A method, system, medium, and equipment for extracting pitting corrosion features of ribbed steel bars.

Abstract

This invention discloses a method, system, medium, and device for extracting pitting corrosion features of ribbed steel bars. The method includes the following steps: S1, acquiring and preprocessing point cloud data of rusted ribbed steel bars to obtain a three-dimensional surface of the surface contour of the rusted ribbed steel bars; S2, removing the envelope from the three-dimensional surface to obtain a lower envelope surface and a removed lower envelope surface; S3, estimating the power spectral density of the removed lower envelope surface to obtain a two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bars; S4, designing a band-stop filter based on the two-dimensional power spectral density map and filtering the removed lower envelope surface C(r,θ,z) to obtain the filtered removed lower envelope surface C of the rusted ribbed steel bars. bs (r,θ,z), combined with the lower envelope surface, the ribbed surface contour surface is calculated; S5, the ribbed surface contour surface A is de-trended to obtain the pitting feature surface; S6, the pitting feature surface is extracted by image segmentation algorithm. This invention achieves accurate extraction of pitting features of ribbed steel bars.

Description

Technical Field

[0001] This invention belongs to the field of reinforced concrete structure performance testing technology, specifically, it relates to a method, system, medium and equipment for extracting pitting corrosion features of ribbed steel bars. Background Technology

[0002] Reinforced concrete structures are among the most common structural forms in civil engineering, bridges, ports, and special structures. However, due to long-term use and service environments, the reinforcing steel in reinforced concrete structures is prone to corrosion, leading to reduced ductility and premature yielding, which significantly impacts structural performance and service life. Furthermore, due to variations in the service environment and the non-uniformity of the concrete cover and the passivation film on the reinforcing steel, non-uniform corrosion and pitting corrosion along the length of the reinforcing steel are frequently observed in practice. This affects the failure mode and mechanical properties of the steel, and the randomness of steel corrosion increases the probability of failure in corroded reinforced concrete members. Therefore, extracting the morphological characteristics of steel corrosion is a crucial aspect of research on steel corrosion in concrete.

[0003] With technological advancements, methods for measuring rebar cross-sectional loss have evolved from traditional weighing and drainage methods to modern three-dimensional scanning. Three-dimensional scanners allow for precise and effective measurement of the remaining cross-sectional area of ​​the rebar and the dimensions of rust pits, among other corrosion morphological characteristics. However, current technologies rely solely on extracting and analyzing the corrosion characteristics of plain round rebar to characterize the degree and pattern of corrosion in reinforced concrete structures. Ribbed rebar, due to its excellent bond with concrete, is more widely used in practical engineering. The periodic ribs of ribbed rebar increase the irregularity of the rebar profile and also affect the unevenness of corrosion, making it a crucial factor that cannot be ignored in the performance evaluation of corroded reinforced concrete structures.

[0004] Therefore, effectively filtering out the influence of periodic ribs on the rusted surface of ribbed steel bars with irregular initial contours, and thus accurately extracting the pitting corrosion characteristics of ribbed steel bars, has become an indispensable part of the performance evaluation of rusted reinforced concrete structures. This is beneficial for studying the influence of steel bar type on corrosion mode and for better understanding the development of steel bar corrosion in reinforced concrete structures in actual engineering. Summary of the Invention

[0005] The primary objective of this invention is to overcome the shortcomings and deficiencies of existing technologies and provide a method for extracting pitting corrosion features of ribbed steel bars, thereby solving the problem of irregular outlines of ribbed steel bars and achieving accurate extraction of pitting corrosion features of rusted ribbed steel bars.

[0006] The second objective of this invention is to provide a system for extracting pitting corrosion features of ribbed steel bars.

[0007] A third objective of this invention is to provide a storage medium.

[0008] The fourth objective of this invention is to provide a computing device.

[0009] The objective of this invention is achieved through the following technical solution: a method for extracting pitting corrosion features of ribbed steel bars, comprising the following steps:

[0010] S1. Obtain the point cloud data of the rusted ribbed steel bar and perform preprocessing to obtain the three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar.

[0011] S2. The three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar is de-envelope processed to separate the lower envelope surface B(r,θ,z) and the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar.

[0012] S3. The power spectral density of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar is estimated to obtain a two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar.

[0013] S4. Design a band-stop filter based on the two-dimensional power spectral density diagram and filter out the residual transverse rib signal of the de-envelope surface C(r,θ,z) of the corroded ribbed steel bar to obtain the filtered de-envelope surface C of the corroded ribbed steel bar. bs (r,θ,z), combined with the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar, the deribbed surface profile surface A of the corroded ribbed steel bar is calculated. bs (r,θ,z);

[0014] S5. Deribbed surface profile A of rusted ribbed steel bars. bs (r,θ,z) is detrended to obtain the pitting corrosion characteristic surface P(r,θ,z) of the corroded ribbed steel bar;

[0015] S6. Extract the rust pits from the pitting feature surface P(r,θ,z) using an image segmentation algorithm to obtain the geometric features of the rust pits.

[0016] Preferably, step S1 specifically includes:

[0017] S11. Obtain the three-dimensional point cloud coordinate data of the surface morphology of the rusted ribbed steel bar through three-dimensional scanning technology, and generate a three-dimensional solid model of the rusted ribbed steel bar through three-dimensional modeling software.

[0018] S12. Along the axial direction of the corroded ribbed steel bar, several cross-sections of the corroded ribbed steel bar are intercepted at intervals of a preset length l; a three-dimensional rectangular coordinate system of the corroded ribbed steel bar is established with the length direction of the corroded ribbed steel bar as the z-axis and the cross-section of the corroded ribbed steel bar as the (x, y) plane. Then, the three-dimensional rectangular coordinates of the data points of the cross-section are represented as (x, y, z), and through coordinate transformation, the three-dimensional polar coordinates (r, θ, z) of the data points of the cross-section are obtained:

[0019]

[0020]

[0021] Among them, r is the radius corresponding to the data point of the cross-section of the corroded steel bar, and θ is the angle corresponding to the data point of the cross-section of the corroded steel bar;

[0022] S13. The data points of all cross-sections of the corroded ribbed steel bar form a three-dimensional surface A(r, θ, z) of the surface profile of the corroded ribbed steel bar in the three-dimensional polar coordinate system.

[0023] Preferably, step S2 specifically includes:

[0024] S21. Take the derivative of the three-dimensional surface A(r, θ, z) of the surface profile of the corroded ribbed steel bar, and find all minimum points A(r low , θ low , z low ) of the three-dimensional surface A(r, θ, z) at a data interval d. The minimum point A(r low , θ low , z low ) satisfies formulas (3) and (4):

[0025] (z0 - z low ) 2 +(θ0 - θ low ) 2 ≤d 2 , Formula (3),

[0026] r low < r0, Formula (4),

[0027] Among them, A(r0, θ0, z0) is the coordinate point on the three-dimensional surface A(r, θ, z) that satisfies formula (3), and the data interval d satisfies l R <d<2l R , l R is the transverse rib spacing of an uncorroded ribbed steel bar with the same diameter as the corroded ribbed steel bar;

[0028] S22. Using all the minimum points as the endpoints of the lower envelope surface, the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar is obtained by cubic spline interpolation.

[0029] S23. Based on the three-dimensional surface A(r,θ,z) of the rusted ribbed steel bar surface profile and the lower envelope surface B(r,θ,z) of the rusted ribbed steel bar, the lower envelope surface C(r,θ,z) is calculated as follows:

[0030] C(r,θ,z)=A(r,θ,z)-B(r,θ,z), formula (5).

[0031] Preferably, step S3 specifically includes:

[0032] S31. Divide the lower envelope surface C(r,θ,z) of the corroded ribbed steel bar into M segments along the θ axis, representing the residual transverse rib signal curves of the corroded ribbed steel bar. Where i is the sequential number, i = 1, 2, 3...M, M ≥ 360;

[0033] S32. The power spectral density of the residual transverse rib signal curves of each segment of rusted ribbed steel bar is estimated by the Welch power spectral estimation method to obtain the corresponding M segments of power spectral density curves. The M segments of power spectral density curves are numbered in sequence i = 1, 2, 3... M and combined to obtain the two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar.

[0034] Preferably, step S4 specifically includes:

[0035] S41. Based on the two-dimensional power spectral density diagram of the residual transverse rib signal of the corroded ribbed steel bar, the main repetition period T and the frequency ω2 of the second harmonic of the residual transverse rib signal are obtained, where T = 1 / ω1, ω1 is the main frequency corresponding to the first frequency band along the frequency increasing direction in the two-dimensional power spectral density diagram, and ω2 is the frequency corresponding to the second frequency band along the frequency increasing direction.

[0036] S42. Design a pair of second-order Butterworth bandstop filters centered on the main frequency ω1 and the second harmonic frequency ω2, respectively. The bandwidth of the second-order Butterworth bandstop filter is b. w Based on the two-dimensional power spectral density plot, the transfer function H of the second-order Butterworth bandstop filter is determined. j (s) is:

[0037]

[0038] Where s is a complex frequency variable, ω j ζ represents the center frequency of the band-stop filter, and j=1 and j=2 represent the main frequency ω1 and the second harmonic frequency ω2 of the residual transverse rib signal of the corroded ribbed steel bar, respectively. jFor the damping ratio, ζ j The calculation formula is:

[0039]

[0040] S43. The residual transverse rib signal of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar is filtered out by a pair of second-order Butterworth band-stop filters from step S42, to obtain the filtered de-envelope surface C of the rusted ribbed steel bar. bs (r,θ,z);

[0041] S44. The lower envelope surface B(r,θ,z) of the rusted ribbed steel bar and the lower envelope surface C of the filtered rusted ribbed steel bar are obtained. bs Adding (r, θ, z) yields the deribbed surface profile A of the rusted ribbed steel bar. bs (r,θ,z):

[0042] A bs (r,θ,z)=B(r,θ,z)+C bs (r,θ,z), formula (8).

[0043] Preferably, step S5 specifically includes:

[0044] S51, the deribbed surface contour surface A of the rusted ribbed steel bar is... bs (r,θ,z) is the deribbed surface profile curve of the rusted ribbed steel bar divided into M segments along the θ axis. Where i is the sequential number, i = 1, 2, 3...M, M ≥ 360;

[0045] S52. Using a pre-defined trend model, the deribbed surface profile curve of each segment of corroded ribbed steel bar is calculated using the least squares method. Fit the trend line Where i is the sequential number, i = 1, 2, 3…M, M ≥ 360, and M segments of trend lines are defined. The ribbed surface profile trend surface U(r,θ,z) of the deribbed steel bar is obtained by arranging them in sequence according to the numbering order.

[0046] S53, the deribbed surface contour surface A of the rusted ribbed steel bar. bs Subtract the de-ribbed surface profile trend surface U(r,θ,z) of the rusted ribbed steel bar from (r,θ,z) to obtain the de-ribbed surface profile trend surface D(r,θ,z) of the rusted ribbed steel bar.

[0047] S54. Select the data points with r greater than 0 in the detrended surface D(r,θ,z) of the de-stressed surface profile of the rusted ribbed steel bar, set the r value of the data point to 0, and smooth the detrended surface D(r,θ,z) to obtain the pitting feature surface P(r,θ,z).

[0048] Preferably, in step S6, the image segmentation algorithm includes the watershed segmentation method, specifically including: extending the pitting feature surface P(r,θ,z) along the θ axis and performing watershed segmentation processing to extract each rust pit, and calculating the geometric features of each rust pit, including area, length, width and depth.

[0049] A system for extracting pitting corrosion features of ribbed steel bars, specifically comprising:

[0050] The data preprocessing module is used to acquire point cloud data of rusted ribbed steel bars and perform preprocessing to obtain the three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bars.

[0051] The de-envelope processing module is used to perform de-envelope processing on the three-dimensional curved surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar, and separate the lower envelope curved surface B(r,θ,z) and the de-envelope curved surface C(r,θ,z) of the rusted ribbed steel bar.

[0052] The power spectral density estimation module is used to estimate the power spectral density of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar, and obtain a two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar.

[0053] The filtering module is used to design a band-stop filter based on the two-dimensional power spectral density map and filter out the residual transverse rib signal of the lower envelope surface C(r,θ,z) of the rusted ribbed steel bar, thereby obtaining the filtered lower envelope surface C of the rusted ribbed steel bar. bs (r,θ,z), combined with the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar, the deribbed surface profile surface A of the corroded ribbed steel bar is calculated. bs (r,θ,z);

[0054] The detrending processing module is used to process the deribbed surface profile A of rusted ribbed steel bars. bs (r,θ,z) is detrended to obtain the pitting corrosion characteristic surface P(r,θ,z) of the corroded ribbed steel bar;

[0055] The image segmentation module is used to extract the rust pits from the pitting feature surface P(r,θ,z) using an image segmentation algorithm to obtain the geometric features of the rust pits.

[0056] A storage medium storing a program, which, when executed by a processor, implements the above-described method for extracting pitting corrosion features of ribbed steel bars.

[0057] A computing device includes a processor and a memory for storing processor-executable programs. When the processor executes the program stored in the memory, it implements the above-described method for extracting pit erosion features of ribbed steel bars.

[0058] The present invention has the following advantages and effects compared with the prior art:

[0059] (1) The present invention forms a three-dimensional curved surface of the surface profile of the rusted ribbed steel bar in a three-dimensional polar coordinate system through coordinate transformation, which simplifies the description of the geometric features of the rusted ribbed steel bar. It can clearly represent the depth and distribution of pitting corrosion by the change of radial distance and azimuth angle in polar coordinates, which makes it easier to identify and extract the pitting corrosion distribution characteristics of the rusted ribbed steel bar more intuitively.

[0060] (2) The present invention separates the lower envelope surface and the lower envelope surface of the rusted ribbed steel bar from the three-dimensional curved surface of the surface contour of the rusted ribbed steel bar by the de-envelope method, effectively suppressing the mutual influence between the remaining inner diameter surface and the residual transverse rib signal surface, and obtaining the lower envelope surface with more significant transverse rib signal periodic characteristics.

[0061] (3) The present invention identifies the main repetition period and the frequency of the second harmonic of the residual transverse rib signal of the rusted ribbed steel bar by means of two-dimensional power spectral density estimation method. Then, by designing a band-stop filter, the residual transverse rib signal in the de-envelope surface of the rusted ribbed steel bar is filtered out, and the de-ribbed surface profile of the rusted ribbed steel bar is obtained, which makes it easier to describe and analyze the pitting corrosion characteristics of the rusted ribbed steel bar more accurately.

[0062] (4) By detrending, the present invention not only eliminates the influence of non-uniform corrosion on the identification of rust pits, but also makes the extracted geometric features of rust pits more accurate and reliable. Compared with existing rust pit feature extraction technology, the present invention greatly improves the accuracy in distinguishing adjacent rust pits and calculating the size of rust pits.

[0063] (5) This invention provides a method for extracting pitting corrosion features of ribbed steel bars based on the de-envelope method and frequency domain filtering, which solves the problem of irregular initial contours of ribbed steel bars, can effectively identify the corrosion mode of ribbed steel bars and accurately extract pitting corrosion features, and provides technical support for better performance evaluation of corroded reinforced concrete. Attached Figure Description

[0064] Figure 1 This is a flowchart illustrating the method for extracting pitting corrosion features of ribbed steel bars according to the present invention.

[0065] Figure 2This is a three-dimensional scanning model of a corroded ribbed steel bar, which is part of the method for extracting pitting corrosion features of ribbed steel bars according to the present invention.

[0066] Figure 3 (a) and (b) are respectively the surface contour surface and top view of the rusted ribbed steel bar and the ribbed steel bar pitting corrosion feature extraction method of the present invention.

[0067] Figure 4 (a) and (b) are two-dimensional power spectral density maps and top views of the residual transverse rib signals of rusted ribbed steel bars according to the method for extracting pitting corrosion features of ribbed steel bars of the present invention.

[0068] Figure 5 (a) and (b) are respectively the surface contour of the deribbed steel bar and its top view of the ribbed steel bar pitting corrosion feature extraction method of the present invention.

[0069] Figure 6 This is a three-dimensional scanning model of a rusted steel bar after rib removal, based on the present invention's method for extracting pitting corrosion features of ribbed steel bars.

[0070] Figure 7 (a) and (b) are respectively the detrended surface and top view of the deribbed surface profile of the rusted steel bar and the top view of the method for extracting pitting corrosion features of ribbed steel bars according to the present invention.

[0071] Figure 8 (a) and (b) are respectively the extended surface contour of the corroded steel bar and the watershed segmentation results of the method for extracting pit corrosion features of ribbed steel bars according to the present invention. Detailed Implementation

[0072] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0073] Example 1

[0074] like Figure 1 The diagram shows a flowchart of a method for extracting pitting corrosion features of ribbed steel bars, including the following steps:

[0075] S1. Obtain the point cloud data of the rusted ribbed steel bar and perform preprocessing to obtain the three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar.

[0076] S2. The three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar is de-envelope processed to separate the lower envelope surface B(r,θ,z) and the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar.

[0077] S3. The power spectral density of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar is estimated to obtain a two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar.

[0078] S4. Design a band-stop filter based on the two-dimensional power spectral density diagram and filter out the residual transverse rib signal of the de-envelope surface C(r,θ,z) of the corroded ribbed steel bar to obtain the filtered de-envelope surface C of the corroded ribbed steel bar. bs (r,θ,z), combined with the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar, the deribbed surface profile surface A of the corroded ribbed steel bar is calculated. bs (r,θ,z);

[0079] S5. Deribbed surface profile A of rusted ribbed steel bars. bs (r,θ,z) is detrended to obtain the pitting corrosion characteristic surface P(r,θ,z) of the corroded ribbed steel bar;

[0080] S6. Extract the rust pits from the pitting feature surface P(r,θ,z) using an image segmentation algorithm to obtain the geometric features of the rust pits.

[0081] Specifically, this invention provides a method for extracting pitting corrosion features of ribbed steel bars based on de-envelope method and frequency domain filtering. The de-envelope method separates the lower envelope surface and the de-envelope surface of the rusted ribbed steel bar from the three-dimensional surface of the surface contour of the rusted ribbed steel bar. These correspond to the remaining inner diameter surface and the residual transverse rib signal surface of the rusted ribbed steel bar, respectively. The remaining inner diameter surface of the rusted ribbed steel bar belongs to the concept of the steel bar solid level, while the lower envelope surface belongs to the concept of the digital signal level. The remaining inner diameter surface is a surface mixed with some noise signal. This step effectively suppresses the mutual influence between the remaining inner diameter surface and the residual transverse rib signal surface, and obtains the de-envelope surface with more significant transverse rib signal periodic characteristics.

[0082] Then, by using two-dimensional power spectral density estimation, the main repetition period and the frequency of the second harmonic of the residual transverse rib signal of the rusted ribbed steel bar are accurately identified. Thus, by designing a band-stop filter, the residual transverse rib signal in the de-envelope surface of the rusted ribbed steel bar is filtered out. Combined with the lower envelope surface, the de-ribbed surface profile of the rusted ribbed steel bar is obtained. This step facilitates a more accurate description and analysis of the pitting corrosion characteristics of the rusted ribbed steel bar.

[0083] Then, the detrending process is performed on the deribbed surface contour of the rusted ribbed steel bar to obtain the pitting feature surface. This step not only eliminates the influence of non-uniform corrosion on the identification of rust pits, but also makes the extracted geometric features of rust pits more accurate and reliable.

[0084] Finally, image segmentation processing is performed on the pitting feature surface to extract each rust pit and obtain its geometric features. Compared with existing rust pit feature extraction techniques, this invention greatly improves accuracy in distinguishing adjacent rust pits and calculating rust pit dimensions. This invention solves the problem of irregular initial contours of ribbed steel bars, effectively identifies the rust patterns of ribbed steel bars, and accurately extracts pitting features, providing technical support for better performance evaluation of rusted reinforced concrete.

[0085] Step S1 specifically includes:

[0086] S11. Obtain the three-dimensional point cloud coordinate data of the surface morphology of the rusted ribbed steel bar through three-dimensional scanning technology, and generate a three-dimensional solid model of the rusted ribbed steel bar through three-dimensional modeling software.

[0087] S12. Along the axial direction of the corroded ribbed steel bar, cut several cross sections of the corroded ribbed steel bar at predetermined lengths l; establish a three-dimensional rectangular coordinate system for the corroded ribbed steel bar with the length direction of the corroded ribbed steel bar as the z-axis and the cross section of the corroded ribbed steel bar as the (x,y) plane. Then, the three-dimensional rectangular coordinates of the data points of the cross section are represented as (x,y,z). Then, through coordinate transformation, obtain the three-dimensional polar coordinates (r,θ,z) of the data points of the cross section:

[0088]

[0089]

[0090] Where r is the radius corresponding to the data point on the cross-section of the corroded steel bar, and θ is the angle corresponding to the data point on the cross-section of the corroded steel bar;

[0091] S13. Form a three-dimensional surface A(r,θ,z) of the surface profile of the rusted ribbed steel bars in a three-dimensional polar coordinate system using the data points of the cross-sections of all the rusted ribbed steel bars.

[0092] Specifically, in this embodiment, a rusted ribbed steel bar with a corrosion rate of 9.91% is removed from the demolition components of a high-pile wharf in a certain seaport. The rusted ribbed steel bar is scanned using a laser scanner or a structured light scanner, and the three-dimensional point cloud data of the scanned rusted ribbed steel bar is saved as a point cloud format.

[0093] Preprocessing of point cloud data includes: modeling, establishing a 3D coordinate system, and transforming the coordinates into 3D polar coordinates. First, a 3D solid model of the rusted ribbed steel bar (Rebar) is generated using 3D modeling software, such as... Figure 2As shown. Then, along the length direction of the three-dimensional solid model of the corroded ribbed steel bar Rebar, that is, the axial direction, several cross-sections are intercepted. The interception interval is determined according to a preset length, and the preset length is selected according to the characteristic dimension accuracy required by the research. In this embodiment, l = 1 mm. Taking the length direction of the three-dimensional solid model as the z-axis and the cross-section as the (x, y) plane, the data points of each cross-section are represented by three-dimensional coordinates (x, y, z). Then, through coordinate transformation, the three-dimensional point cloud coordinates of the corroded ribbed steel bar Rebar are observed in polar coordinate form. Finally, the data points of each cross-section form a three-dimensional surface A(r, θ, z) of the surface contour of the corroded ribbed steel bar in the polar coordinate system (r, θ, z). Figure 3 (a) and (b) respectively show the three-dimensional surface of the surface contour of the corroded ribbed steel bar and its top view. In the present invention, a three-dimensional surface of the surface contour of the corroded ribbed steel bar is formed in a three-dimensional polar coordinate system through coordinate transformation, simplifying the geometric feature description of the corroded ribbed steel bar. The depth and distribution of pitting corrosion can be clearly represented by the changes in the radial distance (corresponding to the radius r of the data point corresponding to the cross-section of the corroded steel bar) and the azimuth angle (corresponding to the angle θ of the data point corresponding to the cross-section of the corroded steel bar) in the polar coordinate system, facilitating the more intuitive identification and extraction of the pitting corrosion distribution characteristics of the corroded ribbed steel bar.

[0094] Step S2 specifically includes:

[0095] S21. Differentiate the three-dimensional surface A(r, θ, z) of the surface contour of the corroded ribbed steel bar, and find all the minimum points A(r low , θ low , z low ) of the three-dimensional surface A(r, θ, z) with a data interval d. The minimum point A(r low , θ low , z low ) satisfies formulas (3) and (4):

[0096] (z0 - z low ) 2 +(θ0 - θ low ) 2 ≤ d 2 , Formula (3).

[0097] r low < r0, Formula (4).

[0098] Where A(r0, θ0, z0) is the coordinate point in the three-dimensional surface A(r, θ, z) that satisfies formula (3), and the data interval d satisfies l R < d < 2l R , l R is the cross-rib spacing of an uncorroded ribbed steel bar with the same diameter as the corroded ribbed steel bar.

[0099] S22. Using all the minimum points as the endpoints of the lower envelope surface, the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar is obtained by cubic spline interpolation.

[0100] S23. Based on the three-dimensional surface A(r,θ,z) of the rusted ribbed steel bar surface profile and the lower envelope surface B(r,θ,z) of the rusted ribbed steel bar, the lower envelope surface C(r,θ,z) is calculated as follows:

[0101] C(r,θ,z)=A(r,θ,z)-B(r,θ,z), formula (5).

[0102] Specifically, the transverse rib spacing is a parameter determined and unchanged during the production of the same batch of reinforcing bars. When it is impossible to obtain uncorroded ribbed reinforcing bars from the same batch due to long service time, this parameter is often obtained by measuring the spacing of the remaining transverse ribs of the corroded reinforcing bars. Multiple sets of relatively intact adjacent transverse ribs in the corroded reinforcing bars can be selected for measurement, and their average value is taken as the transverse rib spacing. In this embodiment, the transverse rib spacing l of the Rebar is obtained by physically measuring the main repetition cycle of the transverse ribs of the corroded ribbed reinforcing bars Rebar. R The value is taken as 6.4 mm. Therefore, according to l... R <d<2l R In this embodiment, the data interval d can be 7mm. Using a data interval d = 7mm, all minimum points of the three-dimensional surface A(r,θ,z) of the rusted ribbed steel bar surface profile are found. A(r0,θ0,z0) is the minimum point on the surface A(r,θ,z) that satisfies the condition (θ...). low ,z low Let A(r, θ, z) be the points within a radius of a circle with center d. The method for finding the minimum point is to compare the coordinates (θ, θ, z) on the surface A(r, θ, z). low ,z low The corresponding r low Find the point that satisfies both equations (3) and (4) by considering the magnitude of r0 corresponding to the coordinates (θ0, z0). This point is the minimum point A(r). low ,θ low ,z low Using all the found minimum points as the endpoints of the lower envelope surface, and then processing them using the common cubic spline interpolation method, the lower envelope surface B(r,θ,z) of the rusted ribbed steel bar is obtained. Subtracting the lower envelope surface B(r,θ,z) of the rusted ribbed steel bar from the three-dimensional surface A(r,θ,z) of the surface profile of the rusted ribbed steel bar, we obtain the de-lower envelope surface C(r,θ,z) of the rusted ribbed steel bar.

[0103] Step S3 specifically includes:

[0104] S31. Divide the lower envelope surface C(r,θ,z) of the corroded ribbed steel bar into M segments along the θ axis, representing the residual transverse rib signal curves of the corroded ribbed steel bar. Where i is the sequential number, i = 1, 2, 3...M, M ≥ 360;

[0105] S32. The power spectral density of the residual transverse rib signal curves of each segment of rusted ribbed steel bar is estimated by the Welch power spectral estimation method to obtain the corresponding M segments of power spectral density curves. The M segments of power spectral density curves are numbered in sequence i = 1, 2, 3... M and combined to obtain the two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar.

[0106] Specifically, in this embodiment, when the value M is 512, the lower envelope surface C(r,θ,z) of the corroded ribbed steel bar is divided into 512 segments along the θ axis, representing the residual transverse rib signal curves of the corroded ribbed steel bar. The residual transverse rib signal curves of 512 rusted ribbed steel bars were analyzed using the common Welch power spectrum estimation method. Two-dimensional power spectral density estimation is performed to obtain the power spectral density curve corresponding to the residual transverse rib signal of each segment of the corroded ribbed steel bar. The M segments of power spectral density curves are then sequentially numbered i, i.e., these power spectral density curves are numbered 1, 2, 3, ..., 512 in ascending order along the θ-axis, resulting in a two-dimensional power spectral density map of the residual transverse rib signal of the corroded ribbed steel bar. (See figure) Figure 4 As shown in (a) and (b), these are two-dimensional power spectral density maps and top views of the residual transverse rib signals of rusted ribbed steel bars.

[0107] Step S4 specifically includes:

[0108] S41. Based on the two-dimensional power spectral density diagram of the residual transverse rib signal of the corroded ribbed steel bar, the main repetition period T and the frequency ω2 of the second harmonic of the residual transverse rib signal are obtained, where T = 1 / ω1, ω1 is the main frequency corresponding to the first frequency band along the frequency increasing direction in the two-dimensional power spectral density diagram, and ω2 is the frequency corresponding to the second frequency band along the frequency increasing direction.

[0109] S42. Design a pair of second-order Butterworth bandstop filters centered on the main frequency ω1 and the second harmonic frequency ω2, respectively. The bandwidth of the second-order Butterworth bandstop filter is b. w Based on the two-dimensional power spectral density plot, the transfer function H of the second-order Butterworth bandstop filter is determined. j (s) is:

[0110]

[0111] Where s is a complex frequency variable, ω jζ represents the center frequency of the band-stop filter, and j=1 and j=2 represent the main frequency ω1 and the second harmonic frequency ω2 of the residual transverse rib signal of the corroded ribbed steel bar, respectively. j For the damping ratio, ζ j The calculation formula is:

[0112] ζ j =2ω j / b w , formula (7);

[0113] S43. The residual transverse rib signal of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar is filtered out by a pair of second-order Butterworth band-stop filters from step S42, to obtain the filtered de-envelope surface C of the rusted ribbed steel bar. bs (r,θ,z);

[0114] S44. The lower envelope surface B(r,θ,z) of the rusted ribbed steel bar and the lower envelope surface C of the filtered rusted ribbed steel bar are obtained. bs Adding (r, θ, z) yields the deribbed surface profile A of the rusted ribbed steel bar. bs (r,θ,z):

[0115] A bs (r,θ,z)=B(r,θ,z)+C bs (r,θ,z), Formula (8).

[0116] Specifically, by observing the two-dimensional power spectral density diagram, the dominant repetition period T and the second harmonic frequency ω2 of the rebar rib signal are identified. In this embodiment, as... Figure 4 As shown in (b), the frequencies of the principal frequency ω1 and the second harmonic ω2 of the corroded ribbed steel bar are calculated to be 0.1563 mm. -1 and 0.2849mm -1 ;

[0117] like Figure 4 As shown in (b), the bandwidth b is determined based on the two-dimensional power spectral density map of the residual transverse rib signal of the reinforcing bar. W =0.4mm -1 The dominant frequency ω1 of the residual transverse rib signal of the corroded ribbed steel bar was 0.1563 mm. -1 The frequency of the second harmonic is ω2 = 0.2849 mm. -1 Centered on b W =0.4mm -1 To determine the bandwidth, a pair of second-order Butterworth bandstop filters are designed. According to formula (7), ζ1 = 0.7815 and ζ2 = 1.425 are calculated. Therefore, the transfer functions H1(s) and H2(s) of the pair of second-order Butterworth bandstop filters are respectively:

[0118]

[0119]

[0120] The residual transverse rib signal on the surface of the rusted ribbed steel bar Rebar is filtered out by the aforementioned pair of second-order Butterworth band-stop filters, resulting in the filtered surface profile of the rusted ribbed steel bar Rebar, as shown below. Figure 5 Figures (a) and (b) show the surface contour of the deribbed steel bar and its top view, respectively. Figure 6 The image shown is a 3D scan model of a rusted ribbed steel bar after rib removal. It can be seen that the transverse ribs on the surface of the rusted ribbed steel bar have clearly disappeared. The method of this invention can effectively filter out the influence of periodic rib shapes, thereby accurately identifying each rust pit in the rusted ribbed steel bar.

[0121] Step S5 specifically includes:

[0122] S51, the deribbed surface contour surface A of the rusted ribbed steel bar is... bs (r,θ,z) is the deribbed surface profile curve of the rusted ribbed steel bar divided into M segments along the θ axis. Where i is the sequential number, i = 1, 2, 3...M, M ≥ 360;

[0123] S52. Select the preset trend model, and use the least squares method to calculate the deribbed surface profile curve of each segment of rusted ribbed steel bar. Fit the trend line Where i is the sequential number, i = 1, 2, 3…M, M ≥ 360, and M segments of trend lines are defined. The ribbed surface profile trend surface U(r,θ,z) of the deribbed steel bar is obtained by arranging them in sequence according to the numbering order.

[0124] S53, the deribbed surface contour surface A of the rusted ribbed steel bar. bs Subtract the de-ribbed surface profile trend surface U(r,θ,z) of the rusted ribbed steel bar from (r,θ,z) to obtain the de-ribbed surface profile trend surface D(r,θ,z) of the rusted ribbed steel bar.

[0125] S54. Select the data points with r greater than 0 in the detrended surface D(r,θ,z) of the de-stressed surface profile of the rusted ribbed steel bar, set the r value of the data point to 0, and smooth the detrended surface D(r,θ,z) to obtain the pitting feature surface P(r,θ,z).

[0126] Specifically, the preset trend models include second-order polynomial functions, third-order polynomial functions, fourth-order polynomial functions, and Gaussian functions, with one selected based on the actual situation. In this embodiment, a third-order polynomial function is selected, and the deribbed surface profile curve of each segment of rusted ribbed steel bar is calculated using the common least squares method. Fit the trend line Figure 7 (a) and (b) correspond to the detrended surface and top view of the de-ribbed surface profile of the rusted ribbed steel bar, respectively. The smoothing process includes the conventional two-dimensional Gaussian kernel smoothing technique in the field of digital image processing. The r value of all data points with r>0 in the detrended surface D(r,θ,z) of the de-ribbed surface profile of the rusted steel bar is set to 0, and two-dimensional Gaussian kernel smoothing with a standard deviation of 2 is performed to obtain the pitting feature surface P(r,θ,z).

[0127] In step S6, the image segmentation algorithm includes the watershed segmentation method, specifically including: extending the pitting feature surface P(r,θ,z) along the θ axis and performing watershed segmentation processing to extract each rust pit, and calculating the geometric features of each rust pit, including area, length, width and depth.

[0128] Specifically, in this embodiment, the present invention uses the common watershed segmentation method to segment the pitting feature surface. Figure 8 (a) and (b) correspond to the extended surface contour of the corroded rebar and its watershed segmentation results, respectively. Corrosion pits that connect to the boundary of the watershed segmentation image or have the same shape are removed, resulting in all 13 corrosion pits on the rebar. Their corresponding geometric features are shown in Table 1.

[0129] Table 1 Geometric characteristics of rust pits

[0130]

[0131]

[0132] Example 2

[0133] A system for extracting pitting corrosion features of ribbed steel bars, specifically comprising:

[0134] The data preprocessing module is used to acquire point cloud data of rusted ribbed steel bars and perform preprocessing to obtain the three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bars.

[0135] The de-envelope processing module is used to perform de-envelope processing on the three-dimensional curved surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar, and separate the lower envelope curved surface B(r,θ,z) and the de-envelope curved surface C(r,θ,z) of the rusted ribbed steel bar.

[0136] The power spectral density estimation module is used to estimate the power spectral density of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar, and obtain a two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar.

[0137] The filtering module is used to design a band-stop filter based on the two-dimensional power spectral density map and filter out the residual transverse rib signal of the lower envelope surface C(r,θ,z) of the rusted ribbed steel bar, thereby obtaining the filtered lower envelope surface C of the rusted ribbed steel bar. bs (r,θ,z), combined with the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar, the deribbed surface profile surface A of the corroded ribbed steel bar is calculated. bs (r,θ,z);

[0138] The detrending processing module is used to process the deribbed surface profile A of rusted ribbed steel bars. bs (r,θ,z) is detrended to obtain the pitting corrosion characteristic surface P(r,θ,z) of the corroded ribbed steel bar;

[0139] The image segmentation module is used to extract the rust pits from the pitting feature surface P(r,θ,z) using an image segmentation algorithm to obtain the geometric features of the rust pits.

[0140] Specifically, this invention separates the lower envelope surface and the de-envelope surface of the rusted ribbed steel bar from the three-dimensional surface of the rusted ribbed steel bar surface profile using a de-envelope method. This effectively suppresses the mutual influence between the remaining inner diameter surface and the residual transverse rib signal surface. Furthermore, it accurately identifies the main repetition period and the frequency of its second harmonic of the residual transverse rib signal of the rusted ribbed steel bar through two-dimensional power spectral density estimation. A band-stop filter is then designed to filter out the residual transverse rib signal in the de-envelope surface of the rusted ribbed steel bar. Next, detrending processing is performed on the de-ribbed surface profile of the rusted ribbed steel bar to obtain the pitting corrosion feature surface. Finally, image segmentation processing is performed on the pitting corrosion feature surface to extract each corrosion pit. This invention solves the problem of irregular initial contours of ribbed steel bars, effectively identifies the corrosion patterns of ribbed steel bars, and accurately extracts pitting corrosion features, providing technical support for better performance evaluation of rusted reinforced concrete.

[0141] Example 3

[0142] A storage medium storing a program, which, when executed by a processor, implements the method for extracting pitting corrosion features of ribbed steel bars as described in Embodiment 1, as follows:

[0143] S1. Obtain the point cloud data of the rusted ribbed steel bar and perform preprocessing to obtain the three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar.

[0144] S2. The three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar is de-envelope processed to separate the lower envelope surface B(r,θ,z) and the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar.

[0145] S3. The power spectral density of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar is estimated to obtain a two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar.

[0146] S4. Design a band-stop filter based on the two-dimensional power spectral density diagram and filter out the residual transverse rib signal of the de-envelope surface C(r,θ,z) of the corroded ribbed steel bar to obtain the filtered de-envelope surface C of the corroded ribbed steel bar. bs (r,θ,z), combined with the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar, the deribbed surface profile surface A of the corroded ribbed steel bar is calculated. bs (r,θ,z);

[0147] S5. Deribbed surface profile A of rusted ribbed steel bars. bs (r,θ,z) is detrended to obtain the pitting corrosion characteristic surface P(r,θ,z) of the corroded ribbed steel bar;

[0148] S6. Extract the rust pits from the pitting feature surface P(r,θ,z) using an image segmentation algorithm to obtain the geometric features of the rust pits.

[0149] The specific processing steps described above are as in Example 1 and will not be repeated here.

[0150] In this embodiment, the storage medium may be a disk, optical disk, computer memory, read-only memory, random access memory, USB flash drive, portable hard drive, or other media.

[0151] Example 4

[0152] A computing device includes a processor and a memory for storing processor-executable programs. When the processor executes the program stored in the memory, it implements the method for extracting pitting corrosion features of ribbed steel bars as described in Embodiment 1, as follows:

[0153] S1. Obtain the point cloud data of the rusted ribbed steel bar and perform preprocessing to obtain the three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar.

[0154] S2. The three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar is de-envelope processed to separate the lower envelope surface B(r,θ,z) and the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar.

[0155] S3. The power spectral density of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar is estimated to obtain a two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar.

[0156] S4. Design a band-stop filter based on the two-dimensional power spectral density diagram and filter out the residual transverse rib signal of the de-envelope surface C(r,θ,z) of the corroded ribbed steel bar to obtain the filtered de-envelope surface C of the corroded ribbed steel bar. bs (r,θ,z), combined with the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar, the deribbed surface profile surface A of the corroded ribbed steel bar is calculated. bs (r,θ,z);

[0157] S5. Deribbed surface profile A of rusted ribbed steel bars. bs (r,θ,z) is detrended to obtain the pitting corrosion characteristic surface P(r,θ,z) of the corroded ribbed steel bar;

[0158] S6. Extract the rust pits from the pitting feature surface P(r,θ,z) using an image segmentation algorithm to obtain the geometric features of the rust pits.

[0159] The specific processing steps described above are as in Example 1 and will not be repeated here.

[0160] In this embodiment, the computing device can be a desktop computer, a laptop computer, a PDA handheld terminal, a tablet computer, or other terminal devices.

[0161] The above embodiments are preferred embodiments of the present invention and are not intended to limit the present invention. Any changes or other equivalent substitutions made without departing from the technical solution of the present invention are included within the protection scope of the present invention.

Claims

1. A method for extracting pitting corrosion features of ribbed steel bars, characterized in that, Including the following steps: S1. Obtain the point cloud data of the rusted ribbed steel bar and perform preprocessing to obtain the three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar. S2. The three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar is de-envelope processed to separate the lower envelope surface B(r,θ,z) and the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar. S3. The power spectral density of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar is estimated to obtain a two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar. S4. Design a band-stop filter based on the two-dimensional power spectral density map and filter out the residual transverse rib signal of the lower envelope surface C(r,θ,z) of the corroded ribbed steel bar to obtain the filtered lower envelope surface C of the corroded ribbed steel bar. bs (r,θ,z), combined with the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar, the deribbed surface profile surface A of the corroded ribbed steel bar is calculated. bs (r,θ,z); S5. Deribbed surface profile A of rusted ribbed steel bars. bs (r,θ,z) is detrended to obtain the pitting corrosion characteristic surface P(r,θ,z) of the rusted ribbed steel bar. Step S5 specifically includes: S51, the deribbed surface contour surface A of the rusted ribbed steel bar is... bs (r,θ,z) is the profile curve of the de-ribbed surface of the rusted ribbed steel bar, divided into M segments along the θ axis. Where i is the sequential number, i = 1, 2, 3…M, M ≥ 360; S52. Using a pre-defined trend model, the deribbed surface profile curve of each segment of corroded ribbed steel bar is calculated using the least squares method. Fit the trend line Where i is the sequential number, i=1,2,3…M, M≥360, and M segments of trend lines are defined. The ribbed surface profile trend surface U(r,θ,z) of the deribbed steel bar is obtained by arranging the ribbed steel bars in sequence. S53, the deribbed surface contour surface A of the rusted ribbed steel bar. bs Subtract the de-ribbed surface profile trend surface U(r,θ,z) of the rusted ribbed steel bar from (r,θ,z) to obtain the de-ribbed surface profile trend surface D(r,θ,z) of the rusted ribbed steel bar. S54. Select the data points with r greater than 0 in the detrending surface D(r,θ,z) of the de-stressed surface profile of the rusted ribbed steel bar, set the r value of the data point to 0, and smooth the detrending surface D(r,θ,z) to obtain the pitting feature surface P(r,θ,z). S6. Extract the rust pits from the pitting feature surface P(r,θ,z) using an image segmentation algorithm to obtain the geometric features of the rust pits.

2. The method for extracting pitting corrosion features of ribbed steel bars according to claim 1, characterized in that, Step S1 specifically includes: S11. Obtain the three-dimensional point cloud coordinate data of the surface morphology of the rusted ribbed steel bar through three-dimensional scanning technology, and generate a three-dimensional solid model of the rusted ribbed steel bar through three-dimensional modeling software. S12. Along the axial direction of the corroded ribbed steel bar, cut several cross sections of the corroded ribbed steel bar at predetermined lengths l; establish a three-dimensional rectangular coordinate system for the corroded ribbed steel bar with the length direction of the corroded ribbed steel bar as the z-axis and the cross section of the corroded ribbed steel bar as the (x,y) plane. Then, the three-dimensional rectangular coordinates of the data points of the cross section are represented as (x,y,z). Then, through coordinate transformation, obtain the three-dimensional polar coordinates (r,θ,z) of the data points of the cross section: , formula (1), , Official (2), Where r is the radius corresponding to the data point on the cross-section of the corroded steel bar, and θ is the angle corresponding to the data point on the cross-section of the corroded steel bar; S13. Form a three-dimensional surface A(r,θ,z) of the surface profile of the rusted ribbed steel bars in a three-dimensional polar coordinate system by using the data points of the cross-sections of all the rusted ribbed steel bars.

3. The method for extracting pitting corrosion features of ribbed steel bars according to claim 1, characterized in that, Step S2 specifically includes: S21. Differentiate the three-dimensional surface A(r,θ,z) of the rusted ribbed steel bar surface profile, and find all minimum points A(r,θ,z) of the three-dimensional surface A(r,θ,z) at data intervals d. low ,θ low ,z low The minimum point A(r) low ,θ low ,z low ) satisfies formulas (3) and (4): , Official (3), , Official (4), Where A(r0,θ0,z0) are the coordinate points in the three-dimensional surface A(r,θ,z) that satisfy formula (3), and the data interval d satisfies l R <d<2l R , l R The spacing between the transverse ribs of the uncorroded ribbed steel bars having the same diameter as the corroded ribbed steel bars; S22. Using all the minimum points as the endpoints of the lower envelope surface, the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar is obtained by cubic spline interpolation. S23. Based on the three-dimensional surface A(r,θ,z) of the rusted ribbed steel bar surface profile and the lower envelope surface B(r,θ,z) of the rusted ribbed steel bar, the lower envelope surface C(r,θ,z) is calculated as follows: C(r,θ,z) = A(r,θ,z)-B(r,θ,z), formula (5).

4. The method for extracting pitting corrosion features of ribbed steel bars according to claim 1, characterized in that, Step S3 specifically includes: S31. Divide the lower envelope surface C(r,θ,z) of the corroded ribbed steel bar into M segments along the θ axis to obtain the residual transverse rib signal curves of the corroded ribbed steel bar. Where i is the sequential number, i = 1, 2, 3…M, M ≥ 360; S32. The power spectral density of the residual transverse rib signal curves of each segment of rusted ribbed steel bar is estimated by the Welch power spectral estimation method to obtain the corresponding M segments of power spectral density curves. The M segments of power spectral density curves are numbered in sequence i=1,2,3…M and combined to obtain the two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar.

5. The method for extracting pitting corrosion features of ribbed steel bars according to claim 1, characterized in that, Step S4 specifically includes: S41. Based on the two-dimensional power spectral density diagram of the residual transverse rib signal of the corroded ribbed steel bar, the main repetition period T and the frequency ω2 of the second harmonic of the residual transverse rib signal are obtained, where T=1 / ω1, ω1 is the main frequency corresponding to the first frequency band along the frequency increasing direction in the two-dimensional power spectral density diagram, and ω2 is the frequency corresponding to the second frequency band along the frequency increasing direction. S42. Design a pair of second-order Butterworth bandstop filters centered on the main frequency ω1 and the second harmonic frequency ω2, respectively. The bandwidth of the second-order Butterworth bandstop filter is b. w Based on the two-dimensional power spectral density plot, the transfer function H of the second-order Butterworth bandstop filter is determined. j (s) is: Official (6), Where s is a complex frequency variable, ω j ζ represents the center frequency of the band-stop filter, and j=1 and j=2 represent the main frequency ω1 and the second harmonic frequency ω2 of the residual transverse rib signal of the corroded ribbed steel bar, respectively. j For the damping ratio, ζ j The calculation formula is: , formula (7); S43. The residual transverse rib signal of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar is filtered out by a pair of second-order Butterworth band-stop filters from step S42, to obtain the filtered de-envelope surface C of the rusted ribbed steel bar. bs (r,θ,z); S44. The lower envelope surface B(r,θ,z) of the rusted ribbed steel bar and the lower envelope surface C of the filtered rusted ribbed steel bar are... bs Adding (r, θ, z) yields the ribbed surface profile A of the deribbed ribbed steel bar. bs (r,θ,z): A bs (r, θ, z) = B(r, θ, z) + C bs (r, θ, z), Equation (8).

6. The method for extracting pitting corrosion features of ribbed steel bars according to claim 1, characterized in that, In step S6, the image segmentation algorithm includes the watershed segmentation method, specifically including: extending the pitting feature surface P(r,θ,z) along the θ axis and performing watershed segmentation processing to extract each rust pit, and calculating the geometric features of each rust pit, including area, length, width and depth.

7. A system for extracting pitting corrosion features of ribbed steel bars, applied to the method of claim 1, characterized in that, Specifically, it includes: The data preprocessing module is used to acquire point cloud data of rusted ribbed steel bars and perform preprocessing to obtain a three-dimensional surface A(r,θ,z) of the surface contour of the rusted ribbed steel bars. The de-envelope processing module is used to perform de-envelope processing on the three-dimensional curved surface A(r,θ,z) of the surface contour of the rusted ribbed steel bar, and separate the lower envelope curved surface B(r,θ,z) and the de-envelope curved surface C(r,θ,z) of the rusted ribbed steel bar. The power spectral density estimation module is used to estimate the power spectral density of the de-envelope surface C(r,θ,z) of the rusted ribbed steel bar, and obtain a two-dimensional power spectral density map of the residual transverse rib signal of the rusted ribbed steel bar. The filtering module is used to design a band-stop filter based on the two-dimensional power spectral density map and filter out the residual transverse rib signal of the lower envelope surface C(r,θ,z) of the rusted ribbed steel bar, thereby obtaining the filtered lower envelope surface C of the rusted ribbed steel bar. bs (r,θ,z), combined with the lower envelope surface B(r,θ,z) of the corroded ribbed steel bar, the deribbed surface profile surface A of the corroded ribbed steel bar is calculated. bs (r,θ,z); The detrending processing module is used to process the deribbed surface profile A of rusted ribbed steel bars. bs (r,θ,z) is detrended to obtain the pitting corrosion characteristic surface P(r,θ,z) of the rusted ribbed steel bar. The image segmentation module is used to extract the rust pits from the pitting feature surface P(r,θ,z) using an image segmentation algorithm to obtain the geometric features of the rust pits.

8. A storage medium, characterized in that, The system contains a program that, when executed by a processor, implements the method for extracting pitting corrosion features of ribbed steel bars as described in any one of claims 1-6.

9. A computing device, characterized in that, The method includes a processor and a memory for storing processor-executable programs. When the processor executes the program stored in the memory, it implements the method for extracting pit erosion features of ribbed steel bars as described in any one of claims 1-6.

Images