Bridge surface crack detection method based on unmanned aerial vehicle

Abstract

The invention relates to a bridge surface crack detection method based on an unmanned aerial vehicle, and relates to the field of bridge detection, and the method comprises the steps: controlling the unmanned aerial vehicle to carry out the image information collection and point cloud data collection of each lane of a to-be-detected bridge, carrying out the crack detection of the image information, and obtaining the crack characteristics of each lane, modeling each lane based on the point cloud data so as to obtain a three-dimensional model of the surface of each lane, determining the crack depth of each lane based on the three-dimensional model and the flatness of the lane surface of the area where each crack feature is located, and determining the crack anomaly of each lane based on the flatness, the crack depth and the crack feature. And determining the crack anomaly degree of the to-be-detected bridge based on the crack anomaly degree of each lane. According to the invention, the bridge deck crack can be detected more accurately and comprehensively, and the abnormal degree of the whole bridge can be obtained more accurately.

Description

Technical Field

[0001] The present application relates to the field of bridge detection, and in particular to a method for detecting cracks on a bridge surface based on an unmanned aerial vehicle. Background Art

[0002] Bridges are an important part of modern transportation, and bridge safety is the top priority of road traffic safety. As bridges age, the risk of damage or even collapse increases, so they need to be repaired.

[0003] At present, an important aspect of judging whether a bridge is abnormal is to judge the cracks on the bridge deck. However, the abnormality analysis is mainly based on some properties of the cracks themselves, such as length, which leads to a one-sided detection of bridge cracks and the inability to obtain a more accurate degree of abnormality of the cracks, and thus the degree of abnormality of the entire bridge cannot be obtained. Summary of the invention

[0004] In order to conduct more accurate and comprehensive detection of bridge deck cracks and thereby obtain a more accurate degree of abnormality of the entire bridge, the present application provides a bridge surface crack detection method based on a drone.

[0005] In the first aspect, the present application provides a method for detecting cracks on a bridge surface based on a drone, which adopts the following technical solution: A bridge surface crack detection method based on drone, comprising: Control the UAV to collect image information and point cloud data for each lane of the bridge to be tested; Performing crack detection on the image information to obtain crack features on each lane; Modeling each lane based on the point cloud data to obtain a three-dimensional model of the surface of each lane; Determining the crack depth of each lane and the flatness of the lane surface in the area where each crack feature is located based on the three-dimensional model; Determine the crack abnormality of each lane based on the flatness, crack depth and crack characteristics; The abnormal degree of cracks of the bridge to be tested is determined based on the abnormal degree of cracks of each lane.

[0006] By adopting the above technical solution, the drone is controlled to fly along each lane according to the multiple lanes divided on the bridge, and the image information and point cloud data of the lane surface are collected during the flight. The image information records the specific condition of the lane surface and whether there are cracks. Therefore, crack detection on the image information can obtain the crack characteristics on each lane. The point cloud data records the three-dimensional terrain of the lane surface. Therefore, a three-dimensional model of the surface of each lane is obtained based on the point cloud data. According to the three-dimensional model, the crack depth of the crack on each lane and the flatness of the lane surface in the crack area can be accurately determined. The worse the flatness, the worse and more uneven the lane condition, and the greater the risk of damage or abnormality to the overall structure of the bridge here, which indicates that the crack abnormality here is high. Therefore, the crack abnormality of each lane is comprehensively determined according to the flatness, crack depth and crack characteristics, that is, the flatness of the lane surface at the crack and the properties of the crack itself are comprehensively analyzed to obtain a more accurate crack abnormality. Then, the crack abnormality of the entire bridge to be tested can be determined in combination with the crack abnormality. Compared with the current analysis of only the specific situation of the crack itself, the present application combines the undulation and flatness at the crack to determine the abnormality of the entire bridge more accurately.

[0007] In another possible implementation, determining the flatness of the lane surface in the area where each crack feature is located based on the three-dimensional model includes: Determine from the three-dimensional model a model segment of the lane area corresponding to each crack feature; The model segment is segmented according to a first preset plane to obtain a first undulating curve on the surface of the lane area, wherein the first preset plane is a plane perpendicular to the bridge deck and consistent with the extension direction of the lane; The model segment is segmented multiple times according to a second preset plane to obtain multiple second undulating curves on the surface of the lane area, wherein the second preset plane is a plane perpendicular to the bridge deck and perpendicular to the extension direction of the lane; The flatness of the lane surface in the area where each crack feature is located is determined based on the first undulation curve and the second undulation curve.

[0008] In another possible implementation, determining the flatness of the lane surface in the area where each crack feature is located based on the first undulation curve and the second undulation curve includes: Calculating the similarity between the first undulating curve and a preset straight line to obtain a first similarity; Segmenting the first undulating curve according to the second preset plane to obtain a plurality of curve segments; Calculating the first curvature corresponding to each of the plurality of curve segments and the second curvature of the plurality of second undulating curves; Determine a undulation score of the lane corresponding to each curved segment based on the first curvature and the second curvature; The flatness of the lane surface in the area where each crack feature is located is determined based on the undulation score and the first similarity.

[0009] In another possible implementation, determining the flatness of the lane surface in the area where each crack feature is located based on the undulation score and the first similarity includes: Determine the target curve segment whose ups and downs score reaches a preset score threshold; Determining a first ratio of the number of the target curve segments to the number of all curve segments; Determining an average value of the ups and downs scores of the target curve segment; The flatness of the lane surface in the area where each crack feature is located is determined based on the first proportion of the target curve segment, the average value of the undulation score, the first similarity and the corresponding coefficients.

[0010] In another possible implementation, the determining of the crack abnormality of each lane based on the flatness, crack depth and crack characteristics includes: Determine the length and width of each crack feature; Determine the attribute score of each crack based on the length, width, crack depth and their corresponding coefficients; Determine the abnormality score of each crack based on the attribute score of each crack, the flatness corresponding to each crack and the corresponding coefficients; The abnormal scores of all crack features of each lane are summed to obtain the crack abnormality degree of each lane.

[0011] In another possible implementation, the controlling the drone to collect image information and point cloud data for each lane of the bridge to be tested includes: Obtain surveillance video of each lane within a preset historical time period; Performing feature recognition on the surveillance video to obtain vehicle features and the total number of vehicles passing through each lane within a preset historical time period; Classifying the vehicle features according to preset categories to obtain the number of vehicles corresponding to each preset category, wherein the preset categories include small vehicles and large vehicles; Determining the number of vehicles corresponding to the large vehicles, and calculating a second ratio of the number of vehicles to the total number of vehicles; Sorting the second proportion of each lane to obtain a sorting result, wherein the sorting result represents the order of importance of all lanes; The drone is controlled to collect image information and point cloud data for each lane of the bridge to be tested according to the sorting result.

[0012] In another possible implementation, the method further includes: Obtaining a historical crack detection result for each lane, wherein the historical crack detection result includes the number of historical cracks on each lane and the crack abnormality of each historical crack; Determining an average value of the fracture anomaly degrees of all historical fractures based on the fracture anomaly degrees of the historical fractures; Determining a correction score based on the average value of the crack abnormality, the number of historical cracks, and the position of each lane in the sorting result; The crack abnormality of each lane is corrected and calculated based on the correction score to obtain a corrected crack abnormality.

[0013] In the second aspect, the present application provides a bridge surface crack detection device based on a drone, which adopts the following technical solution: A bridge surface crack detection device based on an unmanned aerial vehicle, comprising: A control module, used to control the UAV to collect image information and point cloud data for each lane of the bridge to be tested; A crack detection module, used to perform crack detection on the image information to obtain crack features on each lane; A modeling module, used for modeling each lane based on the point cloud data, so as to obtain a three-dimensional model of the surface of each lane; A flatness determination module, used to determine the crack depth of each lane and the flatness of the lane surface in the area where each crack feature is located based on the three-dimensional model; A first abnormality determination module, configured to determine the crack abnormality of each lane based on the flatness, crack depth and crack characteristics; The second abnormality determination module is used to determine the abnormality degree of cracks of the bridge to be tested based on the abnormality degree of cracks of each lane.

[0014] By adopting the above technical solution, the control module controls the drone to fly along each lane according to the multiple lanes divided on the bridge, and collects image information and point cloud data of the lane surface during the flight. The image information records the specific condition of the lane surface and whether there are cracks. Therefore, the crack detection module performs crack detection on the image information to obtain the crack characteristics on each lane. The point cloud data records the three-dimensional terrain conditions of the lane surface. Therefore, the modeling module performs modeling based on the point cloud data to obtain a three-dimensional model of the surface of each lane. The flatness determination module can accurately determine the crack depth of the cracks on each lane and the flatness of the lane surface in the area where the cracks are located according to the three-dimensional model. The worse the flatness, the worse and more uneven the lane condition is, and the greater the risk of damage or abnormality to the overall structure of the bridge here, which means that the degree of abnormality of the cracks here is high. Therefore, the first abnormality determination module comprehensively determines the crack abnormality of each lane based on the flatness, crack depth and crack characteristics, that is, a more accurate crack abnormality is obtained by combining the flatness of the lane surface at the crack and the properties of the crack itself. Then, the second abnormality determination module combines the crack abnormality of each lane to determine the crack abnormality of the entire bridge to be tested. Compared with the current analysis of only the specific situation of the crack itself, the present application combines the undulation and flatness at the crack to determine the abnormality of the entire bridge more accurately.

[0015] In another possible implementation, when the flatness determination module determines the flatness of the lane surface in the area where each crack feature is located based on the three-dimensional model, it is specifically used to: Determine from the three-dimensional model a model segment of the lane area corresponding to each crack feature; The model segment is segmented according to a first preset plane to obtain a first undulating curve on the surface of the lane area, wherein the first preset plane is a plane perpendicular to the bridge deck and consistent with the extension direction of the lane; The model segment is segmented multiple times according to a second preset plane to obtain multiple second undulating curves on the surface of the lane area, wherein the second preset plane is a plane perpendicular to the bridge deck and perpendicular to the extension direction of the lane; The flatness of the lane surface in the area where each crack feature is located is determined based on the first undulation curve and the second undulation curve.

[0016] In another possible implementation, when the flatness determination module determines the flatness of the lane surface in the area where each crack feature is located based on the first undulation curve and the second undulation curve, it is specifically used to: Calculating the similarity between the first undulating curve and a preset straight line to obtain a first similarity; Segmenting the first undulating curve according to the second preset plane to obtain a plurality of curve segments; Calculating the first curvature corresponding to each of the plurality of curve segments and the second curvature of the plurality of second undulating curves; Determine a undulation score of the lane corresponding to each curved segment based on the first curvature and the second curvature; The flatness of the lane surface in the area where each crack feature is located is determined based on the undulation score and the first similarity.

[0017] In another possible implementation, when the flatness determination module determines the flatness of the lane surface in the area where each crack feature is located based on the undulation score and the first similarity, it is specifically configured to: Determine the target curve segment whose ups and downs score reaches a preset score threshold; Determining a first ratio of the number of the target curve segments to the number of all curve segments; Determining an average value of the ups and downs scores of the target curve segment; The flatness of the lane surface in the area where each crack feature is located is determined based on the first proportion of the target curve segment, the average value of the undulation score, the first similarity and the corresponding coefficients.

[0018] In another possible implementation, when the first abnormality determination module determines the crack abnormality of each lane based on the flatness, crack depth, and crack characteristics, it is specifically used to: Determine the length and width of each crack feature; Determine the attribute score of each crack based on the length, width, crack depth and their corresponding coefficients; Determine the abnormality score of each crack based on the attribute score of each crack, the flatness corresponding to each crack and the corresponding coefficients; The abnormal scores of all crack features of each lane are summed to obtain the crack abnormality degree of each lane.

[0019] In another possible implementation, when the control module controls the drone to collect image information and point cloud data for each lane of the bridge to be tested, it is specifically used to: Obtain surveillance video of each lane within a preset historical time period; Performing feature recognition on the surveillance video to obtain vehicle features and the total number of vehicles passing through each lane within a preset historical time period; Classifying the vehicle features according to preset categories to obtain the number of vehicles corresponding to each preset category, wherein the preset categories include small vehicles and large vehicles; Determining the number of vehicles corresponding to the large vehicles, and calculating a second ratio of the number of vehicles to the total number of vehicles; Sorting the second proportion of each lane to obtain a sorting result, wherein the sorting result represents the order of importance of all lanes; The drone is controlled to collect image information and point cloud data for each lane of the bridge to be tested according to the sorting result.

[0020] In another possible implementation, the device further includes: An acquisition module, used to acquire a historical crack detection result of each lane, wherein the historical crack detection result includes the number of historical cracks on each lane and the crack abnormality of each historical crack; An average value determination module, used to determine an average value of the fracture abnormality of all historical fractures based on the fracture abnormality of the historical fractures; A correction score determination module, used to determine a correction score based on the average value of the crack abnormality, the number of historical cracks, and the position of each lane in the sorting result; The correction module is used to perform correction calculation on the crack abnormality of each lane based on the correction score to obtain a corrected crack abnormality.

[0021] In a third aspect, the present application provides an electronic device, which adopts the following technical solution: An electronic device, comprising: at least one processor; Memory; At least one application, wherein at least one application is stored in a memory and configured to be executed by at least one processor, and at least one is configured to: execute a drone-based bridge surface crack detection method shown in any possible implementation of the first aspect.

[0022] In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium, when the computer program is executed in a computer, enables the computer to execute the bridge surface crack detection method based on an unmanned aerial vehicle as described in any one of the first aspects.

[0023] In summary, the present application includes at least one of the following beneficial technical effects: The drone is controlled to fly along each lane according to the multiple lanes divided on the bridge, and image information and point cloud data of the lane surface are collected during the flight. The image information records the specific condition of the lane surface and whether there are cracks. Therefore, crack detection on the image information can obtain the crack characteristics on each lane. The point cloud data records the three-dimensional terrain of the lane surface. Therefore, a three-dimensional model of the surface of each lane is obtained based on the point cloud data. According to the three-dimensional model, the crack depth of the cracks on each lane and the flatness of the lane surface in the crack area can be accurately determined. The worse the flatness, the worse and more uneven the lane condition, and the greater the risk of damage or abnormality to the overall structure of the bridge here, which indicates that the crack abnormality here is high. Therefore, the crack abnormality of each lane is comprehensively determined according to the flatness, crack depth and crack characteristics, that is, the flatness of the lane surface at the crack and the properties of the crack itself are comprehensively analyzed to obtain a more accurate crack abnormality. Then, the crack abnormality of the entire bridge to be tested can be determined in combination with the crack abnormality. Compared with the current analysis of only the specific situation of the crack itself, the present application combines the undulation and flatness at the crack to determine the abnormality of the entire bridge more accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a flow chart of a method for detecting cracks on a bridge surface based on a drone according to an embodiment of the present application.

[0025] Figure 2 It is a structural schematic diagram of a bridge surface crack detection device based on a drone according to an embodiment of the present application.

[0026] Figure 3 It is a structural schematic diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] The present application is further described in detail below in conjunction with the accompanying drawings.

[0028] After reading this specification, those skilled in the art may make non-creative modifications to this embodiment as needed, but such modifications are protected by patent law as long as they are within the scope of the claims of this application.

[0029] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are 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.

[0030] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally means that the associated objects before and after are in an "or" relationship.

[0031] The embodiments of the present application are further described in detail below in conjunction with the drawings in the specification.

[0032] The embodiment of the present application provides a method for detecting cracks on a bridge surface based on a drone, which is executed by an electronic device, which can be a server or a terminal device, wherein the server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The terminal device can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc., but is not limited thereto. The terminal device and the server can be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiment of the present application. Figure 1 As shown, the method includes step S101, step S102, step S103, step S104, step S105 and step S106, wherein: S101, controlling the UAV to collect image information and point cloud data for each lane of the bridge to be tested.

[0033] In the embodiment of the present application, the electronic device and the drone are connected by wireless communication, so as to control the drone to fly along the route of each lane. The drone is equipped with a camera device to collect image information of the lane surface, and is also equipped with a laser radar or a three-dimensional laser scanner, etc., to collect point cloud data of the lane surface. After collecting the image information and point cloud data, the drone sends them to the electronic device, so that the electronic device obtains the image information and point cloud data.

[0034] S102, performing crack detection on the image information to obtain crack features on each lane.

[0035] For the embodiment of the present application, since the image information records the specific conditions of the lane surface, it is possible to analyze whether there are cracks in the lane from the image information, and the electronic device inputs the image information into the trained network model for crack detection and identification, thereby identifying the crack features on the lane. Alternatively, the electronic device performs edge detection on the image information, first denoises the image information, and then performs grayscale transformation on the denoised image information to obtain a grayscale image, and the position where the grayscale value in the grayscale image has a step is the crack feature.

[0036] S103, modeling each lane based on the point cloud data, thereby obtaining a three-dimensional model of the surface of each lane.

[0037] For the embodiment of the present application, the point cloud data represents the three-dimensional features of the lane surface, so the electronic device performs modeling operations on relevant software based on the point cloud data to obtain a three-dimensional model of each lane surface.

[0038] S104: Determine the crack depth of each lane and the flatness of the lane surface in the area where each crack feature is located based on the three-dimensional model.

[0039] For the embodiment of the present application, the three-dimensional model records the spatial performance of each position on the lane. Therefore, after the electronic device identifies the crack feature through image information, it determines the location of the crack feature from the three-dimensional model according to the location of the crack feature, thereby determining the crack depth of each crack feature in the three-dimensional model. Specifically, after determining the location of the crack feature in the three-dimensional model, the electronic device can determine the depth of each crack feature, and then average the depths at each location to obtain the crack depth of each crack. The three-dimensional model also records the ups and downs of the lane where the crack is located, that is, as the bridge increases in service life, the rolling wear of vehicles, and the impact of car accidents, the surface flatness of the bridge decreases, thereby affecting the quality of the bridge and increasing the risk of damage and abnormality of the bridge. Therefore, the electronic device determines the flatness of the lane surface where the crack is located according to the three-dimensional model.

[0040] S105: Determine the crack abnormality of each lane based on the flatness, crack depth, and crack characteristics.

[0041] For the embodiment of the present application, after the electronic device determines the flatness of the area where the crack feature is located, it combines the crack depth and the crack feature itself to conduct a comprehensive analysis of multiple factors to determine a more accurate crack abnormality for each lane. The worse the flatness of the lane surface at the crack, the higher the degree of crack abnormality, and the greater the risk of damage or abnormality to the bridge. Therefore, the determination of crack abnormality by combining flatness, crack depth and crack feature itself is more accurate than considering only the characteristic attributes of the crack itself.

[0042] S106: Determine the abnormality degree of cracks of the bridge to be tested based on the abnormality degree of cracks of each lane.

[0043] For the embodiment of the present application, after the electronic device determines the abnormality of the cracks in each lane, the abnormality of the cracks in all lanes is summed to obtain the abnormality of the cracks in the entire bridge to be tested. By controlling the drone to collect images of the lane surface and point cloud data of the lane surface, the abnormality of the cracks is more comprehensive than simply using image information to identify cracks and analyzing the properties of the cracks themselves. Further, the abnormality of the cracks in the bridge to be tested is determined more accurately and comprehensively based on the flatness of the lane where the cracks are located, the crack depth and the properties of the cracks themselves.

[0044] In a possible implementation of the embodiment of the present application, in step S104, the flatness of the lane surface in the area where each crack feature is located is determined based on the three-dimensional model, including step S1041 (not shown in the figure), step S1042 (not shown in the figure), step S1043 (not shown in the figure), and step S1044 (not shown in the figure), wherein: S1041, determining a model segment of the lane area corresponding to each crack feature from the three-dimensional model.

[0045] For the embodiment of the present application, after the electronic device determines the position of each crack feature in the three-dimensional model and image information, it can map the image information and the three-dimensional model to overlap according to the range of the lane, so as to obtain the position of the crack feature in the three-dimensional model, and then divide the three-dimensional model according to the preset range to obtain the model fragment of the lane area corresponding to each crack feature. The preset range can be a specified number of pixels, that is, after the electronic device determines the crack feature in the three-dimensional model, it expands the specified number of pixels around the outline of the crack feature to obtain the model fragment of each crack feature.

[0046] S1042: Segment the model segment according to a first preset plane to obtain a first undulating curve on the surface of the lane area.

[0047] The first preset plane is a plane perpendicular to the bridge deck and consistent with the extension direction of the lane.

[0048] For the embodiment of the present application, the first undulating curve obtained by the electronic device according to the segmentation of the first preset plane is equivalent to profile the lane to obtain a side view of the model. The first undulating curve of the side view records the undulation of the crack characteristics along the extension direction of the lane.

[0049] S1043: Segment the model segment multiple times according to the second preset plane to obtain multiple second undulating curves on the surface of the lane area.

[0050] The second preset plane is a plane perpendicular to the bridge deck and perpendicular to the extension direction of the lane.

[0051] For the embodiment of the present application, in order to determine the undulation in another direction of the lane area where the crack feature is located, since the crack is split along the lane extension direction, the electronic device divides the model segment of the crack feature multiple times according to the second preset plane to obtain multiple second undulation curves, which can be regarded as cutting the model segment corresponding to the crack feature multiple times. Each second undulation curve is the undulation of the lane surface perpendicular to the lane extension direction at each cut. Since the crack feature extends within a certain range, the undulation at each cut is different. It is convenient to obtain more accurate flatness based on the second undulation curves at multiple cuts. The length corresponding to the second undulation curve is the width of the model segment in step S1041, that is, the span in the model segment along the direction perpendicular to the lane extension direction.

[0052] S1044: Determine the flatness of the lane surface in the area where each crack feature is located based on the first undulation curve and the second undulation curve.

[0053] For the embodiment of the present application, the first undulating curve is the undulation in the direction in which the lane extends, and the second undulating curve is the undulation perpendicular to the direction in which the lane extends. Therefore, a more accurate flatness can be obtained by combining the undulations in the two directions.

[0054] In other embodiments, if the crack feature is perpendicular to the lane extension direction, or the angle between the crack feature and the lane extension direction is greater than 45°, a plurality of first undulating curves and first and second undulating curves are determined to determine the flatness.

[0055] In a possible implementation of the embodiment of the present application, in step S1044, the flatness of the lane surface in the area where each crack feature is located is determined based on the first undulation curve and the second undulation curve, specifically including step Sa (not shown in the figure), step Sb (not shown in the figure), step Sc (not shown in the figure), step Sd (not shown in the figure) and step Se (not shown in the figure), wherein Sa, calculating the similarity between the first undulating curve and the preset straight line to obtain a first similarity.

[0056] For the embodiment of the present application, the preset straight line represents that the bridge surface is completely horizontal, that is, an ideal state. Therefore, the electronic device can input the first undulating curve and the preset straight line into the trained network model to perform similarity calculation, thereby obtaining the first similarity. Specifically, the network model can be a convolutional neural network model, a recurrent neural network model, or other types of network models. In other embodiments, the first similarity can also be calculated by calculating the Euclidean distance, cosine similarity, etc. between the first undulating curve and the preset straight line. The higher the first similarity, the closer the first undulating curve is to the preset straight line, the higher the flatness of the first undulating curve, the lower the abnormality of the crack feature in the lane extension direction, and the lower the probability and risk of bridge damage or abnormality.

[0057] Sb, dividing the first undulating curve according to the second preset plane to obtain a plurality of curve segments.

[0058] For the embodiment of the present application, the electronic device divides the first undulating curve into multiple curve segments according to the second preset plane at preset intervals and at equal intervals, so as to facilitate the subsequent analysis of the undulation degree of the crack characteristics in multiple regional segments along the lane extension direction.

[0059] Sc, calculating the first curvature corresponding to each of the plurality of curve segments and the second curvature of the plurality of second undulating curves.

[0060] For the embodiment of the present application, the electronic device fits each curve segment into a function, and then calculates the curvature of the function, i.e., the first curvature, which characterizes the curvature of each curve segment. The electronic device calculates the second curvatures corresponding to each of the plurality of second undulating curves in the same manner, and the first curvature and the second curvature are the curvatures corresponding to the same crack feature in two directions in the same crack region.

[0061] Sd, based on the first curvature and the second curvature, determine the undulation score of the lane corresponding to each curve segment.

[0062] For the embodiment of the present application, after the electronic device determines the first curvature and the second curvature, it can add the first curvature and the second curvature to obtain the ups and downs score of the lane corresponding to the curve segment, or the staff can set the corresponding coefficients of the first curvature and the second curvature according to the different importance of the first curvature and the second curvature, and then perform weighted calculation on the first curvature and the second curvature according to the corresponding coefficients to obtain the ups and downs score of the lane area where the curve segment is located. It is more accurate to divide a crack feature into multiple areas and comprehensively determine the ups and downs score of each area based on the curvatures in two directions corresponding to each area.

[0063] Se, the flatness of the lane surface in the area where each crack feature is located is determined based on the undulation score and the first similarity.

[0064] In the embodiment of the present application, the fluctuation scores of multiple areas of the crack feature and the first similarity between the first fluctuation curve of the crack feature along the extension direction of the lane and the preset straight line are key factors for characterizing the flatness of the lane surface in the entire crack feature area. Therefore, the electronic device can determine the flatness of the lane surface in each crack feature area based on the fluctuation score and the first similarity. The electronic device calculates and analyzes the flatness of the lane surface in each crack feature area through the contents disclosed in steps Sa to Se.

[0065] A possible implementation of the embodiment of the present application is to determine the flatness of the lane surface in the area where each crack feature is located based on the undulation score and the first similarity in step Se, which specifically includes step 1, step 2, step 3 and step 4, wherein: Step 1: determine the target curve segment whose fluctuation score reaches a preset score threshold.

[0066] For the embodiment of the present application, the preset score threshold is used as the dividing point of whether the fluctuation degree is too large. After the electronic device determines the fluctuation scores of multiple intervals of a crack feature, they are all compared with the preset score threshold, so as to determine the target curve segment whose fluctuation score reaches the preset score threshold. The more target curve segments there are, the greater the probability and risk of damage or abnormality on the lane surface in the area where the crack is located.

[0067] Step 2: Determine a first ratio of the number of target curve segments to the number of all curve segments.

[0068] For the embodiment of the present application, the electronic device divides the model fragment of the crack feature multiple times in the manner in step S1043 to obtain multiple curve segments, the electronic device counts the number of multiple curve segments to obtain the number of all curve segments, the electronic device counts the number of target curve segments to obtain the number of target curve segments, and then the electronic device calculates the ratio of the number of target curve segments to the number of all curve segments, that is, the first proportion. The larger the first proportion, the more curve segments with excessive undulations there are, and the worse the flatness of the corresponding lane surface.

[0069] Step three, determine the average value of the fluctuation score of the target curve segment.

[0070] For the embodiment of the present application, the electronic device calculates the average value of the fluctuation scores of all target curve segments through the average value calculation formula. The average value of the fluctuation score represents the specific fluctuation level of the curve segment with excessive fluctuation, that is, excessive flatness. The larger the average value of the fluctuation score, the greater the fluctuation degree and the worse the flatness.

[0071] Step 4: Determine the flatness of the lane surface in the area where each crack feature is located based on the first proportion of the target curve segment, the average value of the undulation score, the first similarity, and the corresponding coefficients.

[0072] For the embodiment of the present application, in summary, the first proportion of the target curve segment, the average value of the fluctuation score, and the first similarity are all key factors affecting the smoothness of the lane surface in the area where the crack feature is located, and the degree of influence on the smoothness is different. Therefore, the staff sets different coefficients for the first proportion, the average value of the fluctuation score, and the first similarity and stores them in the electronic device. After the electronic device determines the first proportion, the average value of the fluctuation score, and the first similarity, it calls the corresponding coefficients for weighted calculation to obtain a score, which can characterize the smoothness of the lane surface in the area where each crack feature is located. The smoothness determined by comprehensive analysis of the first proportion, the average value of the fluctuation score, and the first similarity is more accurate.

[0073] In a possible implementation of the embodiment of the present application, in step S105, the crack abnormality of each lane is determined based on the flatness, crack depth and crack characteristics, and specifically includes step S1051 (not shown in the figure), step S1052 (not shown in the figure), step S1053 (not shown in the figure) and step S1054 (not shown in the figure), wherein: S1051, determine the length and width of each crack feature.

[0074] For the embodiment of the present application, after the electronic device recognizes the crack feature through the image information, the number of pixels in the length direction of the crack feature from the image information is counted, and the number of pixels can represent the length of the crack feature. Then the electronic device takes multiple sampling points for the crack feature, counts the number of pixels along the direction perpendicular to the length at the sampling points, thereby obtaining the number of pixels along the direction perpendicular to the length at multiple sampling points, and then calculates the average number of pixels for the number of pixels at all sampling points through the average value calculation formula, and the average number of pixels can be used to represent the width of the crack feature.

[0075] S1052: Determine the attribute score of each crack based on the length, width, crack depth and their corresponding coefficients.

[0076] For the embodiments of the present application, the length, width and crack depth of the crack feature are all key factors that characterize the severity of the crack feature itself. The staff can set the corresponding coefficients of the length, width and crack depth in advance and store them in the electronic device. After the electronic device determines the length, width and crack depth of the crack feature, it calls the corresponding coefficients for weighted calculation to obtain the attribute score of the crack feature itself.

[0077] S1053: Determine the abnormality score of each crack based on the attribute score of each crack, the flatness corresponding to each crack, and the corresponding coefficients.

[0078] For the embodiment of the present application, after the electronic device determines the attribute score and flatness of each crack, the abnormality of each crack can be comprehensively determined. The staff sets corresponding coefficients for the attribute score of the crack feature itself and the flatness of the lane surface area and stores them in the electronic device. After the electronic device determines the attribute score and flatness, it calls the corresponding coefficients for weighted calculation to determine the abnormality score of each crack. It is more accurate to comprehensively analyze and determine the abnormality score through the attribute score of the crack itself and the flatness of the surface of the lane area.

[0079] S1054: Sum the abnormal scores of all crack features of each lane to obtain the crack abnormality degree of each lane.

[0080] For the embodiment of the present application, after the electronic device determines the abnormal scores of all crack features on each lane, the abnormal scores of all crack features are summed up to obtain the total score corresponding to each lane, that is, the crack abnormality degree.

[0081] In a possible implementation of the embodiment of the present application, in step S101, the drone is controlled to collect image information and point cloud data for each lane of the bridge to be tested, specifically including step S1011 (not shown in the figure), step S1012 (not shown in the figure), step S1013 (not shown in the figure), step S1014 (not shown in the figure), step S1015 (not shown in the figure) and step S1016 (not shown in the figure), wherein: S1011, obtaining surveillance video of each lane within a preset historical time period.

[0082] For the embodiment of the present application, the preset historical time period can be the past month, the past three months or other time periods, which can be set by the staff through the visual operation interface. The surveillance video is collected by multiple surveillance cameras installed on the bridge and obtained by the electronic device.

[0083] S1012, performing feature recognition on the surveillance video to obtain the characteristics of vehicles passing through each lane within a preset historical time period and the total number of vehicles.

[0084] For the embodiment of the present application, the electronic device inputs the monitoring video into the trained network model for vehicle feature recognition, thereby obtaining the vehicle features and the total number of vehicles passing through each lane of the bridge within a preset historical time period. The network model can be a convolutional neural network model, a recurrent neural network model, or other types of network models, which are not limited here.

[0085] S1013, classifying the vehicle features according to preset categories to obtain the number of vehicles corresponding to each preset category.

[0086] Among them, the preset categories include small vehicles and large vehicles.

[0087] For the embodiment of the present application, small vehicles and large vehicles can be classified according to the length and volume of the vehicle, for example, private cars are small vehicles and trucks are large vehicles. The electronic device classifies the identified vehicle features to determine the number of vehicles corresponding to each preset category. Specifically, the vehicle features can be classified according to the outline size in the surveillance video.

[0088] S1014, determining the number of vehicles corresponding to the large vehicles, and calculating a second ratio of the number of vehicles to the total number of vehicles.

[0089] For the embodiment of the present application, large vehicles, due to their heavy weight, will generate vibration or impact when passing through a bridge, which will have a significant impact on the durability and wear of the bridge. Therefore, the electronic device determines the number of vehicles corresponding to the large vehicles passing through each lane in a preset historical time period, and calculates the second ratio of the number of vehicles to the total number of vehicles passing through each lane in the preset historical time period. The larger the second ratio, the more large vehicles pass through the lane, the more serious the impact on the bridge, and the higher the priority when controlling the drone to collect image information and point cloud data for each lane.

[0090] S1015, sorting the second proportion of each lane to obtain a sorting result.

[0091] The sorting result represents the order of importance of all lanes.

[0092] For the embodiment of the present application, the electronic device sorts the second proportion of each lane from large to small to obtain a sorting result. The higher the ranking, the more important the lane is, and the more serious the impact of wear and damage in the preset historical time period.

[0093] S1016, controlling the UAV to collect image information and point cloud data for each lane of the bridge to be tested according to the sorting result.

[0094] For the embodiment of the present application, after the electronic device determines the sorting result, it generates a control instruction according to the sorting result and sends it to the drone, so that the drone determines the flight order of each lane according to the sorting result, flies according to the flight order, and collects image information and point cloud data for each lane. The collection order of each lane is determined according to the specific situation of the size of the vehicles passing through each lane in the preset historical time period, thereby optimizing the entire collection process.

[0095] In a possible implementation of the embodiment of the present application, step S105 further includes step S107 (not shown in the figure), step S108 (not shown in the figure), step S109 (not shown in the figure) and step S110 (not shown in the figure), wherein: S107, obtaining historical crack detection results for each lane.

[0096] The historical crack detection results include the number of historical cracks on each lane and the crack abnormality of each historical crack.

[0097] For the embodiment of the present application, after each crack detection on each lane, each detection result is stored in the local storage medium or server of the electronic device, so as to be stored and convenient for subsequent viewing and calling. The electronic device can retrieve the historical crack detection results from the local storage medium or server of the electronic device. The historical crack detection results record the crack conditions that have appeared on each lane, that is, the actual wear, damage or abnormality of each lane, so as to facilitate the subsequent correction of the crack abnormality of each lane according to the historical detection results.

[0098] S108, determining an average value of the fracture anomaly degrees of all historical fractures based on the fracture anomaly degrees of the historical fractures.

[0099] For the embodiment of the present application, the electronic device calls the crack abnormality of each lane's history cracks and calculates the average crack abnormality of all historical cracks through an average value calculation formula. The average value is used to more accurately characterize the overall level of historical cracks in each lane.

[0100] S109, determining a correction score based on the average value of the crack abnormality, the number of historical cracks, and the position of each lane in the sorting result.

[0101] For the embodiment of the present application, the more the number of historical cracks, the more severe the abnormal influences such as damage and wear on the lane, and the more forward the lane is in the sorting result, the more large vehicles have passed by, and the more severe the abnormal influences such as damage and wear on the lane. The average value of crack abnormality, the number of historical cracks, and the position of each lane in the sorting result all represent the actual level of damage or abnormality suffered by each lane at present. Therefore, the staff sets different coefficients for the average value of crack abnormality, the number of historical cracks, and the position of the lane in the sorting result, and the electronic device calls the corresponding coefficients to perform weighted calculation on the average value of crack abnormality, the number of historical cracks, and the position of the lane in the sorting result to obtain the correction score. The correction score obtained according to the above three factors is more accurate. Since the value of the position of the lane in the sorting result is inversely proportional to the abnormal influence suffered by the vehicle, in order to facilitate calculation and ensure the correctness and logic of the correction score, the electronic device can determine the inverse of the position of the lane in the sorting result for calculation.

[0102] S110, performing correction calculation on the crack abnormality of each lane based on the correction score to obtain a corrected crack abnormality.

[0103] For the embodiment of the present application, after the electronic device determines the correction score and crack abnormality of each lane, it can directly calculate the sum of the crack abnormality value and the correction score to perform a correction calculation to obtain the corrected crack abnormality. By correcting the crack abnormality of each lane, the crack abnormality of each lane is made more accurate.

[0104] The above-mentioned embodiment introduces a method for detecting bridge surface cracks based on a drone from the perspective of method flow. The following embodiment introduces a device for detecting bridge surface cracks based on a drone from the perspective of a virtual module or a virtual unit. Please refer to the following embodiment for details.

[0105] The present application embodiment provides a bridge surface crack detection device 20 based on a drone, such as Figure 2 As shown, the UAV-based bridge surface crack detection device 20 may specifically include: The control module 201 is used to control the UAV to collect image information and point cloud data for each lane of the bridge to be tested; A crack detection module 202 is used to perform crack detection on the image information to obtain crack features on each lane; A modeling module 203, used to model each lane based on the point cloud data, so as to obtain a three-dimensional model of the surface of each lane; A flatness determination module 204 is used to determine the crack depth of each lane and the flatness of the lane surface in the area where each crack feature is located based on the three-dimensional model; A first abnormality determination module 205, for determining the abnormality of cracks of each lane based on the flatness, crack depth and crack characteristics; The second abnormality determination module 206 is used to determine the abnormality degree of cracks of the bridge to be tested based on the abnormality degree of cracks of each lane.

[0106] The embodiment of the present application discloses a bridge surface crack detection device 20 based on a drone, wherein a control module 201 controls the drone to fly along each lane according to the multiple lanes divided on the bridge, and collects image information and point cloud data of the lane surface during the flight. The image information records the specific condition of the lane surface and whether there are cracks. Therefore, the crack detection module 202 performs crack detection on the image information to obtain the crack characteristics on each lane. The point cloud data records the three-dimensional terrain of the lane surface. Therefore, the modeling module 203 performs modeling based on the point cloud data to obtain a three-dimensional model of the surface of each lane. The flatness determination module 204 can accurately determine the crack depth and crack size of the cracks on each lane according to the three-dimensional model. The flatness of the lane surface in the area, the worse the flatness, the worse and more uneven the lane condition, the greater the risk of damage or abnormality to the overall structure of the bridge here, which means that the degree of abnormality of the cracks here is high. Therefore, the first abnormality determination module 205 comprehensively determines the crack abnormality of each lane based on the flatness, crack depth and crack characteristics, that is, a more accurate crack abnormality is obtained by combining the flatness of the lane surface at the crack and the properties of the crack itself. Then the second abnormality determination module 206 combines the crack abnormality of each lane to determine the crack abnormality of the entire bridge to be tested. Compared with the current analysis of only the specific situation of the crack itself, the present application combines the undulation and flatness at the crack to determine the abnormality of the entire bridge more accurately.

[0107] In a possible implementation of the embodiment of the present application, when the flatness determination module 204 determines the flatness of the lane surface in the area where each crack feature is located based on the three-dimensional model, it is specifically used to: Determine from the three-dimensional model a model segment of the lane area corresponding to each crack feature; The model segment is segmented according to a first preset plane to obtain a first undulating curve on the surface of the lane area, where the first preset plane is a plane perpendicular to the bridge deck and consistent with the extension direction of the lane; The model segment is segmented multiple times according to a second preset plane to obtain multiple second undulating curves on the surface of the lane area, where the second preset plane is a plane perpendicular to the bridge deck and perpendicular to the extension direction of the lane; The flatness of the lane surface in the area where each crack feature is located is determined based on the first undulation curve and the second undulation curve.

[0108] In a possible implementation of the embodiment of the present application, when the flatness determination module 204 determines the flatness of the lane surface in the area where each crack feature is located based on the first undulation curve and the second undulation curve, it is specifically used to: Calculating the similarity between the first undulating curve and the preset straight line to obtain a first similarity; Segmenting the first undulating curve according to a second preset plane to obtain a plurality of curve segments; Calculating first curvatures corresponding to the plurality of curve segments and second curvatures of the plurality of second undulating curves; Determine the undulation score of the lane corresponding to each curved segment based on the first curvature and the second curvature; The flatness of the lane surface in the area where each crack feature is located is determined based on the undulation score and the first similarity.

[0109] In a possible implementation of the embodiment of the present application, when the flatness determination module 204 determines the flatness of the lane surface in the area where each crack feature is located based on the undulation score and the first similarity, it is specifically used to: Determine the target curve segment whose ups and downs score reaches a preset score threshold; Determine a first ratio of the number of target curve segments to the number of all curve segments; Determine the average value of the ups and downs score of the target curve segment; The flatness of the lane surface in the area where each crack feature is located is determined based on the first proportion of the target curve segment, the average value of the undulation score, the first similarity, and the corresponding coefficients.

[0110] In a possible implementation of the embodiment of the present application, when the first abnormality determination module 205 determines the crack abnormality of each lane based on the flatness, crack depth and crack characteristics, it is specifically used to: Determine the length and width of each crack feature; Determine the attribute score of each crack based on the length, width, crack depth and their corresponding coefficients; The abnormality score of each crack is determined based on the attribute score of each crack, the flatness corresponding to each crack and the corresponding coefficients; The anomaly scores of all crack features of each lane are summed to obtain the crack anomaly degree of each lane.

[0111] In a possible implementation of the embodiment of the present application, the control module 201 is specifically used to control the drone to collect image information and point cloud data for each lane of the bridge to be tested: Obtain surveillance video of each lane within a preset historical time period; Perform feature recognition on the surveillance video to obtain the vehicle features and total number of vehicles passing through each lane within a preset historical time period; Classifying vehicle features according to preset categories to obtain the number of vehicles corresponding to each preset category, where the preset categories include small vehicles and large vehicles; Determine the number of vehicles corresponding to the large vehicles, and calculate the second ratio of the number of vehicles to the total number of vehicles; The second proportion of each lane is sorted to obtain a sorting result, and the sorting result represents the order of importance of all lanes; Control the UAV to collect image information and point cloud data for each lane of the bridge to be tested according to the sorting results.

[0112] In a possible implementation of the embodiment of the present application, the device 20 further includes: An acquisition module is used to acquire historical crack detection results of each lane, where the historical crack detection results include the number of historical cracks on each lane and the crack abnormality of each historical crack; An average value determination module, used to determine the average value of the fracture abnormality of all historical fractures based on the fracture abnormality of the historical fractures; A correction score determination module, used to determine a correction score based on an average value of crack abnormality, a number of historical cracks, and a position of each lane in the sorting result; The correction module is used to perform correction calculation on the crack abnormality of each lane based on the correction score to obtain a corrected crack abnormality.

[0113] Technicians in the relevant field can clearly understand that, for the convenience and brevity of description, the specific working process of the drone-based bridge surface crack detection device 20 described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0114] An electronic device is provided in an embodiment of the present application, such as Figure 3 As shown, Figure 3 The electronic device 30 shown includes: a processor 301 and a memory 303. The processor 301 and the memory 303 are connected, such as through a bus 302. Optionally, the electronic device 30 may also include a transceiver 304. It should be noted that in actual applications, the transceiver 304 is not limited to one, and the structure of the electronic device 30 does not constitute a limitation on the embodiments of the present application.

[0115] The processor 301 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. The processor 301 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0116] The bus 302 may include a path to transmit information between the above components. The bus 302 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 302 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but it does not mean that there is only one bus or only one type of bus.

[0117] The memory 303 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0118] The memory 303 is used to store the application code for executing the solution of the present application, and the execution is controlled by the processor 301. The processor 301 is used to execute the application code stored in the memory 303 to implement the contents shown in the above method embodiment.

[0119] The electronic devices include, but are not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), and fixed terminals such as digital TVs, desktop computers, etc. It can also be a server, etc. Figure 3 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0120] An embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer-readable storage medium is run on a computer, the computer can execute the corresponding contents of the aforementioned method embodiment. Compared with the related art, in the embodiment of the present application, a drone is controlled to fly along each lane according to the multiple lanes divided on the bridge, and image information and point cloud data of the lane surface are collected during the flight. The image information records the specific condition of the lane surface and whether there are cracks. Therefore, crack detection on the image information can obtain the crack characteristics on each lane. The point cloud data records the three-dimensional terrain conditions of the lane surface. Therefore, a three-dimensional model of the surface of each lane is obtained by modeling based on the point cloud data. According to the three-dimensional model, the crack depth of the cracks on each lane and the flatness of the lane surface in the area where the cracks are located can be accurately determined. The worse the smoothness, the worse and more uneven the lane condition is, and the greater the risk of damage or abnormality to the overall structure of the bridge here, which means that the degree of abnormality of the cracks here is high. Therefore, the crack abnormality of each lane is determined comprehensively based on the smoothness, crack depth and crack characteristics, that is, a more accurate crack abnormality is obtained by combining the smoothness of the lane surface at the crack and the properties of the crack itself. Then, the crack abnormality of the entire bridge to be tested can be determined by combining the crack abnormality of each lane. Compared with the current analysis of only the specific situation of the crack itself, this application combines the undulation and smoothness of the crack to determine the abnormality of the entire bridge more accurately.

[0121] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0122] The above description is only a partial implementation method of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A bridge surface crack detection method based on drone, characterized in that: include: Control the UAV to collect image information and point cloud data for each lane of the bridge to be tested; Performing crack detection on the image information to obtain crack features on each lane; Modeling each lane based on the point cloud data to obtain a three-dimensional model of the surface of each lane; Determining the crack depth of each lane and the flatness of the lane surface in the area where each crack feature is located based on the three-dimensional model; Determine the crack abnormality of each lane based on the flatness, crack depth and crack characteristics; The abnormal degree of cracks of the bridge to be tested is determined based on the abnormal degree of cracks of each lane.

2. The method for detecting cracks on a bridge surface based on an unmanned aerial vehicle according to claim 1, characterized in that: Determining the flatness of the lane surface in the area where each crack feature is located based on the three-dimensional model includes: Determine from the three-dimensional model a model segment of the lane area corresponding to each crack feature; The model segment is segmented according to a first preset plane to obtain a first undulating curve on the surface of the lane area, wherein the first preset plane is a plane perpendicular to the bridge deck and consistent with the extension direction of the lane; The model segment is segmented multiple times according to a second preset plane to obtain multiple second undulating curves on the surface of the lane area, wherein the second preset plane is a plane perpendicular to the bridge deck and perpendicular to the extension direction of the lane; The flatness of the lane surface in the area where each crack feature is located is determined based on the first undulation curve and the second undulation curve.

3. The method for detecting cracks on a bridge surface based on an unmanned aerial vehicle according to claim 2, characterized in that: The step of determining the flatness of the lane surface in the area where each crack feature is located based on the first undulation curve and the second undulation curve includes: Calculating the similarity between the first undulating curve and a preset straight line to obtain a first similarity; Segmenting the first undulating curve according to the second preset plane to obtain a plurality of curve segments; Calculating the first curvature corresponding to each of the plurality of curve segments and the second curvature of the plurality of second undulating curves; Determine a undulation score of the lane corresponding to each curved segment based on the first curvature and the second curvature; The flatness of the lane surface in the area where each crack feature is located is determined based on the undulation score and the first similarity.

4. The method for detecting cracks on a bridge surface based on an unmanned aerial vehicle according to claim 3 is characterized in that: The step of determining the flatness of the lane surface in the area where each crack feature is located based on the undulation score and the first similarity includes: Determine the target curve segment whose ups and downs score reaches a preset score threshold; Determining a first ratio of the number of the target curve segments to the number of all curve segments; Determining an average value of the ups and downs scores of the target curve segment; The flatness of the lane surface in the area where each crack feature is located is determined based on the first proportion of the target curve segment, the average value of the undulation score, the first similarity and the corresponding coefficients.

5. The method for detecting cracks on a bridge surface based on an unmanned aerial vehicle according to claim 1, characterized in that: The determining of the crack abnormality of each lane based on the flatness, crack depth and crack characteristics includes: Determine the length and width of each crack feature; Determine the attribute score of each crack based on the length, width, crack depth and their corresponding coefficients; Determine the abnormality score of each crack based on the attribute score of each crack, the flatness corresponding to each crack and the corresponding coefficients; The abnormal scores of all crack features of each lane are summed to obtain the crack abnormality degree of each lane.

6. The method for detecting cracks on a bridge surface based on an unmanned aerial vehicle according to claim 1, characterized in that: The controlling of the UAV to collect image information and point cloud data for each lane of the bridge to be tested includes: Obtain surveillance video of each lane within a preset historical time period; Performing feature recognition on the surveillance video to obtain vehicle features and the total number of vehicles passing through each lane within a preset historical time period; Classifying the vehicle features according to preset categories to obtain the number of vehicles corresponding to each preset category, wherein the preset categories include small vehicles and large vehicles; Determining the number of vehicles corresponding to the large vehicles, and calculating a second ratio of the number of vehicles to the total number of vehicles; Sorting the second proportion of each lane to obtain a sorting result, wherein the sorting result represents the order of importance of all lanes; The drone is controlled to collect image information and point cloud data for each lane of the bridge to be tested according to the sorting result.

7. The method for detecting cracks on a bridge surface based on an unmanned aerial vehicle according to claim 6, characterized in that: The method further comprises: Acquire a historical crack detection result for each lane, wherein the historical crack detection result includes the number of historical cracks on each lane and the crack abnormality of each historical crack; Determining an average value of the fracture anomaly degrees of all historical fractures based on the fracture anomaly degrees of the historical fractures; Determining a correction score based on the average value of the crack abnormality, the number of historical cracks, and the position of each lane in the sorting result; The crack abnormality of each lane is corrected and calculated based on the correction score to obtain a corrected crack abnormality.

8. A bridge surface crack detection device based on drone, characterized in that: include: A control module, used to control the UAV to collect image information and point cloud data for each lane of the bridge to be tested; A crack detection module, used to perform crack detection on the image information to obtain crack features on each lane; A modeling module, used for modeling each lane based on the point cloud data, so as to obtain a three-dimensional model of the surface of each lane; A flatness determination module, used to determine the crack depth of each lane and the flatness of the lane surface in the area where each crack feature is located based on the three-dimensional model; A first abnormality determination module, configured to determine the crack abnormality of each lane based on the flatness, crack depth and crack characteristics; The second abnormality determination module is used to determine the abnormality degree of cracks of the bridge to be tested based on the abnormality degree of cracks of each lane.

9. An electronic device, characterized in that: It includes: at least one processor; Memory; At least one application, wherein the at least one application is stored in the memory and configured to be executed by the at least one processor, and the at least one application is used to execute a bridge surface crack detection method based on a drone according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed in a computer, the computer is caused to execute the method for detecting bridge surface cracks based on an unmanned aerial vehicle as described in any one of claims 1 to 7.

Images