Crack detection device and detection method for prestressed board

Abstract

The invention provides a crack detection device and detection method for a prestressed board, and the method comprises the steps: extracting a crack trunk image of a to-be-detected prestressed board from a spectral image, carrying out the neighborhood division of the crack trunk image, and obtaining a plurality of crack connected domains; determining a crack characteristic index of the to-be-detected prestressed board, and determining an effective identification coefficient of a crack in the crack trunk image according to the crack characteristic index and all crack connected domains; all the crack connected domains are effectively identified based on the effective identification coefficient, and effective distribution of cracks of the to-be-detected prestressed board is obtained; extracting a spectral characteristic quantity from the spectral data, and carrying out characteristic combination on the spectral characteristic quantity and the effective distribution of the crack to obtain a combined crack characteristic value; and performing crack marking on the to-be-detected prestressed plate based on the joint crack characteristic value to complete crack detection of the to-be-detected prestressed plate. By adopting the scheme of the invention, the characteristic joint detection of the crack in the prestressed board can be realized, so that the crack misjudgment of the prestressed board is reduced.

Description

Technical Field

[0001] The present application relates to the technical field of prestressed panels, and more specifically, to a crack detection device and a crack detection method for a prestressed panel surface. Background Art

[0002] Prestressed slabs are a special type of concrete slab that is prestressed by steel strands during production, allowing it to withstand greater external loads. Compared with traditional concrete slabs, prestressed slabs have higher compressive and flexural strengths and can support greater loads. Prestressed slabs are commonly used in industrial and civil buildings, such as factories, airport runways, bridges, tunnels, etc.

[0003] During the manufacturing process of the prestressed plate, pressure needs to be applied to it in advance so that the internal force of the concrete in the tension zone of the prestressed plate when the external load acts on it will generate compressive stress to offset or reduce the tensile stress generated by the external load. Under the action of external forces and natural factors, cracks and other defects will inevitably appear on the surface of the prestressed plate, so it is necessary to detect cracks on the prestressed plate surface; however, in the prior art, when detecting cracks on the prestressed plate surface, traditional manual detection is inefficient, greatly affected by subjective factors, and cannot detect tiny cracks and hidden cracks. In addition, when using spectra to detect cracks on the prestressed plate surface, crack detection of the prestressed plate surface through a single spectral feature has limitations, and when detecting situations with complex prestressed plate surface textures or small crack widths, it is also difficult to accurately extract crack information from the background, resulting in incomplete crack detection results, resulting in misjudgment and missed detection of cracks on the prestressed plate surface. Therefore, how to achieve joint detection of cracks in the prestressed plate surface, thereby reducing misjudgment of cracks on the prestressed plate surface, has become a difficult problem faced by the industry. Summary of the invention

[0004] The present application provides a crack detection device and a detection method for a prestressed plate surface, which can realize the joint detection of the characteristics of cracks in the prestressed plate surface, thereby reducing the misjudgment of cracks in the prestressed plate surface.

[0005] In a first aspect, the present application provides a method for detecting cracks in a prestressed plate surface, comprising the following steps: Collecting spectral images of the prestressed plate surface to be tested; Extracting the crack trunk image of the prestressed plate surface to be detected from the spectral image based on the difference in spectral information, performing neighborhood division on the crack trunk image to obtain a plurality of crack connected domains; Determine the crack characteristic index of the prestressed plate surface to be detected, and determine the effective identification coefficient of the crack in the crack trunk image based on the crack characteristic index and all crack connected domains; Based on the effective identification coefficient, all crack connected domains are effectively identified to obtain the effective distribution of cracks on the prestressed plate surface to be detected; Acquire spectral data in the spectral image, and then extract spectral feature quantities from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected, and perform feature combination on the spectral feature quantities and the effective distribution of cracks to obtain a combined crack feature value when crack detection is performed on the prestressed plate surface to be detected; The surface of the prestressed plate to be detected is crack-marked based on the combined crack characteristic value, thereby completing crack detection of the prestressed plate to be detected.

[0006] In some embodiments, determining the crack characteristic index of the prestressed plate surface to be detected specifically includes: Obtain historical crack characteristic data of prestressed slabs with historical defects; The crack characteristic index of the prestressed plate surface to be detected is determined based on the historical crack characteristic data.

[0007] In some embodiments, determining the effective identification coefficient of the cracks in the crack trunk image based on the crack characteristic index and all crack connected domains specifically includes: Determine the fracture characteristic quantity of each fracture connection domain; The effective identification coefficient of the cracks in the crack trunk image is determined by all crack feature quantities and the crack feature index.

[0008] In some embodiments, effectively identifying all crack connected domains based on the effective identification coefficient to obtain the effective distribution of cracks on the prestressed plate surface to be detected specifically includes: Select a fracture connection domain as the selected fracture connection domain; Determine the fracture description value of the selected fracture connectivity domain; The effective identification coefficient and the fracture description value are used to perform identification response to obtain an identification result of the selected fracture connectivity domain; Continue to determine the identification results of the remaining fracture connection domains; The effective distribution of cracks on the prestressed plate surface to be detected is determined based on all identification results.

[0009] In some embodiments, extracting spectral feature quantities from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected specifically includes: Determine the spectral reflection characteristic curve of the prestressed plate surface to be detected from the spectral data, wherein the spectral reflection characteristic curve is the spectral reflection characteristic of the prestressed plate surface to be detected; The spectral characteristic quantity of the prestressed plate surface to be detected is extracted through the spectral reflection characteristic curve.

[0010] In some embodiments, marking cracks on the prestressed plate surface to be detected based on the combined crack characteristic value specifically includes: Acquire a spectral image of the prestressed plate surface to be tested; Traversing each pixel point in the spectral image; All pixel points whose pixel values ​​are greater than the combined crack characteristic value are marked at their corresponding positions on the prestressed plate surface to be detected, so as to obtain the crack positions of the prestressed plate surface to be detected and complete the crack detection of the prestressed plate to be detected.

[0011] In some embodiments, a spectral image of the prestressed panel surface to be inspected is collected by a hyperspectral imaging system.

[0012] In a second aspect, the present application provides a crack detection device for a prestressed plate surface, comprising: An acquisition module, used for acquiring a spectral image of the prestressed plate surface to be detected; A processing module, used for extracting the crack trunk image of the prestressed plate surface to be detected from the spectral image based on the spectral information difference, performing neighborhood division on the crack trunk image, and obtaining a plurality of crack connected domains; The processing module is further used to effectively identify all crack connected domains based on the effective identification coefficient to obtain the effective distribution of cracks on the prestressed plate surface to be detected; The processing module is further used to determine the crack characteristic index of the prestressed plate surface to be detected, and determine the effective identification coefficient of the crack in the crack trunk image based on the crack characteristic index and all crack connected domains; The processing module is further used to obtain spectral data in the spectral image, and then extract spectral feature quantities from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected, and perform feature combination of the spectral feature quantities and the effective distribution of cracks to obtain a combined crack feature value when crack detection is performed on the prestressed plate surface to be detected; An execution module is used to mark cracks on the surface of the prestressed plate to be detected based on the combined crack characteristic value, so as to complete crack detection on the prestressed plate to be detected.

[0013] In a third aspect, the present application provides a computer device, comprising a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the above-mentioned prestressed plate surface crack detection method.

[0014] In a fourth aspect, the present application provides a computer-readable storage medium, in which instructions or codes are stored. When the instructions or codes are run on a computer, the computer implements the above-mentioned prestressed plate surface crack detection method when executed.

[0015] The technical solution provided by the embodiments disclosed in this application has the following beneficial effects: In the present application, a spectral image of a prestressed plate surface to be detected is collected; a crack trunk image of the prestressed plate surface to be detected is extracted from the spectral image based on spectral information differences, and the crack trunk image is divided into neighborhoods to obtain a plurality of crack connected domains; crack characteristic indicators of the prestressed plate surface to be detected are determined, and an effective identification coefficient of cracks in the crack trunk image is determined by the crack characteristic indicators and all crack connected domains; all crack connected domains are effectively identified based on the effective identification coefficient to obtain an effective distribution of cracks on the prestressed plate surface to be detected; spectral data in the spectral image is acquired, and spectral characteristic quantities are extracted from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected, and the spectral characteristic quantities and the crack effective distribution are feature-combined to obtain a combined crack characteristic value when crack detection is performed on the prestressed plate surface to be detected; cracks on the prestressed plate surface to be detected are marked based on the combined crack characteristic value to complete crack detection of the prestressed plate to be detected.

[0016] It can be seen that in the present application, the spectral feature quantity and the effective distribution of the cracks are combined to obtain the combined crack feature value when the prestressed plate surface to be detected is subjected to crack detection, and then the prestressed plate surface to be detected is crack marked based on the combined crack feature value; firstly, the coefficient used to identify the significance of the cracks in the crack trunk image is determined by the crack feature index and the crack scale of the crack connected domain, that is, the effective identification coefficient of the cracks in the crack trunk image, which can more accurately evaluate the severity of the cracks in the prestressed plate surface to be detected and reduce false detection and missed detection of cracks; secondly, all crack connected domains are effectively identified according to the effective identification coefficient and the crack degree of the crack connected domain to obtain the effective distribution of cracks in the prestressed plate surface to be detected, which can present the significance and severity of the cracks in the prestressed plate surface to be detected in the form of image marking for subsequent analysis and processing; furthermore, based on the spectral reflectance characteristics of the prestressed plate surface to be detected (i.e., the reflectivity change of the prestressed plate surface to be detected at different wavelengths) extracting the spectral feature of the prestressed plate surface to be detected can help identify subtle changes in cracks in the prestressed plate surface to be detected, and then, the spectral feature is combined with the distribution density of cracks in the prestressed plate surface to be detected to obtain a joint crack feature value when crack detection is performed on the prestressed plate surface to be detected. By combining the crack feature and the spectral feature, the misjudgment caused by relying solely on a single feature for crack detection can be avoided, thereby reducing the probability of false detection and missed detection of cracks. Finally, the crack detection of the prestressed plate surface to be detected is completed based on the joint crack feature value, which can more accurately detect and mark the crack position in the prestressed plate surface to be detected, reduce manual intervention, and save manpower and time costs, and can also improve the efficiency and consistency of prestressed plate surface crack detection. In summary, this scheme can realize the feature joint detection of cracks in the prestressed plate surface, thereby reducing false judgment of cracks in the prestressed plate surface. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0018] Figure 1 is an exemplary flow chart of a crack detection method for a prestressed slab according to some embodiments of the present application; Figure 2 is an exemplary flow chart of determining a crack trunk image according to some embodiments of the present application; Figure 3 is an exemplary flow chart of determining fracture connectivity domains according to some embodiments of the present application; Figure 4 is a schematic diagram of exemplary hardware and / or software of a crack detection device for a prestressed slab according to some embodiments of the present application; Figure 5 It is a structural schematic diagram of a computer device for implementing a method for detecting cracks in a prestressed panel surface according to some embodiments of the present application. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0020] The embodiment of the present application provides a crack detection device and a detection method for a prestressed plate surface, the core of which is to collect a spectral image of the prestressed plate surface to be detected; extract a crack trunk image of the prestressed plate surface to be detected from the spectral image based on spectral information differences, perform neighborhood division on the crack trunk image, and obtain multiple crack connected domains; determine crack characteristic indicators of the prestressed plate surface to be detected, and determine an effective identification coefficient of cracks in the crack trunk image based on the crack characteristic indicators and all crack connected domains; perform effective identification on all crack connected domains based on the effective identification coefficient to obtain the crack trunk image to be detected. The invention discloses a method for detecting cracks in a prestressed plate surface by using a method for detecting cracks in a prestressed plate surface. The method comprises the following steps: obtaining spectral data in the spectral image, and then extracting spectral feature quantities from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected, and performing feature combination of the spectral feature quantities and the crack effective distribution to obtain a joint crack feature value when performing crack detection on the prestressed plate surface to be detected; marking cracks on the prestressed plate surface to be detected based on the joint crack feature value to complete crack detection on the prestressed plate surface to be detected; the feature joint detection of cracks in the prestressed plate surface can be realized, thereby reducing misjudgment of cracks in the prestressed plate surface.

[0021] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods. Figure 1 , which is an exemplary flow chart of a method for detecting cracks in a prestressed plate surface according to some embodiments of the present application. The method 100 for detecting cracks in a prestressed plate surface mainly includes the following steps: In step 101, a spectral image of the prestressed plate surface to be inspected is collected.

[0022] In specific implementation, the spectral image of the prestressed plate surface to be inspected can be collected by a hyperspectral imaging system, that is, the surface of the prestressed plate to be inspected is scanned using the hyperspectral imaging system to obtain the spectral image of the prestressed plate surface to be inspected. In other embodiments, other methods can also be used for collection, which is not limited here.

[0023] It should be noted that the spectral image in the present application represents a hyperspectral data cube that contains spatial information and spectral information of the detected prestressed panel surface. The spectral image contains spectral reflectance data of each pixel (i.e., the reflectance value of each pixel). In addition, the spectral image contains multiple bands, each band represents reflection or emission information within a different wavelength range in the spectral image. Therefore, the spectral image can provide spectral information of the prestressed panel surface to be detected in multiple bands, which can reflect slight changes in the material so as to identify defects such as cracks, and the spectral image provides a data basis for subsequent crack detection.

[0024] In step 102, a crack trunk image of the prestressed plate surface to be detected is extracted from the spectral image based on the difference in spectral information, and the crack trunk image is divided into neighborhoods to obtain a plurality of crack connected domains.

[0025] In some embodiments, reference Figure 2 As shown in FIG. 1 , this figure is an exemplary flow chart for determining a crack trunk image in some embodiments of the present application. In this embodiment, extracting the crack trunk image of the prestressed plate surface to be detected from the spectral image based on the spectral information difference can be achieved by the following steps: First, in step 1021, different bands in the spectral image are extracted to obtain multiple band images; Next, in step 1022, pixel difference is performed on all band images to obtain a plurality of difference images, wherein all the difference images are used as spectral information differences; Finally, in step 1023, all the differential images are segmented by threshold value to obtain the crack trunk image of the prestressed plate surface to be detected.

[0026] In specific implementation, the band image in the present application represents image data within the corresponding band wavelength range. Different band images reflect the reflection characteristics of the prestressed plate surface to be detected at different wavelengths, which is helpful to identify and distinguish different materials and surface features in the prestressed plate to be detected; extracting different bands in the spectral image to obtain multiple band images can be achieved in the following way, namely: the band images corresponding to each band can be extracted from the spectral image through the library in Python for processing hyperspectral data in spectral images (such as hyperspy, spectral and numpy libraries), thereby obtaining multiple band images. In other embodiments, other methods can also be used for extraction, which is not limited here.

[0027] In a specific implementation, pixel difference differentiation is performed on all band images to obtain multiple differential images, which can be achieved in the following manner, namely: for every two band images in all band images, the pixel difference of each corresponding pixel point in the two band images (that is, the pixel difference between the two band images) can be calculated by the differential image method, and an image composed of all the calculated pixel differences is used as the differential image of the two band images, thereby obtaining a differential image between every two band images in all band images. In other embodiments, other methods can also be used for determination, which is not limited here.

[0028] It should be noted that the differential image in the present application refers to an image composed of all pixel differences between two band images.

[0029] In specific implementation, threshold segmentation is performed on all differential images to obtain the crack trunk image of the prestressed plate surface to be detected, which can be achieved in the following manner, namely: first, a segmentation threshold is set, and the segmentation threshold can be set based on historical experimental data and experience through an adaptive threshold method, so that the set segmentation threshold is suitable for differential images with uneven brightness or complex backgrounds, so that better results can be achieved when crack detection is performed through differential images; secondly, a differential image is selected as a selected differential image, and then, the pixel points in the selected differential image whose pixel difference is greater than the segmentation threshold are marked as crack areas, and then, the crack areas in the remaining differential images are continued to be marked; finally, all the marked crack areas are binarized to obtain a binary image that constitutes the crack trunk image of the prestressed plate surface to be detected, wherein white pixels in the crack trunk image represent crack areas, and black pixels represent background areas; other methods may also be used for determination in other embodiments, which are not limited here.

[0030] It should be noted that the crack trunk image in the present application represents a binary image extracted from the spectral image of the prestressed plate surface, representing the main structure and morphology of the crack. In addition, in the present application, the crack is separated from the background to obtain the crack trunk image, which reduces the interference of background noise on crack detection and highlights the main body of the crack to lay the foundation for subsequent crack analysis and processing.

[0031] In some embodiments, reference Figure 3 As shown in FIG. 1 , this figure is an exemplary flow chart of determining a fracture connection domain in some embodiments of the present application. In this embodiment, the fracture trunk image is divided into neighborhoods to obtain multiple fracture connection domains, which can be achieved by using the following steps: First, in step 1031, burrs are removed from the crack trunk image to obtain a crack cleansing image; Next, in step 1032, a neighborhood partition factor of the crack cleansing image is determined; Finally, in step 1033, a plurality of fracture connected domains are identified from the fracture trunk image using the neighborhood partitioning factor.

[0032] In a specific implementation, the crack trunk image is subjected to burr removal to obtain a crack purified image, which can be achieved in the following manner, namely: noise and fine burrs in the crack trunk image are removed by morphological operations (such as corrosion and expansion), the crack trunk structure is retained, and a crack purified image is obtained. In other embodiments, other methods may also be used for determination, which are not limited here. In addition, it should be noted that the crack purified image represents an image of the crack trunk image with noise and irrelevant fine structures removed, and the crack purified image is determined for further crack analysis and evaluation.

[0033] It should be noted that the neighborhood division factor in the present application is a factor used to divide the neighborhood pixel relationship in the crack cleansing image. The neighborhood division factor determines which adjacent pixels are considered to be connected in the process of marking the connected domain of the crack cleansing image. Therefore, the neighborhood division factor can better identify the crack connected area in the crack cleansing image. In specific implementation, the neighborhood division factor of the crack cleansing image can be determined in the following manner, that is, the neighborhood division factor of the crack cleansing image can be determined based on the characteristics of the connected area in the crack cleansing image (such as area, aspect ratio and shape, etc.) through a machine learning model. Since for finer cracks, the 8-neighborhood connectivity analysis method (i.e., considering the connectivity of the pixels in the crack cleansing image in four directions of up, down, left, right and diagonal) can usually better identify the connected area, the neighborhood division factor in the present application is an 8-neighborhood factor, wherein the machine learning model is, for example, a decision tree, a random forest, a support vector machine, etc., and other methods can also be used for determination in other embodiments, which is not limited here.

[0034] In specific implementation, identifying multiple fracture connected domains from the fracture trunk image by the neighborhood division factor can be achieved in the following manner, namely: the fracture cleansed image can be identified and marked by the neighborhood division factor using the existing 8-neighborhood connected domain marking method, and then the area of ​​each connected domain in the fracture cleansed image is extracted, wherein the fracture connected domain is composed of connected foreground pixels (white pixels), and background pixels (black pixels) separate each fracture connected domain, thereby obtaining multiple fracture connected domains. In other embodiments, other methods can also be used for determination, which is not limited here.

[0035] It should be noted that the crack connected domain in the present application refers to the connected area composed of adjacent pixels in the crack purification image and representing the crack. By identifying and dividing the crack connected domain from the crack trunk image, noise and isolated small areas can be effectively removed, and the main position and expansion trend of the crack can be determined, thereby improving the accuracy of subsequent prestressed plate surface crack detection.

[0036] In step 103, crack characteristic indices of the prestressed plate surface to be detected are determined, and effective identification coefficients of cracks in the crack trunk image are determined based on the crack characteristic indices and all crack connected domains.

[0037] In some embodiments, determining the crack characteristic index of the prestressed plate surface to be detected can be achieved by using the following steps: Obtain historical crack characteristic data of prestressed slabs with historical defects; The crack characteristic index of the prestressed plate surface to be detected is determined based on the historical crack characteristic data.

[0038] It should be noted that in the present application, the historical crack database of the prestressed slab includes crack characteristic data of all historical defective prestressed slabs, and the crack characteristic data include crack area, circumference, aspect ratio, density, shape characteristics (such as roundness, compactness, etc.) and other crack characteristic data on the defective prestressed slab.

[0039] In specific implementation, obtaining the historical crack characteristic data of the historical defective prestressed slab can be achieved in the following manner, namely: the crack characteristic data of the historical defective prestressed slab can be obtained from the historical crack database of the prestressed slab, and all the obtained crack characteristic data are combined to form the historical crack characteristic data of the historical defective prestressed slab. In other embodiments, other methods can also be used to obtain the data, which is not specifically limited here.

[0040] It should be noted that the crack characteristic index in the present application is a parameter used to detect the crack characteristics of the prestressed plate surface. The crack characteristic index is a comprehensive quantitative value. The crack characteristic index can quantify the size, shape and distribution of the cracks in the prestressed plate surface, and provide a specific description of the cracks in the prestressed plate surface. In specific implementation, the crack characteristic index of the prestressed plate surface to be detected is determined by the historical crack characteristic data. The following method can be used, namely: the historical crack characteristic data can be feature extracted and selected through the existing feature selection algorithm to extract the most useful crack characteristic data for the crack detection of the prestressed plate surface to be detected, and then the existing The comprehensive quantification algorithm will comprehensively quantify all the extracted crack feature data, and finally use the quantified results as the crack feature indicators of the prestressed plate surface to be detected. Through the above-mentioned comprehensive quantification, all the crack feature data can be quantified into a comprehensive value to provide a comprehensive evaluation of the cracks in the prestressed plate surface to be detected, and can also provide a scientific basis for the detection of cracks in the prestressed plate surface. Among them, the feature selection algorithm is, for example, recursive feature elimination (RFE), random forest feature selection, etc., and the comprehensive quantification algorithm is, for example, fuzzy comprehensive evaluation method, cluster analysis method, weighted average method, etc. In other embodiments, other methods can also be used for determination, which is not limited here.

[0041] In some embodiments, determining the effective identification coefficient of the cracks in the crack trunk image based on the crack characteristic index and all crack connected domains can be achieved by using the following steps: Determine the fracture characteristic quantity of each fracture connection domain; The effective identification coefficient of the cracks in the crack trunk image is determined by all crack feature quantities and the crack feature index.

[0042] It should be noted that the crack characteristic quantity in the present application reflects the crack scale of the crack connected domain. The larger the crack characteristic quantity is, the larger the crack scale of the crack connected domain is, indicating that the potential structural risk of the crack connected domain is also higher. On the contrary, the smaller the crack characteristic quantity is, the smaller the crack scale of the crack connected domain is, indicating that the potential structural risk of the crack connected domain is also lower. In specific implementation, the crack characteristic quantity of each crack connected domain can be determined in the following way, that is, the total number of pixels in each crack connected domain is used as the crack characteristic quantity of each crack connected domain. In other embodiments, other methods can also be used for determination, which is not limited here.

[0043] In specific implementation, determining the effective identification coefficient of cracks in the crack trunk image through all crack feature quantities and the crack feature index can be achieved in the following manner, namely: multiplying each crack feature quantity with the crack feature index respectively to obtain multiple product values, and then using the statistics of all product values ​​as the effective identification coefficient of cracks in the crack trunk image, wherein the statistics are represented by the average value, and in other embodiments, the median, standard deviation and other methods can also be used for determination, which is not limited here.

[0044] It should be noted that the effective identification coefficient in the present application is a coefficient used to identify the significance of cracks in the crack trunk image. The effective identification coefficient can be used to more accurately evaluate the severity of cracks in the prestressed slab to be detected, reduce false detection and missed detection of cracks, and thus improve the accuracy and automation of prestressed slab crack detection, so as to comprehensively evaluate the health status of the prestressed slab. It will not be repeated here.

[0045] In step 104, all crack connected domains are effectively identified based on the effective identification coefficient to obtain the effective distribution of cracks on the prestressed plate surface to be detected.

[0046] In some embodiments, effective identification of all crack connected domains based on the effective identification coefficient to obtain the effective distribution of cracks on the prestressed plate surface to be detected can be achieved by the following steps: Select a fracture connection domain as the selected fracture connection domain; Determine the fracture description value of the selected fracture connectivity domain; The effective identification coefficient and the fracture description value are used to perform identification response to obtain an identification result of the selected fracture connectivity domain; Continue to determine the identification results of the remaining fracture connection domains; The effective distribution of cracks on the prestressed plate surface to be detected is determined based on all identification results.

[0047] In specific implementation, the crack description value in the present application reflects the degree of cracks in the crack connection domain. The larger the crack description value, the larger the degree of cracks in the crack connection domain, and the smaller the crack description value, the smaller the degree of cracks in the crack connection domain. The crack description value of the selected crack connection domain can be determined in the following manner, namely: first, the crack characteristic quantity of the selected crack connection domain is obtained, and secondly, the ratio of the maximum length and the minimum width of the crack in the selected crack connection domain is determined to obtain the aspect ratio of the selected crack connection domain. The aspect ratio reflects the slenderness of the crack connection domain. Finally, the result obtained by multiplying the crack characteristic quantity and the aspect ratio is used as the crack description value of the selected crack connection domain. In other embodiments, other methods can also be used for determination, which is not limited here.

[0048] In specific implementation, the effective identification coefficient and the crack description value are identified and responded to obtain the identification result of the selected fracture connection domain, which can be achieved in the following manner, namely: if the fracture description value is greater than the effective identification coefficient, the identification response of the selected fracture connection domain is a significant fracture area, that is, the identification result of the selected fracture connection domain is a significant fracture area, indicating that the degree of fracture in the selected fracture connection domain is relatively serious; if the fracture description value is less than or equal to the effective identification coefficient, the identification response of the selected fracture connection domain is a non-significant fracture area, that is, the identification result of the selected fracture connection domain is a non-significant fracture area, indicating that the degree of fracture in the selected fracture connection domain is relatively mild, and the identification response of the effective identification coefficient and the fracture description value can be completed. In other embodiments, other methods can also be used for implementation, which is not limited here.

[0049] In specific implementation, determining the effective distribution of cracks in the prestressed plate surface to be detected based on all identification results can be achieved in the following manner, namely: the crack connected domain that is a significant crack area in all identification results can be used as the effective crack area of ​​the prestressed plate surface to be detected, and then all effective crack areas are marked in white, and other areas are marked in black, and finally the image obtained after marking is used as the effective distribution of cracks in the prestressed plate surface to be detected, wherein the area marked in white is the effective distribution area of ​​cracks in the prestressed plate surface to be detected. In other embodiments, other methods can also be used for determination, which is not limited here.

[0050] It should be noted that the effective crack distribution in the present application represents all significant crack areas in the prestressed slab to be detected. The effective crack distribution presents the significance and severity of the cracks in the prestressed slab to be detected in the form of image markings to facilitate subsequent analysis and processing, which will not be repeated here.

[0051] It should be noted that the effective identification in the present application is a process of comparing the coefficient used to identify the degree of significance of cracks in the crack trunk image with the degree of cracks in the crack connected domain, so as to identify all crack connected domains that are significant crack areas, and then regard all crack connected domains that are significant crack areas as effective crack areas, and finally form all effective crack areas into an effective distribution of cracks on the prestressed slab surface to be detected. In addition, the effective identification can reduce misjudgments in manual detection and improve the accuracy and reliability of prestressed slab surface crack detection results.

[0052] In step 105, spectral data in the spectral image is obtained, and then spectral feature quantities are extracted from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected, and the spectral feature quantities and the effective distribution of cracks are combined to obtain a combined crack feature value when crack detection is performed on the prestressed plate surface to be detected.

[0053] In specific implementation, the spectral data in the spectral image can be obtained in the following manner, namely: the reflectance value of each pixel can be obtained from the spectral image, and the reflectance value is used to describe the spectral characteristics of the corresponding pixel point, and then all the obtained reflectance values ​​are combined into the spectral data in the spectral image. In other embodiments, other methods can also be used for acquisition, which is not specifically limited here.

[0054] It should be noted that the spectral data in the present application represents a collection of reflectance values ​​of all pixels in the spectral image.

[0055] In some embodiments, the following steps may be used to extract spectral feature quantities from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected: Determine the spectral reflection characteristic curve of the prestressed plate surface to be detected from the spectral data, wherein the spectral reflection characteristic curve is the spectral reflection characteristic of the prestressed plate surface to be detected; The spectral characteristic quantity of the prestressed plate surface to be detected is extracted through the spectral reflection characteristic curve.

[0056] It should be noted that the spectral reflection characteristic curve in the present application reflects the reflectivity change of the prestressed plate surface to be detected at different wavelengths. In specific implementation, the spectral reflection characteristic curve of the prestressed plate surface to be detected can be determined by the spectral data in the following manner, namely: first, all the reflectivity values ​​corresponding to each band in the spectral image can be obtained from the spectral data, and secondly, the average value of all the reflectivity values ​​corresponding to each band is used as the spectral vector of each band, and a spectral vector represents a spectral feature. Then, all the obtained spectral vectors are plotted into a curve through data analysis software (such as MATLAB), wherein the horizontal axis of the curve is the wavelength, and the vertical axis is the spectral vector. Finally, the curve obtained is used as the spectral reflection characteristic curve of the prestressed plate surface to be detected. Other methods can also be used for determination in other embodiments, which are not limited here.

[0057] In specific implementation, the spectral feature quantity of the prestressed plate surface to be detected can be extracted through the spectral reflection characteristic curve in the following manner, namely: the spectral features related to the crack detection of the prestressed plate surface to be detected can be identified and extracted from the spectral reflection characteristic curve based on spectral reflection characteristics (such as the shape of the spectral reflection characteristic curve, the peak position, the bandwidth, the change of reflectivity, etc.) through a machine learning model (such as a support vector machine, a decision tree, a random forest, etc.), and then the spectral features obtained are reduced in dimension using methods such as principal component analysis (PCA) to obtain multiple principal components and corresponding principal component scores to reduce the computational complexity and extract the main features, wherein the principal component score refers to the projection value of each spectral feature on each principal component, that is, it represents the position of the spectral feature in the principal component dimensionality reduction space, and finally the average value of the multiple principal component scores is obtained as the spectral feature quantity of the prestressed plate surface to be detected. In other embodiments, other methods can also be used for extraction, which is not limited here.

[0058] It should be noted that the spectral feature quantities in the present application are used to represent the overall spectral characteristics of the prestressed panel surface to be tested. The spectral feature quantities can help identify anomalies or defects and identify subtle changes caused by cracks. In addition, in the crack detection of the prestressed panel surface to be tested, the spectral feature quantities can reveal the impact of cracks on the spectral characteristics, thereby helping to detect and evaluate cracks and improve the sensitivity of crack detection. It will not be repeated here.

[0059] In some embodiments, the spectral feature quantity and the crack effective distribution are combined to obtain the combined crack feature value when the prestressed plate surface to be detected is subjected to crack detection, which can be achieved by the following steps: Determining the crack characteristic density of the prestressed plate surface to be tested based on the effective crack distribution; The crack characteristic density and the spectral characteristic quantity are combined to obtain a combined crack characteristic value when crack detection is performed on the prestressed plate surface to be detected.

[0060] It should be noted that the crack characteristic density in the present application reflects the distribution density of cracks in the prestressed plate surface to be detected. The larger the crack characteristic density, the larger the distribution density of cracks in the prestressed plate surface to be detected. Conversely, the smaller the crack characteristic density, the smaller the distribution density of cracks in the prestressed plate surface to be detected. In addition, the crack characteristic density can evaluate the distribution of cracks on the entire prestressed plate surface to be detected, which will not be repeated here. In specific implementation, the crack characteristic density of the prestressed plate surface to be detected can be determined by the crack effective distribution. The following method is adopted, namely: first, all significant crack areas in the crack effective distribution are obtained, and secondly, the total number of pixels of the prestressed plate surface to be detected is obtained. The ratio of the total number of pixels contained in all significant crack areas to the total number of pixels of the prestressed plate surface to be detected can be used as the crack characteristic density of the prestressed plate surface to be detected. In other embodiments, it can also be determined by other methods, which are not limited here.

[0061] In specific implementation, the crack characteristic density and the spectral characteristic quantity are combined to obtain the combined crack characteristic value when the prestressed plate surface to be detected is subjected to crack detection. This can be achieved in the following manner, namely: first, the influence of the crack characteristic density on the prestressed plate surface to be detected and the influence of the spectral characteristic quantity on the prestressed plate surface to be detected are determined based on historical experimental data through a weight ratio algorithm in the prior art, and a density characteristic weight coefficient corresponding to the crack characteristic density and a spectral characteristic weight coefficient corresponding to the spectral characteristic quantity are obtained, wherein the density characteristic weight coefficient is used to balance the crack characteristic density. The contribution of the spectral feature to the feature fusion is calculated, and the spectral feature weight coefficient is used to balance the contribution of the spectral feature to the feature fusion. Secondly, the crack feature density is multiplied by the density feature weight coefficient to obtain a first product value, and then the spectral feature is multiplied by the spectral feature weight coefficient to obtain a second product value. Finally, the two product values ​​obtained above are added together to obtain a result as the joint crack feature value for crack detection on the prestressed plate surface to be detected, thereby completing the feature combination of the spectral feature and the effective distribution of the crack. Other methods may also be used for determination in other embodiments, which are not limited here.

[0062] It should be noted that the combined crack feature in the present application is a combined characteristic index used to detect the specific position of the crack in the prestressed plate surface to be detected. The combined crack feature is a comprehensive characteristic parameter obtained by combining the spectral characteristic quantity and the crack characteristic density. By determining the combined crack feature, the crack position in the prestressed plate surface can be more accurately detected and marked, avoiding misjudgment caused by relying solely on a single crack detection feature. It will not be repeated here.

[0063] In step 106, crack marking is performed on the surface of the prestressed plate to be detected based on the combined crack characteristic value, thereby completing crack detection of the prestressed plate to be detected.

[0064] In some embodiments, crack marking of the prestressed plate surface to be detected based on the combined crack characteristic value may be implemented by the following steps: Acquire a spectral image of the prestressed plate surface to be tested; Traversing each pixel point in the spectral image; All pixel points whose pixel values ​​are greater than the combined crack characteristic value are marked at their corresponding positions on the prestressed plate surface to be detected, so as to obtain the crack positions of the prestressed plate surface to be detected and complete the crack detection of the prestressed plate to be detected.

[0065] In a specific implementation, each pixel point in the spectral image can be traversed through machine learning and image processing technology, and all pixel points with pixel values ​​greater than the combined crack characteristic value are marked at their corresponding positions on the surface of the prestressed plate to be detected, while pixel points with pixel values ​​less than or equal to the combined crack characteristic value in the spectral image are not processed to present the specific positions of the cracks on the surface of the prestressed plate to be detected, thereby realizing automatic crack detection of the prestressed plate to be detected. Other methods can also be used to achieve this in other embodiments, which are not limited here.

[0066] It should be noted that the present application combines crack characteristics and spectral characteristics to mark the crack positions of the prestressed panel surface to be inspected, so that cracks in the prestressed panel surface can be detected more accurately, the probability of false detection and missed detection of cracks can be reduced, and manual intervention can be reduced, manpower and time costs can be saved, and the efficiency and consistency of crack detection on the prestressed panel surface can be improved, making crack detection on the prestressed panel surface more accurate and reliable. It will not be repeated here.

[0067] In addition, in another aspect of the present application, in some embodiments, the present application provides a crack detection device for a prestressed plate surface, referring to Figure 4 , which is a schematic diagram of exemplary hardware and / or software of a crack detection device for a prestressed plate surface according to some embodiments of the present application, the crack detection device 400 for a prestressed plate surface includes: an acquisition module 401, a processing module 402 and an execution module 403, which are described as follows: Acquisition module 401, in this application, acquisition module 401 is mainly used to collect spectral images of the prestressed plate surface to be detected; Processing module 402, in the present application, the processing module 402 is mainly used to extract the crack trunk image of the prestressed plate surface to be detected from the spectral image based on the spectral information difference, perform neighborhood division on the crack trunk image, and obtain multiple crack connected domains; The processing module 402 in the present application is also used to determine the crack characteristic index of the prestressed plate surface to be detected, and determine the effective identification coefficient of the crack in the crack trunk image based on the crack characteristic index and all crack connected domains; In specific implementation, the processing module 402 in the present application is also used to effectively identify all crack connected domains based on the effective identification coefficient to obtain the effective distribution of cracks on the prestressed plate surface to be detected; In addition, the processing module 402 in the present application is also used to obtain spectral data in the spectral image, and then extract spectral feature quantities from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected, and perform feature combination of the spectral feature quantities and the effective distribution of the cracks to obtain a combined crack feature value when crack detection is performed on the prestressed plate surface to be detected; Execution module 403, in the present application, execution module 403 is mainly used to mark cracks on the prestressed plate surface to be detected based on the combined crack characteristic value, and complete crack detection on the prestressed plate to be detected.

[0068] The above describes in detail the examples of the crack detection device and detection method for the prestressed panel provided by the embodiments of the present application. It can be understood that in order to realize the above functions, the corresponding device includes hardware structures and / or software modules corresponding to the execution of each function. It should be easily appreciated by those skilled in the art that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present application.

[0069] In some embodiments, the present application also provides a computer device, comprising a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the above-mentioned prestressed plate surface crack detection method.

[0070] In some embodiments, reference Figure 5 , the dotted line in the figure indicates that the unit or the module is optional, and the figure is a schematic diagram of the structure of a computer device for implementing the crack detection method of the prestressed plate surface of the present application. The crack detection method of the prestressed plate surface in the above embodiment can be Figure 5The computer device 500 shown in the figure is implemented, and the computer device 500 includes at least one processor 501, a memory 502 and at least one communication unit 505. The computer device 500 can be a terminal device, a server or a chip.

[0071] The processor 501 may be a general-purpose processor or a special-purpose processor. For example, the processor 501 may be a central processing unit (CPU), which may be used to control the computer device 500, execute software programs, and process data of the software programs. The computer device 500 may also include a communication unit 505 to implement signal input (reception) and output (transmission).

[0072] For example, the computer device 500 may be a chip, the communication unit 505 may be an input and / or output circuit of the chip, or the communication unit 505 may be a communication interface of the chip, and the chip may be a component of a terminal device, a network device, or other devices.

[0073] For another example, the computer device 500 may be a terminal device or a server, and the communication unit 505 may be a transceiver of the terminal device or the server, or the communication unit 505 may be a transceiver circuit of the terminal device or the server.

[0074] The computer device 500 may include one or more memories 502, on which a program 504 is stored. The program 504 can be executed by the processor 501 to generate instructions 503, so that the processor 501 performs the method described in the above method embodiment according to the instructions 503. Optionally, data (such as a target audit model) can also be stored in the memory 502. Optionally, the processor 501 can also read the data stored in the memory 502, and the data can be stored at the same storage address as the program 504, or the data can be stored at a different storage address from the program 504.

[0075] The processor 501 and the memory 502 may be provided separately or integrated together, for example, integrated on a system on chip (SOC) of the terminal device.

[0076] It should be understood that each step of the above method embodiment can be completed by a hardware-based logic circuit or software-based instructions in the processor 501. The processor 501 can be a CPU, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, such as discrete gates, transistor logic devices, or discrete hardware components.

[0077] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0078] For example, in some embodiments, the present application also provides a computer-readable storage medium, in which instructions or codes are stored. When the instructions or codes are run on a computer, the computer implements the above-mentioned prestressed plate surface crack detection method when executed.

[0079] In summary, in the crack detection device and detection method for the prestressed plate surface disclosed in the embodiment of the present application, a spectral image of the prestressed plate surface to be detected is collected; based on the difference in spectral information, a crack trunk image of the prestressed plate surface to be detected is extracted from the spectral image, and the crack trunk image is divided into neighborhoods to obtain a plurality of crack connected domains; the crack characteristic index of the prestressed plate surface to be detected is determined, and the effective identification coefficient of the crack in the crack trunk image is determined by the crack characteristic index and all crack connected domains; based on the effective identification coefficient, all crack connected domains are effectively identified to obtain the crack trunk image to be detected. The invention discloses a method for detecting cracks in a prestressed plate surface by using a method for detecting cracks in a prestressed plate surface. The method comprises the following steps: obtaining spectral data in the spectral image, and then extracting spectral feature quantities from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected, and performing feature combination of the spectral feature quantities and the crack effective distribution to obtain a joint crack feature value when performing crack detection on the prestressed plate surface to be detected; marking cracks on the prestressed plate surface to be detected based on the joint crack feature value to complete crack detection on the prestressed plate surface to be detected; the feature joint detection of cracks in the prestressed plate surface can be realized, thereby reducing misjudgment of cracks in the prestressed plate surface.

[0080] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0081] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A method for detecting cracks in a prestressed plate surface, characterized in that: The steps include: Collecting spectral images of the prestressed plate surface to be tested; Extracting the crack trunk image of the prestressed plate surface to be detected from the spectral image based on the difference in spectral information, performing neighborhood division on the crack trunk image to obtain a plurality of crack connected domains; Determine the crack characteristic index of the prestressed plate surface to be detected, and determine the effective identification coefficient of the crack in the crack trunk image based on the crack characteristic index and all crack connected domains; Based on the effective identification coefficient, all crack connected domains are effectively identified to obtain the effective distribution of cracks on the prestressed plate surface to be detected; Acquire spectral data in the spectral image, and then extract spectral feature quantities from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected, and perform feature combination on the spectral feature quantities and the effective distribution of cracks to obtain a combined crack feature value when crack detection is performed on the prestressed plate surface to be detected; The surface of the prestressed plate to be detected is crack-marked based on the combined crack characteristic value, thereby completing crack detection of the prestressed plate to be detected.

2. The method according to claim 1, characterized in that Determining the crack characteristic index of the prestressed plate surface to be detected specifically includes: Obtain historical crack characteristic data of prestressed slabs with historical defects; The crack characteristic index of the prestressed plate surface to be detected is determined based on the historical crack characteristic data.

3. The method according to claim 1, characterized in that Determining the effective identification coefficient of the cracks in the crack trunk image based on the crack characteristic index and all crack connected domains specifically includes: Determine the fracture characteristic quantity of each fracture connection domain; The effective identification coefficient of the cracks in the crack trunk image is determined by all crack feature quantities and the crack feature index.

4. The method according to claim 1, characterized in that Effectively identifying all crack connected domains based on the effective identification coefficient to obtain the effective distribution of cracks on the prestressed plate surface to be detected specifically includes: Select a fracture connection domain as the selected fracture connection domain; Determine the fracture description value of the selected fracture connectivity domain; The effective identification coefficient and the fracture description value are used to perform identification response to obtain an identification result of the selected fracture connectivity domain; Continue to determine the identification results of the remaining fracture connection domains; The effective distribution of cracks on the prestressed plate surface to be detected is determined based on all identification results.

5. The method according to claim 1, characterized in that Extracting spectral feature quantities from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected specifically includes: Determine the spectral reflection characteristic curve of the prestressed plate surface to be detected from the spectral data, wherein the spectral reflection characteristic curve is the spectral reflection characteristic of the prestressed plate surface to be detected; The spectral characteristic quantity of the prestressed plate surface to be detected is extracted through the spectral reflection characteristic curve.

6. The method according to claim 1, characterized in that The crack marking of the prestressed plate surface to be detected based on the combined crack characteristic value specifically includes: Acquire a spectral image of the prestressed plate surface to be tested; Traversing each pixel point in the spectral image; All pixel points whose pixel values ​​are greater than the combined crack characteristic value are marked at their corresponding positions on the prestressed plate surface to be detected, so as to obtain the crack positions of the prestressed plate surface to be detected and complete the crack detection of the prestressed plate to be detected.

7. The method according to claim 1, characterized in that The spectral image of the prestressed plate surface to be inspected is collected through a hyperspectral imaging system.

8. A crack detection device for a prestressed plate surface, characterized in that: include: An acquisition module, used for acquiring a spectral image of the prestressed plate surface to be detected; A processing module, used for extracting the crack trunk image of the prestressed plate surface to be detected from the spectral image based on the spectral information difference, performing neighborhood division on the crack trunk image, and obtaining a plurality of crack connected domains; The processing module is further used to effectively identify all crack connected domains based on the effective identification coefficient to obtain the effective distribution of cracks on the prestressed plate surface to be detected; The processing module is further used to determine the crack characteristic index of the prestressed plate surface to be detected, and determine the effective identification coefficient of the crack in the crack trunk image based on the crack characteristic index and all crack connected domains; The processing module is further used to obtain spectral data in the spectral image, and then extract spectral feature quantities from the spectral data based on the spectral reflectance characteristics of the prestressed plate surface to be detected, and perform feature combination of the spectral feature quantities and the effective distribution of cracks to obtain a combined crack feature value when crack detection is performed on the prestressed plate surface to be detected; An execution module is used to mark cracks on the surface of the prestressed plate to be detected based on the combined crack characteristic value, so as to complete crack detection on the prestressed plate to be detected.

9. A computer device, characterized in that: The computer device includes a memory and a processor, the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the prestressed plate surface crack detection method described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions or codes, and when the instructions or codes are executed on a computer, the computer implements the method for detecting cracks in a prestressed slab as described in any one of claims 1 to 7.

Images