number of neighboring pixels of pixel (p, q), and s(x) be the binary mapping function. S44. Vectorize the calculation results of contrast, entropy, correlation, energy, and local binary mode at all grid segmentation scales to construct a set of concrete surface roughness feature vectors: 。 5. The method for detecting concrete roughness based on the improved bitterfish optimization algorithm according to claim 4, characterized in that, S5 includes the following steps: S51. Input the set of concrete surface roughness feature vectors V into the improved Bitterfish optimization algorithm, and calculate the roughness contrast coefficient of each grid based on the information contrast for different grid regions of the roughened concrete surface: ； in, To prevent constants with a denominator of zero; Roughness comparison coefficient As the initial population for generating weighting factors Each individual Initial roughness estimate: ； in, and For preset calibration coefficients, For random numbers that follow a uniform distribution, Indicates the number of the candidate solution in the population; S52. Set the roughness difference factor Describes the roughness variation between grid regions to reflect the differences in local roughness on the concrete surface: ； in, In order to be with the first The set of adjacent grids in a scaled grid region. For the first Comprehensive roughness index at the scale: ； in, Preset weighting coefficients; S53. Introduce a dynamic weight adjustment mechanism to adjust the roughness difference factor. Integrating the weight updates of bitter fish and prey, the influence factor of bitter fish is set at the t-th iteration. Influence factor of prey : ； in, and These represent the maximum and minimum values ​​of the bitter fish influence factor, respectively, where T is the maximum number of iterations. This is the amplification factor for the effect of roughness difference; S54. In each iteration, for each candidate solution in the population Adopt improved update rules: ； in, This is the individual bitter fish with the smallest detection error based on concrete surface roughness in the current population. The best prey individual at present. , Uniformly distributed random numbers, This is the roughness gradient adjustment coefficient; S55. Define a fitness function to measure candidate solutions. The deviation between the measured value and the actual concrete surface roughness, fitness function for: ； in, The first obtained by experiment or field testing Actual values ​​of concrete surface roughness under scale; S56. When the fitness function Meets the preset convergence threshold Or, when the maximum number of iterations T is reached, determine the first iteration. Optimal roughness parameters at scale The optimal candidate solution for the current fitness is given, and the set of optimal roughness parameters for all grid regions is represented as: ； Each parameter reflects both the local differences and overall characteristics of the texture of the roughened concrete surface.

6. The method for detecting concrete roughness based on the improved bitterfish optimization algorithm according to claim 1, characterized in that, S6 includes the following steps: S61. Calculate the multi-scale roughness weighting coefficient based on the set of optimal roughness parameter values ​​in each grid region. Multi-scale roughness weighting coefficients are used to reflect the contribution of different scale grid regions to the overall roughness: ； in, For the first Comprehensive roughness index at the scale, This represents the average roughness across all grid scales. This is a local roughness variation influencing factor, used to amplify or reduce the impact of the mesh region on the overall roughness; The multi-scale roughness weighting coefficient compares the differences between the comprehensive roughness index of each grid region and the global mean, giving regions with roughness changes greater than a threshold higher weight during the fusion process. S62. Based on multi-scale roughness weighting coefficients Optimal roughness parameters for each grid region Weighted fusion is performed by introducing a local deviation correction term during the fusion process, so that the overall roughness assessment value reflects the true roughness distribution of the concrete surface, thus generating fused roughness parameters. : ； in, The mean of the optimal roughness parameters for all grid regions. This is the roughness deviation adjustment coefficient, used to correct the impact of local anomalies on the overall evaluation during the fusion process; S63. Calculate the fusion roughness error This is used to evaluate the rationale for multi-scale fusion. ； The first term measures the deviation between the local optimal roughness parameter and the fusion parameter, while the second term assesses the degree of deviation in the comprehensive roughness index, ensuring that the fusion result characterizes the roughness features of the roughened concrete surface. If the fusion roughness error meets a preset fusion error threshold... If the result is correct, the fusion is valid; otherwise, the weighting coefficients are adjusted. And recalculate the fusion roughness parameters ; S64. Employing a nonlinear fitting model For fused roughness parameters The fitting and optimization process is performed to make the fused roughness parameters conform to the actual surface roughness distribution of concrete through an adaptive optimization process. ； in, The final output is the overall roughness test of the concrete.

Images