A space-ground two-way cooperative road disease detection system and method based on homologous beidou timing

Abstract

The application discloses a kind of based on homologous road disease detection system and method of two-way cooperation of road based on compass, it is related to road detection technical field, for solving the problem that mutual disjointedness of pavement disease detection and surrounding structure abnormal inspection in existing road detection, air-ground data time reference and spatial reference are difficult to unify, trigger review and cause diagnosis continuity is insufficient, by vehicle detection subsystem and unmanned aerial vehicle detection subsystem receive same group compass satellite signal to carry out hardware synchronization, collect pavement detection data and structure detection data, execute vehicle grading trigger unmanned aerial vehicle and the two-way cooperation of unmanned aerial vehicle abnormal trigger vehicle, based on UTC time and CGCS2000 coordinate system carry out space-time alignment, generate cause label, risk level and maintenance suggestion, realize the cooperative detection and diagnosis evaluation of road global disease.

Description

Technical Field

[0001] This invention relates to the field of road inspection technology, and more specifically, to a two-way collaborative air-ground road defect detection system and method based on BeiDou time synchronization. Background Technology

[0002] Road defect detection is a crucial part of road maintenance and management. It typically involves inspecting pavement defects such as cracks, potholes, alligator cracks, and subsidence, as well as the condition of surrounding structures such as slopes, retaining walls, sections under bridges, and sections under overpasses. Current road detection systems mostly use vehicle-mounted equipment to acquire pavement images and point cloud data, or drone-based inspection equipment to acquire images and point cloud data of surrounding structures. The inspection results are then archived and reported by a backend system.

[0003] The existing technology has the following shortcomings: Currently, existing technologies separate vehicle-mounted road surface detection and unmanned structure inspection, lacking a mechanism for air-to-ground data synchronization and bidirectional triggering under the same time synchronization, making it difficult to establish a correlation between road surface defects and anomalies in surrounding structures. This results in insufficient synergy and diagnostic completeness in the detection of road defects across the entire area. Therefore, this paper proposes an air-to-ground bidirectional collaborative road defect detection system and method based on the same BeiDou time synchronization.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] To overcome the aforementioned deficiencies in the prior art, embodiments of the present invention provide an air-ground two-way collaborative road defect detection system and method based on BeiDou time synchronization. The system uses a vehicle-mounted detection subsystem and a drone detection subsystem to receive the same set of BeiDou satellite signals, generating vehicle-mounted time synchronization markers, drone time synchronization markers, and synchronization status markers. When the synchronization status markers are available, road defect detection parameters and structure anomaly detection parameters are generated respectively. Air-ground two-way collaborative detection is achieved through vehicle-mounted graded triggering of drones and drone anomaly triggering of vehicles. Then, based on UTC time and the CGCS2000 coordinate system, associated load capacity is generated. Based on the associated load capacity and defect anomaly parameters, causal labels, risk levels, maintenance recommendations, and detection reports are generated to solve the problems of mis-association and missed association in cross-platform collaborative detection of road defects.

[0006] This invention provides a two-way collaborative air-ground road defect detection method based on BeiDou time synchronization, comprising the following steps: Step S1: Receive BeiDou satellite signals and output PPS second pulses to the vehicle-mounted detection subsystem and the UAV detection subsystem respectively. When the vehicle-mounted detection subsystem performs road surface data collection, it generates a vehicle-mounted time synchronization mark. When the UAV detection subsystem performs structure data collection, it generates a UAV time synchronization mark. Generate a synchronization status mark based on the vehicle-mounted time synchronization mark and the UAV time synchronization mark. Step S2: When the synchronization status marker is available, collect road surface image data and road surface point cloud data and generate road surface distress coordinates, road surface distress type, road surface distress scale and road surface distress change rate. Control the UAV detection subsystem to collect structure image data and structure point cloud data and generate structure anomaly coordinates, structure anomaly type, structure anomaly scale and structure anomaly change rate. Step S3: Determine whether to generate a UAV trigger command based on the type of pavement distress, the scale of pavement distress, the rate of change of pavement distress, and the marking of key impact areas. After generating the UAV trigger command, the UAV detection subsystem performs multi-angle surround observation of the surrounding structures. Based on the abnormal coordinates and types of the structures, generate a vehicle-mounted fine detection command. After generating the vehicle-mounted fine detection command, the vehicle-mounted detection subsystem performs fine mode detection on the associated road sections. Step S4: Receive pavement distress coordinates, pavement distress type, pavement distress scale, pavement distress change rate, structure anomaly coordinates, structure anomaly type, structure anomaly scale, structure anomaly change rate, vehicle-mounted time markers and UAV time markers, perform spatiotemporal alignment of pavement distress objects and structure anomaly objects based on UTC time and CGCS2000 coordinate system, and generate associated bearing capacity; Step S5: Generate cause labels based on associated load volume, vehicle-mounted time markers, UAV time markers, pavement distress type, pavement distress scale, pavement distress change rate, structure anomaly type, structure anomaly scale, and structure anomaly change rate. Generate risk level and maintenance recommendations based on cause labels, and generate an inspection report based on cause labels, risk level, and maintenance recommendations.

[0007] In a preferred embodiment, in step S1, the vehicle-mounted detection subsystem receives BeiDou timing signals through the first BeiDou RTK module, and the UAV detection subsystem receives BeiDou timing signals through the second BeiDou RTK module. The first and second Beidou RTK modules receive the same set of Beidou satellite signals and output PPS second pulses respectively, so that the vehicle-mounted detection subsystem generates vehicle-mounted timing markers and the UAV detection subsystem generates UAV timing markers. The cloud-based collaborative system reads the vehicle-mounted time synchronization markers and UAV time synchronization markers, calculates the time deviation from the same source, and verifies the time source number, satellite combination number, and PPS pulse number in the vehicle-mounted time synchronization markers and UAV time synchronization markers. Based on the verification results, it generates a synchronization status marker.

[0008] In a preferred embodiment, in step S2, the vehicle-mounted detection subsystem collects road surface image data and road surface point cloud data through road surface sensors, generates road surface defect coordinates, road surface defect types, and road surface defect scales based on the road surface image data and road surface point cloud data, and generates the road surface defect change rate based on the road surface defect scales and historical road surface defect scales.

[0009] In a preferred embodiment, in step S2, the UAV detection subsystem collects structure image data and structure point cloud data, generates structure anomaly coordinates, structure anomaly type and structure anomaly scale based on the structure image data and structure point cloud data, and generates structure anomaly change rate based on the structure anomaly scale and historical structure anomaly scale.

[0010] In a preferred embodiment, in step S3, basic road structure data in the road asset database is read. The basic road structure data includes structure coordinates, structure influence range, and road section mapping relationship. Key influence area markers are generated based on pavement defect coordinates and structure influence range.

[0011] In a preferred embodiment, in step S3, a UAV trigger command is generated based on the matching grading condition table of pavement distress type, pavement distress scale, pavement distress change rate and key influence area markers. The UAV trigger command includes pavement distress coordinates, pavement distress type, pavement distress scale and observation radius of surrounding structures. When the UAV detection subsystem identifies the abnormal coordinates and abnormal type of a structure, it generates the associated road segment based on the abnormal coordinates, the area of ​​influence of the structure, and the mapping relationship of road segments, and generates a vehicle-mounted fine detection command, enabling the vehicle-mounted detection subsystem to perform fine mode detection on the associated road segment.

[0012] In a preferred embodiment, in step S4, when the synchronization status is marked as available, the cloud-based collaborative system performs spatiotemporal alignment of road surface defects and structural anomalies based on UTC time and CGCS2000 coordinate system to generate co-origin time deviation and open-ground spatial interval.

[0013] In a preferred embodiment, in step S4, time numbers are generated for candidate objects within the same spatial window according to the same source time deviation, spatial numbers are generated according to the open space interval, time proximity is generated based on the time numbers, and spatial proximity is generated based on the spatial numbers. The system generates type correlation quantities based on the combined pavement distress types and structure anomaly types, and generates correlation bearing quantities based on temporal proximity quantities, spatial proximity quantities, and type correlation quantities.

[0014] In a preferred embodiment, in step S4, candidate open spaces are determined based on the associated acceptance volume, and causal labels are generated based on the candidate open spaces, vehicle-mounted time markers, UAV time markers, road surface distress types, road surface distress scales, road surface distress change rates, structure anomaly types, structure anomaly scales, and structure anomaly change rates. Risk levels are generated based on causal labels, associated load, pavement distress scale, pavement distress change rate, structural anomaly scale, and structural anomaly change rate. Maintenance recommendations are generated based on cause labels and risk levels, and inspection reports are generated based on pavement distress coordinates, pavement distress types, pavement distress dimensions, structure anomaly coordinates, structure anomaly types, structure anomaly dimensions, cause labels, risk levels, and maintenance recommendations.

[0015] A bidirectional collaborative air-ground road defect detection system based on BeiDou time synchronization includes a hardware synchronization module, a conventional detection module, a bidirectional triggering module, a spatiotemporal alignment module, a diagnostic assessment module, and a report generation module. The functions of each module are as follows: The hardware synchronization module is used to receive the same set of BeiDou satellite signals through the first BeiDou RTK module of the vehicle-mounted detection subsystem and the second BeiDou RTK module of the UAV detection subsystem, respectively output PPS second pulses to the vehicle-mounted detection subsystem and the UAV detection subsystem, and generate vehicle timing mark, UAV timing mark and synchronization status mark; The conventional detection module is used to control the vehicle-mounted detection subsystem to collect road surface image data and road surface point cloud data, and generate road surface distress coordinates, road surface distress type, road surface distress scale, and road surface distress change rate; it is also used to control the UAV detection subsystem to collect structure image data and structure point cloud data, and generate structure anomaly coordinates, structure anomaly type, structure anomaly scale, and structure anomaly change rate. The bidirectional triggering module is used to classify and determine triggering based on pavement distress type, pavement distress scale, pavement distress change rate, and key impact zone markers. When the UAV triggering conditions in the classification condition table are met, it generates a UAV triggering command, enabling the UAV detection subsystem to perform multi-angle surround observation of surrounding structures. It is also used to generate vehicle-mounted fine detection commands when the UAV detection subsystem identifies abnormal coordinates and abnormal types of structures, enabling the vehicle-mounted detection subsystem to perform fine mode detection on the associated road sections. The spatiotemporal alignment module is used to align the coordinates of road surface defects, structural anomalies, vehicle-mounted time markers, and UAV time markers based on UTC time and the CGCS2000 coordinate system when the synchronization status is marked as available, generating the same source time deviation, air-ground spatial interval, time proximity quantity, spatial proximity quantity, type association quantity, and association acceptance quantity. The diagnostic assessment module is used to generate cause labels based on associated load volume, vehicle-mounted time markers, UAV time markers, pavement distress type, pavement distress scale, pavement distress change rate, structure anomaly type, structure anomaly scale, and structure anomaly change rate, and to generate risk level and maintenance recommendations based on the cause labels. The report generation module is used to generate inspection reports based on pavement distress coordinates, pavement distress type, pavement distress scale, structure anomaly coordinates, structure anomaly type, structure anomaly scale, cause label, risk level, and maintenance recommendations.

[0016] The technical effects and advantages of this invention are as follows: This invention achieves time synchronization between the vehicle-mounted detection subsystem and the UAV detection subsystem through BeiDou time synchronization. It acquires road surface defect data and structure anomaly data through conventional detection, and achieves collaborative verification between vehicle-mounted and UAV detection through bidirectional triggering. It establishes the correlation between road surface defects and structure anomalies through spatiotemporal alignment, and generates cause labels, risk levels, maintenance recommendations, and detection reports through diagnostic assessment. This invention can incorporate road surface defect detection and surrounding structure anomaly inspection into the same collaborative detection process, improving the synergy, correlation accuracy, and diagnostic completeness of road surface defect detection. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a two-way collaborative air-ground road defect detection method based on BeiDou time synchronization according to the present invention.

[0018] Figure 2 This is a schematic diagram of the module framework of a bidirectional collaborative road defect detection system based on BeiDou time synchronization according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] This invention enables the vehicle-mounted detection subsystem and the UAV detection subsystem to operate under a unified time reference through BeiDou time synchronization. By integrating road surface defect detection and abnormal inspection of surrounding structures into the same air-ground collaborative detection process through routine detection, bidirectional triggering, spatiotemporal alignment, and diagnostic evaluation, it outputs cause labels, risk levels, maintenance recommendations, and detection reports.

[0021] Example 1 Reference Figure 1A two-way collaborative air-to-ground road defect detection method based on BeiDou time synchronization includes the following steps: Step S1: Receive BeiDou satellite signals and output PPS second pulses to the vehicle-mounted detection subsystem and the UAV detection subsystem respectively. When the vehicle-mounted detection subsystem performs road surface data collection, it generates a vehicle-mounted time synchronization mark. When the UAV detection subsystem performs structure data collection, it generates a UAV time synchronization mark. Generate a synchronization status mark based on the vehicle-mounted time synchronization mark and the UAV time synchronization mark. Step S2: When the synchronization status marker is available, collect road surface image data and road surface point cloud data and generate road surface distress coordinates, road surface distress type, road surface distress scale and road surface distress change rate. Control the UAV detection subsystem to collect structure image data and structure point cloud data and generate structure anomaly coordinates, structure anomaly type, structure anomaly scale and structure anomaly change rate. Step S3: Determine whether to generate a UAV trigger command based on the type of pavement distress, the scale of pavement distress, the rate of change of pavement distress, and the marking of key impact areas. After generating the UAV trigger command, the UAV detection subsystem performs multi-angle surround observation of the surrounding structures. Based on the abnormal coordinates and types of the structures, generate a vehicle-mounted fine detection command. After generating the vehicle-mounted fine detection command, the vehicle-mounted detection subsystem performs fine mode detection on the associated road sections. Step S4: Receive pavement distress coordinates, pavement distress type, pavement distress scale, pavement distress change rate, structure anomaly coordinates, structure anomaly type, structure anomaly scale, structure anomaly change rate, vehicle-mounted time markers and UAV time markers, perform spatiotemporal alignment of pavement distress objects and structure anomaly objects based on UTC time and CGCS2000 coordinate system, and generate associated bearing capacity; Step S5: Generate cause labels based on associated load volume, vehicle-mounted time markers, UAV time markers, pavement distress type, pavement distress scale, pavement distress change rate, structure anomaly type, structure anomaly scale, and structure anomaly change rate. Generate risk level and maintenance recommendations based on cause labels, and generate an inspection report based on cause labels, risk level, and maintenance recommendations.

[0022] The specific implementation is as follows: In step S1, the vehicle-mounted detection subsystem receives the BeiDou timing signal through the first BeiDou RTK module and transmits the PPS second pulse output by the first BeiDou RTK module to the FPGA controller. The UAV detection subsystem receives BeiDou timing signals through the second BeiDou RTK module and transmits the PPS second pulse output by the second BeiDou RTK module to the low-power SOC controller.

[0023] The first and second Beidou RTK modules receive the same set of Beidou satellite signals. The vehicle-mounted detection subsystem records the vehicle-mounted timing mark each time road surface data collection is triggered, and the UAV detection subsystem records the UAV timing mark each time structure data collection is triggered.

[0024] PPS (Pulse Per Second) refers to the hardware synchronization pulse output by the Beidou RTK module every second. In this example, it is used to constrain the acquisition time of the vehicle-mounted detection subsystem and the UAV detection subsystem.

[0025] The vehicle-mounted time stamp is the UTC time stamp written by the vehicle-mounted detection subsystem when the road surface acquisition is triggered, while the UAV time stamp is the UTC time stamp written by the UAV detection subsystem when the structure acquisition is triggered.

[0026] The cloud-based fusion subsystem reads vehicle-mounted time synchronization markers and drone time synchronization markers, and calculates the time deviation between the two sources: , in, For time discrepancies originating from the same source, the vehicle-mounted time synchronization marker is... The drone timing marker is Homogeneous time deviation The smaller the value, the closer the data from the vehicle-mounted detection subsystem and the drone detection subsystem are under the same BeiDou timing reference. The cloud-integrated subsystem simultaneously verifies the timing source number, week-second count, and PPS edge sequence number in both vehicle-mounted and drone-mounted timing markers. When the time source number is consistent, the week-second count is continuous, and the PPS edge sequence number is corresponding, a synchronization status mark is generated as available. If any of these components are inconsistent, the synchronization status is marked as unavailable, triggering the vehicle-mounted detection subsystem and the drone detection subsystem to reacquire the BeiDou timing signal.

[0027] It should be added that the vehicle-mounted inspection subsystem refers to the inspection unit installed on the road inspection vehicle, which is used to collect road surface inspection data and generate road surface defect inspection results; the drone inspection subsystem refers to the structure inspection unit mounted on the drone platform, which is used to collect roadside structure inspection data and generate structure anomaly inspection results.

[0028] In step S2, the vehicle-mounted detection subsystem performs road surface detection in the conventional mode. The wheel encoder sends a mileage trigger signal to the FPGA controller, which controls the road surface sensor to collect road surface image data and road surface point cloud data. The edge computing unit identifies the coordinates, types, and scales of road surface defects from the road surface image data and road surface point cloud data.

[0029] The edge computing unit performs grayscale correction and texture enhancement on the road image data, extracting candidate defect areas corresponding to road cracks, potholes, alligator cracks, and subsidence. Simultaneously, the edge computing unit fits the road point cloud data to a ground reference surface, obtaining the elevation difference of each point within the candidate defect area relative to the ground reference surface. The edge computing unit then matches the pixel positions of the candidate defect areas in the road image data with their spatial positions in the road point cloud data to generate road defect coordinates. Road surface distress types are generated based on the image morphological features and point cloud elevation features of candidate distress areas. Image morphological features include the length, width, area, edge continuity, and texture dispersion of the candidate distress area; point cloud elevation features include the maximum elevation difference, average elevation difference, and elevation abrupt change range within the candidate distress area. When the candidate distress area has a long and continuous shape and the elevation difference is concentrated in a linear region, the generated road surface distress type is crack; when the candidate distress area has a concave surface shape and the elevation difference is concentrated in a closed region, the generated road surface distress type is pothole; when the candidate distress area contains multiple intersecting linear textures, the generated road surface distress type is alligator cracking; when the candidate distress area is continuously concave and has a gentle boundary transition, the generated road surface distress type is subsidence.

[0030] Pavement distress scale refers to the set of geometric quantities corresponding to the types of pavement distress. When the pavement distress type is crack, the pavement distress scale includes crack width and crack length. When the pavement distress type is pothole, the pavement distress scale includes pothole depth and pothole area.

[0031] The vehicle-mounted detection subsystem accesses the historical pavement defect database, reads the historical pavement defect scales corresponding to the pavement defect coordinates, and calculates the pavement defect change rate based on the pavement defect scales and historical pavement defect scales. ,in, This represents the rate of change in road surface defects. This is a scale for road surface defects. This is a scale for historical road surface defects. The historical interval is the time between the current detection time and the corresponding detection time of the historical pavement distress scale. The larger the pavement distress change rate, the faster the corresponding pavement distress develops within the historical interval. If there is no historical pavement distress scale corresponding to the pavement distress coordinates in the historical distress database, the pavement distress change rate is recorded as the initial record state and marked as the first discovery in the detection report.

[0032] The UAV inspection subsystem performs roadside structure inspections in a conventional mode. The low-power SOC controller uses PPS (pixel per second) to control aerial sensors to collect structure image data and point cloud data. It then identifies anomalous coordinates, anomaly types, and scales from these data. Anomaly scales refer to the set of geometric quantities corresponding to the anomaly type; for example, when the anomaly type is slope cracks, the scales include crack width and length; when the anomaly type is bridge scour, the scales include scour depth and scour area.

[0033] The UAV detection subsystem accesses the historical structure database, reads the historical structure anomaly scales corresponding to the structure anomaly coordinates, and calculates the structure anomaly change rate based on the structure anomaly scale and the historical anomaly scales. , in, The abnormal change rate of the structure, For abnormal dimensions of the structure, The scale of historical structural anomalies is defined as follows: the greater the rate of change of structural anomalies, the faster the corresponding structural anomaly develops within the historical interval. If there is no historical structural anomaly scale corresponding to the structural anomaly coordinates in the historical structural database, the rate of change of structural anomalies is recorded as the initial record state and marked as the first discovery in the inspection report.

[0034] In step S3, the vehicle-mounted detection subsystem performs a graded triggering judgment based on the type of pavement distress, the scale of pavement distress, the rate of change of pavement distress, and the key impact zone markers. The graded condition table is a data table formed by road maintenance specifications, historical treatment records, and road asset database. It records the UAV triggering conditions and the search radius of surrounding structures corresponding to different pavement distress types under different key impact zone markers. In this example, it is used to avoid the vehicle-mounted detection subsystem triggering the UAV detection subsystem when it detects minor pavement distress.

[0035] When the pavement distress type, pavement distress scale, pavement distress change rate, and key impact zone markers meet any of the UAV triggering conditions in the classification condition table, the vehicle-mounted detection subsystem generates a UAV triggering command. The UAV triggering command includes pavement distress coordinates, pavement distress type, pavement distress scale, and search radius of surrounding structures. When the pavement distress type, pavement distress scale, pavement distress change rate, and key impact zone markers do not meet the UAV triggering conditions in the classification condition table, the vehicle-mounted detection subsystem will only upload the pavement distress coordinates, pavement distress type, pavement distress scale, and pavement distress change rate to the cloud fusion subsystem.

[0036] After receiving the trigger command from the UAV, the UAV detection subsystem determines the observation area centered on the coordinates of the road surface defect and according to the search radius of surrounding structures. Within the observation area, it performs multi-angle surround observation. Multi-angle surround observation refers to the UAV detection subsystem performing circular or spiral movements around the target structure while simultaneously pointing the gimbal towards the target structure, collecting image data and point cloud data of the structure from different azimuths and pitch angles.

[0037] When the UAV detection subsystem identifies the abnormal coordinates and abnormal type of a structure, the UAV detection subsystem generates a vehicle-mounted fine detection instruction, which includes the abnormal coordinates, abnormal type, and associated road section of the structure. When the UAV inspection subsystem fails to identify abnormal coordinates and abnormal types of structures, it maintains normal mode and continues to upload structure inspection data.

[0038] After receiving the vehicle-mounted fine inspection command, the vehicle-mounted inspection subsystem performs fine mode inspection on the associated road sections.

[0039] Fine-scale pattern detection includes at least one of the following: increasing the trigger density of wheel encoders, reducing the candidate retention boundary for pavement distress identification, and enabling pavement distress scale measurement. The candidate retention boundary is determined by the identification confidence distribution of confirmed minor distress samples in the historical distress database, and the lower quartile of the identification confidence of confirmed minor distress samples is taken as the candidate retention boundary.

[0040] In step S4, the cloud-based fusion subsystem receives pavement distress coordinates, pavement distress types, pavement distress scales, pavement distress change rates, key influence zone markers, structure anomaly coordinates, structure anomaly types, structure anomaly scales, structure anomaly change rates, vehicle-mounted time synchronization markers, UAV time synchronization markers, and synchronization status markers. When the synchronization status marker is available, the cloud-based fusion subsystem performs spatiotemporal alignment based on UTC time and the CGCS2000 coordinate system.

[0041] The CGCS2000 coordinate system is a commonly used geodetic coordinate system in my country. In this example, it is used to transform the coordinates of road surface defects and the coordinates of structural anomalies to the same spatial reference.

[0042] The cloud-based fusion subsystem calculates the spatial interval between ground and air based on the coordinates of road surface defects and structural anomalies, and calculates the time deviation based on vehicle-mounted and UAV-based time markers. For candidate objects within the same inspection batch, the cloud-based fusion subsystem generates time numbers in ascending order of time deviation and spatial numbers in ascending order of spatial interval between ground and air, and calculates temporal and spatial proximity. , , in, For time proximity quantities, For spatial proximity, For time sequence number, For spatial sequence number, The candidate number represents the number of combinations of pavement distress objects and structural anomaly objects participating in alignment within the same detection batch. A larger temporal proximity indicates that the acquisition times of the vehicle-mounted detection subsystem and the UAV detection subsystem are closer; a larger spatial proximity indicates that the coordinates of pavement distress objects and structural anomalies are closer.

[0043] It should be added that the time sequence numbers mentioned above are numbered from 1 in ascending order of the same source time deviation, and the spatial sequence numbers are numbered from 1 in ascending order of the air-ground spatial interval.

[0044] The cloud-based fusion subsystem accesses the pavement distress and structural anomaly association table to read the type association values ​​corresponding to pavement distress types and structural anomaly types. The pavement distress and structural anomaly association table is a data table established based on historically confirmed cases, recording the frequency of co-occurrence of pavement distress types and structural anomaly types in the same causal chain; the type association value is the ratio of the historical co-occurrence frequency of the target combination to the total historical frequency of the pavement distress type.

[0045] The cloud-based fusion subsystem calculates the associated throughput based on temporal proximity, spatial proximity, and type association. , in, For related acceptance volume, As a type-related quantity, the association bearing capacity is used to characterize the spatiotemporal and type-related bearing capacity relationships between pavement distress objects and structural anomaly objects. A decrease in any of the time proximity quantity, spatial proximity quantity, or type-related quantity will decrease the association bearing capacity, thus preventing false associations based solely on temporal or spatial proximity.

[0046] The cloud-based fusion subsystem sorts objects by their associated load from largest to smallest, and identifies the top-ranked road surface defects and structural anomalies as candidate empty land objects; when all associated loads are zero, no candidate empty land objects are generated.

[0047] In step S5, the cloud-based fusion subsystem generates causal labels based on candidate air-to-ground objects, vehicle-mounted time markers, UAV time markers, pavement distress types, pavement distress scales, pavement distress change rates, structure anomaly types, structure anomaly scales, and structure anomaly change rates. Causal labels include structure-induced types, natural aging types, and types awaiting verification.

[0048] When there are open ground candidate objects, and the UAV timing mark is earlier than the vehicle timing mark or is in the same detection batch as the vehicle timing mark, the cloud fusion subsystem will mark the corresponding road surface defect object as a structure-induced type. When there are no open space candidate objects and the pavement distress type matches the distribution of natural aging records in the historical distress database, the cloud fusion subsystem will mark the corresponding pavement distress object as naturally aging type; when there are open space candidate objects but the vehicle-mounted time stamp is earlier than the UAV time stamp, or when there are no comparable records in the historical distress database, the cloud fusion subsystem will mark the corresponding pavement distress object as pending review type.

[0049] The distribution of natural aging records refers to the statistical distribution of pavement distress types, distress scales, and distress change rates corresponding to natural aging cases confirmed by maintenance personnel in the historical distress database. In this case, it is used to determine whether pavement distress conforms to the aging characteristics of conventional materials.

[0050] The cloud-based fusion subsystem generates risk levels based on causal tags, associated load, pavement distress scale, pavement distress change rate, structural anomaly scale, and structural anomaly change rate.

[0051] Risk levels include high risk, medium risk, and low risk, and the risk level is output from the risk level table; The risk level table is formed by archiving historical records of treated defects according to maintenance treatment levels. It records the risk level corresponding to different cause labels, different associated load ranges, different pavement defect scale ranges, different pavement defect change rate ranges, different structure anomaly scale ranges, and different structure anomaly change rate ranges.

[0052] When the cause label is structure-induced type and the risk level table maps the corresponding combination to high risk, the cloud fusion subsystem generates maintenance recommendations that include closing road sections, reinforcing surrounding structures, and repairing pavement defects. When the risk level table maps the corresponding combination to medium risk, the cloud-based fusion subsystem generates maintenance recommendations that include the re-inspection cycle and key monitoring objects. When the risk level table maps the corresponding combination to low risk, the cloud-based fusion subsystem generates maintenance recommendations that include record archiving and routine inspections.

[0053] The cloud-based fusion subsystem generates an inspection report based on pavement distress coordinates, distress types, distress dimensions, structural anomaly coordinates, anomaly types, anomaly dimensions, cause labels, risk levels, and maintenance recommendations. The inspection report outputs the overall road distress inspection results and links and saves pavement distress objects, surrounding structural anomaly objects, cause diagnosis content, and risk assessment content.

[0054] In the system implementation, the vehicle-mounted detection subsystem includes an FPGA controller, a wheel encoder, a road surface sensor, a first Beidou RTK module, and an edge computing unit.

[0055] The FPGA controller receives the mileage trigger signal output by the wheel encoder and controls the road surface sensor to collect road surface image data and road surface point cloud data; The first Beidou RTK module outputs PPS second pulses and vehicle-mounted timing markers to the FPGA controller; The edge computing unit generates road surface distress coordinates, distress types, distress scales, and distress change rates based on road surface image data and road surface point cloud data, and performs hierarchical trigger judgment.

[0056] The UAV detection subsystem includes a UAV platform, a low-power SOC controller, an airborne sensor, a second BeiDou RTK module, and a communication module.

[0057] The second Beidou RTK module outputs PPS second pulses and UAV timing markers to the low-power SOC controller; The low-power SOC controller controls the aerial sensor to collect image data and point cloud data of the structure, and generates the structure anomaly coordinates, structure anomaly type, structure anomaly scale and structure anomaly change rate; The communication module is used to receive drone trigger commands, send vehicle-mounted fine inspection commands, and upload structure inspection data to the cloud fusion subsystem.

[0058] The cloud-based fusion subsystem includes a data receiving module, a spatiotemporal alignment module, a cause diagnosis module, a risk assessment module, and a report generation module. The data receiving module receives data uploaded from the vehicle-mounted detection subsystem and the drone detection subsystem; the spatiotemporal alignment module generates data on source time deviation, air-to-ground spatial interval, temporal proximity, spatial proximity, type correlation, and correlation carrying capacity; the cause diagnosis module generates cause labels based on correlation carrying capacity and timing sequence; the risk assessment module generates risk levels and maintenance recommendations based on a risk level table; and the report generation module generates the detection report.

[0059] Example 2 like Figure 2 As shown, a two-way collaborative air-ground road defect detection system based on BeiDou time synchronization includes a hardware synchronization module, a routine detection module, a two-way triggering module, a spatiotemporal alignment module, a diagnostic evaluation module, and a report generation module. The functions of each module are as follows: The hardware synchronization module is used to receive the same set of BeiDou satellite signals through the first BeiDou RTK module of the vehicle-mounted detection subsystem and the second BeiDou RTK module of the UAV detection subsystem, respectively output PPS second pulses to the vehicle-mounted detection subsystem and the UAV detection subsystem, and generate vehicle timing mark, UAV timing mark and synchronization status mark; The conventional detection module is used to control the vehicle-mounted detection subsystem to collect road surface image data and road surface point cloud data, and generate road surface distress coordinates, road surface distress type, road surface distress scale, and road surface distress change rate; it is also used to control the UAV detection subsystem to collect structure image data and structure point cloud data, and generate structure anomaly coordinates, structure anomaly type, structure anomaly scale, and structure anomaly change rate. The bidirectional triggering module is used to classify and determine triggering based on pavement distress type, pavement distress scale, pavement distress change rate, and key impact zone markers. When the UAV triggering conditions in the classification condition table are met, it generates a UAV triggering command, enabling the UAV detection subsystem to perform multi-angle surround observation of surrounding structures. It is also used to generate vehicle-mounted fine detection commands when the UAV detection subsystem identifies abnormal coordinates and abnormal types of structures, enabling the vehicle-mounted detection subsystem to perform fine mode detection on the associated road sections. The spatiotemporal alignment module is used to align the coordinates of road surface defects, structural anomalies, vehicle-mounted time markers, and UAV time markers based on UTC time and the CGCS2000 coordinate system when the synchronization status is marked as available, generating the same source time deviation, air-ground spatial interval, time proximity quantity, spatial proximity quantity, type association quantity, and association acceptance quantity. The diagnostic assessment module is used to generate cause labels based on associated load volume, vehicle-mounted time markers, UAV time markers, pavement distress type, pavement distress scale, pavement distress change rate, structure anomaly type, structure anomaly scale, and structure anomaly change rate, and to generate risk level and maintenance recommendations based on the cause labels. The report generation module is used to generate inspection reports based on pavement distress coordinates, pavement distress type, pavement distress scale, structure anomaly coordinates, structure anomaly type, structure anomaly scale, cause label, risk level, and maintenance recommendations.

[0060] Finally, it should be noted that in this paper, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0061] Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0062] In this document, the singular forms “a,” “an,” and “the” may also include the plural forms unless the context clearly indicates otherwise. It should also be understood that terms such as “comprising / including” or “having” specify the presence of the stated features, integrals, steps, operations, components, parts, or combinations thereof, but do not preclude the possibility of the presence or addition of one or more other features, integrals, steps, operations, components, parts, or combinations thereof. Meanwhile, the term “and / or” as used in this specification includes any and all combinations of the associated listed items.

[0063] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0064] The above description of the disclosed embodiments will enable those skilled in the art to make or use various modifications to these embodiments. It will be readily apparent to those skilled in the art that the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A two-way collaborative air-ground road defect detection method based on BeiDou time synchronization, characterized in that: Includes the following steps: Step S1: Receive BeiDou satellite signals and output PPS second pulses to the vehicle-mounted detection subsystem and the UAV detection subsystem respectively. When the vehicle-mounted detection subsystem performs road surface data collection, it generates a vehicle-mounted time synchronization mark. When the UAV detection subsystem performs structure data collection, it generates a UAV time synchronization mark. Generate a synchronization status mark based on the vehicle-mounted time synchronization mark and the UAV time synchronization mark. Step S2: When the synchronization status marker is available, collect road surface image data and road surface point cloud data and generate road surface distress coordinates, road surface distress type, road surface distress scale and road surface distress change rate. Control the UAV detection subsystem to collect structure image data and structure point cloud data and generate structure anomaly coordinates, structure anomaly type, structure anomaly scale and structure anomaly change rate. Step S3: Determine whether to generate a UAV trigger command based on the type of pavement distress, the scale of pavement distress, the rate of change of pavement distress, and the marking of key impact areas. After generating the UAV trigger command, the UAV detection subsystem performs multi-angle surround observation of the surrounding structures. Based on the abnormal coordinates and types of the structures, generate a vehicle-mounted fine detection command. After generating the vehicle-mounted fine detection command, the vehicle-mounted detection subsystem performs fine mode detection on the associated road sections. Step S4: Receive pavement distress coordinates, pavement distress type, pavement distress scale, pavement distress change rate, structure anomaly coordinates, structure anomaly type, structure anomaly scale, structure anomaly change rate, vehicle-mounted time markers and UAV time markers, perform spatiotemporal alignment of pavement distress objects and structure anomaly objects based on UTC time and CGCS2000 coordinate system, and generate associated bearing capacity; Step S5: Generate cause labels based on associated load volume, vehicle-mounted time markers, UAV time markers, pavement distress type, pavement distress scale, pavement distress change rate, structure anomaly type, structure anomaly scale, and structure anomaly change rate. Generate risk level and maintenance recommendations based on cause labels, and generate an inspection report based on cause labels, risk level, and maintenance recommendations.

2. The air-ground two-way collaborative road defect detection method based on BeiDou time synchronization as described in claim 1, characterized in that: In step S1, the vehicle-mounted detection subsystem receives BeiDou timing signals through the first BeiDou RTK module, and the UAV detection subsystem receives BeiDou timing signals through the second BeiDou RTK module. The first and second Beidou RTK modules receive the same set of Beidou satellite signals and output PPS second pulses respectively, so that the vehicle-mounted detection subsystem generates vehicle-mounted timing markers and the UAV detection subsystem generates UAV timing markers. The cloud-based collaborative system reads the vehicle-mounted time synchronization markers and UAV time synchronization markers, calculates the time deviation from the same source, and verifies the time source number, satellite combination number, and PPS pulse number in the vehicle-mounted time synchronization markers and UAV time synchronization markers. Based on the verification results, it generates a synchronization status marker.

3. The air-ground two-way collaborative road defect detection method based on BeiDou time synchronization as described in claim 1, characterized in that: In step S2, the vehicle-mounted detection subsystem collects road image data and road point cloud data through road surface sensors, generates road surface distress coordinates, road surface distress types and road surface distress scales based on the road surface image data and road surface point cloud data, and generates the road surface distress change rate based on the road surface distress scale and historical road surface distress scales.

4. The air-ground two-way collaborative road defect detection method based on BeiDou time synchronization as described in claim 1, characterized in that: In step S2, the UAV detection subsystem collects structure image data and structure point cloud data, generates structure anomaly coordinates, structure anomaly type and structure anomaly scale based on the structure image data and structure point cloud data, and generates structure anomaly change rate based on structure anomaly scale and historical structure anomaly scale.

5. The air-ground two-way collaborative road defect detection method based on BeiDou time synchronization as described in claim 3, characterized in that: In step S3, the basic data of road structures in the road asset database is read. The basic data of road structures includes the coordinates of the structures, the influence range of the structures, and the mapping relationship of road sections. Key influence area markers are generated based on the coordinates of pavement defects and the influence range of the structures.

6. The air-ground two-way collaborative road defect detection method based on BeiDou time synchronization as described in claim 5, characterized in that: In step S3, based on the pavement distress type, pavement distress scale, pavement distress change rate and key influence area marker matching classification condition table, a UAV trigger command is generated according to the matching result. The UAV trigger command includes pavement distress coordinates, pavement distress type, pavement distress scale and observation radius of surrounding structures. When the UAV detection subsystem identifies the abnormal coordinates and abnormal type of a structure, it generates the associated road segment based on the abnormal coordinates, the area of ​​influence of the structure, and the mapping relationship of road segments, and generates a vehicle-mounted fine detection command, enabling the vehicle-mounted detection subsystem to perform fine mode detection on the associated road segment.

7. The air-ground two-way collaborative road defect detection method based on BeiDou time synchronization as described in claim 1, characterized in that: In step S4, when the synchronization status is marked as available, the cloud-based collaborative system performs spatiotemporal alignment of road surface defects and structural anomalies based on UTC time and CGCS2000 coordinate system, generating co-origin time deviation and open-ground spatial interval.

8. The air-ground two-way collaborative road defect detection method based on BeiDou time synchronization as described in claim 7, characterized in that: In step S4, time numbers are generated for candidate objects within the same spatial window according to the same source time deviation, spatial numbers are generated according to the open-ground spatial interval, time proximity is generated based on the time numbers, and spatial proximity is generated based on the spatial numbers. The system generates type correlation quantities based on the combined pavement distress types and structure anomaly types, and generates correlation bearing quantities based on temporal proximity quantities, spatial proximity quantities, and type correlation quantities.

9. The air-ground two-way collaborative road defect detection method based on BeiDou time synchronization as described in claim 8, characterized in that: In step S4, candidate open spaces are determined based on the associated acceptance volume, and causal labels are generated based on the candidate open spaces, vehicle-mounted time markers, UAV time markers, pavement distress types, pavement distress scales, pavement distress change rates, structure anomaly types, structure anomaly scales, and structure anomaly change rates. Risk levels are generated based on causal labels, associated load, pavement distress scale, pavement distress change rate, structural anomaly scale, and structural anomaly change rate. Maintenance recommendations are generated based on cause labels and risk levels, and inspection reports are generated based on pavement distress coordinates, pavement distress types, pavement distress dimensions, structure anomaly coordinates, structure anomaly types, structure anomaly dimensions, cause labels, risk levels, and maintenance recommendations.

10. A two-way collaborative air-ground road defect detection system based on BeiDou time synchronization, used to implement the two-way collaborative air-ground road defect detection method based on BeiDou time synchronization as described in claims 1-9, characterized in that: It includes a hardware synchronization module, a routine detection module, a bidirectional triggering module, a spatiotemporal alignment module, a diagnostic assessment module, and a report generation module. The functions of each module are as follows: The hardware synchronization module is used to receive the same set of BeiDou satellite signals through the first BeiDou RTK module of the vehicle-mounted detection subsystem and the second BeiDou RTK module of the UAV detection subsystem, respectively output PPS second pulses to the vehicle-mounted detection subsystem and the UAV detection subsystem, and generate vehicle timing mark, UAV timing mark and synchronization status mark; The conventional detection module is used to control the vehicle-mounted detection subsystem to collect road image data and road point cloud data, and generate road distress coordinates, road distress types, road distress scales and road distress change rates. It is also used to control the UAV detection subsystem to collect image data and point cloud data of structures, and generate the coordinates, types, scales and rates of change of anomalies of structures. The bidirectional triggering module is used to classify and determine triggering based on pavement distress type, pavement distress scale, pavement distress change rate, and key impact zone markers. When the UAV triggering conditions in the classification condition table are met, it generates a UAV triggering command, enabling the UAV detection subsystem to perform multi-angle surround observation of surrounding structures. It is also used to generate vehicle-mounted fine detection commands when the UAV detection subsystem identifies abnormal coordinates and abnormal types of structures, enabling the vehicle-mounted detection subsystem to perform fine mode detection on the associated road sections. The spatiotemporal alignment module is used to align the coordinates of road surface defects, structural anomalies, vehicle-mounted time markers, and UAV time markers based on UTC time and the CGCS2000 coordinate system when the synchronization status is marked as available, generating the same source time deviation, air-ground spatial interval, time proximity quantity, spatial proximity quantity, type association quantity, and association acceptance quantity. The diagnostic assessment module is used to generate cause labels based on associated load volume, vehicle-mounted time markers, UAV time markers, pavement distress type, pavement distress scale, pavement distress change rate, structure anomaly type, structure anomaly scale, and structure anomaly change rate, and to generate risk level and maintenance recommendations based on the cause labels. The report generation module is used to generate inspection reports based on pavement distress coordinates, pavement distress type, pavement distress scale, structure anomaly coordinates, structure anomaly type, structure anomaly scale, cause label, risk level, and maintenance recommendations.

Images